Skip to content
Lenny's PodcastLenny's Podcast

Timothy Davis: How Shopify validates new ad platforms fast

Through signs of life tests on 1% Meta lookalikes, Shopify proves a paid channel; agencies kickstart the work, then internal teams own the playbook fast.

Lenny RachitskyhostTimothy DavisguestChristina (OneSchema CEO/founder)guest
Jul 28, 20241h 42mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 2:31

    Timothy’s background

    1. LR

      Is performance marketing just something every company should be doing?

    2. TD

      Hot take, paid is for everyone. If you look at the way each platform is doing, Google, you have to scroll pretty far down to get to an organic listing. Meta, it's almost a pay-for-play now.

    3. LR

      When you take over for an agency at a company, you crush their performance within a month, I'm curious to what you find they are doing wrong.

    4. TD

      (instrumental music) Instead of thinking about being on top of the page, and that's kind of like ego marketing, "I want to be number one, I want to be there all the time," it's about showing to the right person as often as possible.

    5. LR

      Any other tips for people just to experiment with the platform?

    6. TD

      Each platform is different. The user behavior is different. Make sure you're not too hard on yourself if it doesn't work. It's okay to fail, because we're either winning or we're learning.

    7. LR

      (instrumental music) Today my guest is Timothy Davis. Timothy has led performance marketing for all of Shopify for the past two and a half years, and as a consultant, has helped companies like Pinterest, LinkedIn, Redfin, and Eventbrite kickstart and scale their performance marketing teams. In our conversation, we get incredibly tactical on all things performance marketing and paid growth, when to start investing, how to run signs of life test on each platform, what platforms to investigate and what platforms to bet big on, what types of companies are best suited to invest big on paid growth, whether you should invest pre-product market fit or not, what agencies often get wrong, and what to look for in your investment when you're just getting started, plus what your first three hires should look like, tips for which platforms are most interesting right now, a peek at Timothy's actual reports that he runs to judge performance if you're watching this on YouTube, and so much more. This episode is for anyone who's trying to figure out how to kickstart or improve their performance marketing investment, and I guarantee you'll get something out of this that'll make your life better. If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube. It's the best way to avoid missing future episodes, and it helps the podcast tremendously. With that, I bring you Timothy Davis. Timothy, thank you so much for being here, and welcome to the podcast.

    8. TD

      Yeah, thanks for having me. A long time coming.

    9. LR

      Yeah. I'm really excited we're finally doing this. I actually posted on Twitter for people to suggest questions for this topic, and there's just so much interest in what we're gonna be talking about, which essentially we're gonna be diving deep into all things paid growth, all things performance marketing. Uh, by the way, I have a cold, for people with, in case you wonder why I sound a little weird, but the sh- show must go on. I want to start with just setting a little context for folks that aren't super familiar with performance marketing,

  2. 2:31 – 6:31

    Understanding performance marketing

    1. LR

      paid growth. When we talk about performance marketing/paid growth, first of all, are those two terms interchangeable to you? And second of all, what falls under the umbrella of performance marketing/paid growth?

    2. TD

      Paid, uh, can be a lot of things. It can be online, it can be offline, it can also even be affiliates. Typically, when I talk about performance marketing, it's about online only, but you could argue it's, like, offline, could be performant, affiliates could be performant, stuff like that. Uh, they can be interchangeable, but I would recommend if you are talking about it to kind of specify which one you're talking about, because when you do say paid, because I'm in the industry, I would just, "Oh, they do Google search and they do Meta and they do things like that." But if you do offline and you just say paid, may- ma- we may be, uh, talking about things but missing each other on two, two ships passing in the night. So, I, I think historically, uh, people when they have said paid growth have been fully focused on just online.

    3. LR

      Just online. Okay.

    4. TD

      But I would argue-

    5. LR

      Got it.

    6. TD

      ... in the last couple of years, you could definitely roll, uh, offline into that conversation as well.

    7. LR

      Okay, so performance marketing, when someone hears that term, it's essentially marketing that you can measure the performance of.

    8. TD

      Correct. Yes.

    9. LR

      Nailed-

    10. TD

      Nail on the head.

    11. LR

      (instrumental music) This episode is brought to you by buildbetter.ai. Back in 2020 when AI was just a toy, Build Better bet that it could cut down on product teams' operational BS. Fast-forward to today, 23,000 product teams use purpose-built AI in Build Better every day. First, Build Better uses custom models to turn unstructured data, like product and sales calls, support tickets, internal communications and surveys, into structured insights. It's like having a dedicated data science team. Second, Build Better runs those structured insights into workflows, like weekly reports about customer issues, context-aware PRDs, and user research documents with citations. It even turns stand-ups into action items that automatically get assigned and shared into your tools. Plus, with unlimited seat pricing on all plans, Build Better ensures everyone at your company has access to this knowledge. Truly, no data silos. In a world of AI demos overpromising and under-delivering, see why Build Better has a 93% subscription retention. Get a personalized demo and use code Lenny for $100 credit if you sign up now at buildbetter.ai/lenny. I'm excited to chat with Christina Gilbert, the founder of OneSchema, one of our longtime podcast sponsors. Hi, Christina.

    12. CC

      Yes, thank you for having me on, Lenny.

    13. LR

      What is the latest with OneSchema? I know you now work with some of my favorite companies, like Ramp, Vanta, Skale, and Watershed. I heard that you just launched a new product to help product teams import CSVs from especially tricky systems like ERPs.

    14. CC

      Yes. So we just launched OneSchema File Feeds, which allows you to build an integration with any system in 15 minutes as long as you can export a CSV to an SFTP folder. We see our customers all the time getting stuck with hacks and workarounds, and the product teams that we work with don't have to turn down prospects because their systems are too hard to integrate with. We allow our customers to offer thousands of integrations without involving their engineering team at all.

    15. LR

      I can tell you that if my team had to build integrations like this, how nice would it be to be able to take this off my roadmap and instead use something like OneSchema, and not just to build it but also to maintain it forever?

    16. CC

      Absolutely, Lenny. We've heard so many horror stories of multi-day outages from even just a handful of bad records. We are laser-focused on integration reliability to help teams end all those distractions that come up with integrations.We have a built-in validation layer that stops any bad data from entering your system. And OneSchema will notify your team immediately of any data that looks incorrect.

    17. LR

      I know that importing incorrect data can cause all kinds of pain for your customers and quickly lose their trust. Christina, thank you for joining us. And if you want to learn more, head on over to OneSchema.co. That's OneSchema.co.

  3. 6:31 – 8:39

    The importance of paid search

    1. LR

    2. Is performance marketing just something every company should be doing, or is there certain business models where it's like, no, you're probably gonna grow through other channels mostly, like SEO, or sales, or word of mouth?

    3. TD

      Hot take, uh, I would say paid is for everyone. Uh, I, if you look at the way each platform is doing, Google introducing AI plus with paid taking up, like, the first four spots, you have to scroll pretty far down to get to an organic listing. Meta, it's almost a pay for play now. You have to do promoted posts for people to even see your content. I would say it does depend on the industry you're in. If you're, say, leaning heavily into influencer marketing, that is still a paid component. You may not be doing Meta ads, you may not be doing paid search, but that is a paid component. But I would say at baseline, everyone should be doing paid search. Uh, the way I usually explain it to people is, paid search is user driven. You have to type in a relevant keyword for your ad to show. Anything else is more disruptive media. Uh, you're on Meta and you're looking at pictures of babies and cats, and then all of a sudden you see an ad for Shopify. You're watching a YouTube video, maybe a YouTube podcast, and then in between, you get advertisements, maybe for Shopify, maybe for Pinterest, maybe for Eventbrite. So, all of that being disruptive media will... may not make a whole lot of sense for where you are as a business, but I would say paid search being user driven, having to type in a relevant keyword for your ad to show, should pretty much be for just about everybody.

    4. LR

      Everyone should do it. Super interesting. There's also the, most companies grow through one growth engine primarily, say word of mouth or SEO or sales or paid. So, for some companies, paid will be most of how they grow. Like say, I always think of Booking.com and Credit Karma, and just, like, most of their growth came from paid growth, online paid growth. For some companies it's just, like, a layer, like a small, 10, 20% of

  4. 8:39 – 9:54

    Identifying growth potential

    1. LR

      their growth. What are signs that you have the potential for paid/performance marketing to be most of your growth, like say 70, 80%?

    2. TD

      I always say, where, where are your users? Uh, you will have some data that will allow you to understand that. Say right now you're doing really well on TikTok. Some would argue that's an emerging channel right now. Great. Uh, if that's doing really well for you, maybe you take... because that content could be used on something similar like Snap, see how that does for you. You could also lean into your Google Analytics or whatever analytics tool you are using and see if you're already getting users from that platform, if you're currently managing that. If you don't have a profile on there and obviously you don't have a presence, it would be really hard for users to find you and get there. But always look at the data that's available to you within your analytics platforms and say, "Users are already finding us here. How can we turn that knob up to 11?" And you can do that with paid, because if they're already finding you through that channel, what would happen if you were to just turn it up to 11?

    3. LR

      Mm-hmm. Is there an example of a company you worked with that's known that... as an example that where you're like, "Oh, I see, everyone's looking, finding them on, say, Google. Let's go turn it to 11"?

  5. 9:54 – 12:22

    Case studies: Hairstory and Ipsy

    1. LR

    2. TD

      Yeah. So there was a company I worked with a couple years ago called Hairstory, uh, and Ipsy. I know, very funny, guy with no hair working on a company called Hairstory. But, uh, they were doing really well from a Google Shopping standpoint. But when I started consulting with them, we looked at the analytics and we actually saw that they were getting a lot of people from Meta and, uh, TikTok at the time. And TikTok was very, very new, uh, and they didn't really know what to do with it. So said, "We'll do a small test. We can start on Meta with just some, you know, uh, customer testimonials. Let's see how that does. We'll get interest in, build a funnel, get retargeting, and hopefully get conversions." TikTok was so new at the time that it was like, "We don't know what's gonna work. (laughs) Let's just try with what, what is currently available." Then once we started doing it, we started noticing, hey, this doesn't make a whole lot of sense from a creative standpoint. We were kind of missing the mark of what we were using on Meta was working, was not working on TikTok. So you can't always just take what is currently working on one platform and apply it to another, 'cause it is a different user experience. It is a different mindset that the user has when they're on there. So, that was an example of where we were able to take those instances, look at their, uh, data and say, "They're already finding us here. How can we turn this up a little bit?" And I don't remember the numbers exactly, but it was pretty exponential.

    3. LR

      Okay. So the, the core advice here so far is, look for where people are coming to you from today, and then that gives you a sign of where you should start to think about running paid ads, performance marketing on those platforms.

    4. TD

      Yeah. And also, uh, there's a saying that, um, my current director likes to use, which is, signs of life. You can always do a very, very small test. You can just put a little money into a platform, see if there's a sign of life. If there is, then you can s- you can pull back and say, "Okay, we have signs of life. Now let's build a campaign around that." There's no reason to say, "All right, we have a signs of life. Let's now turn it up all the way. Do we have the right creative? Do we have the right campaign? Do we have the right messaging for the users on this platform?" Then let's take, uh, that approach as opposed to just, "Hey, signs of life. Great. Go. Run 100, uh, miles an hour."... make sure you're doing the right thing when you get into those platforms.

  6. 12:22 – 18:57

    Experimenting with new platforms

    1. TD

    2. LR

      Let's pull that thread. A lot of people run little experiments on, say, TikTok and Snap and Twitter and things, and LinkedIn, and they often don't see a lot of results, and it's always like, "Hey, did we do it wrong? Or is it the platform's not working?" I know it's, like, a very difficult question to give, like, here's how you do a sign of life test correctly, but what are some things that you think either people do often wrong when they're trying to experiment with the platform, or that you just think they should do when they're trying to look for signs of life?

    3. TD

      The number one thing I tell people all the time is, use your own data. Start with your own data. So take the existing customer base you have, load that up into the platform to build lookalikes. From there, you can do ... and I'll use Meta as the example. You can do a 1% all the way up to a 10%. So a 1% match, 2%, 3%, so on, so forth. What I tend to do is build ad sets, uh, one at 1%, because we know that one's going to be highly correlated to what we're looking for. Then a two to four, and then a five to seven, and then an eight plus. More times than not, the eight, eight plus does not work. But there have been times where it has, so it's just, "Hey, let's do it. Let's try it." But if you also have a very limited budget, because some people don't have the luxury of having an un- unlimited budget like we do in, uh, some of these other companies I've worked at, uh, just start with the 1%. See what that sign of life is. Because you already know that this is so tied to your existing customer base. If that sign of life is giving you a positive, uh, signal, now you have the information you need to build a campaign you need to be successful.

    4. LR

      What are these percentages referring to? You may have mentioned it, but wh- when you say 1%, 2%, 3%.

    5. TD

      Yeah, yeah. So the, the, that is how closely tied they are to that user's behavior on the platform. So they're 1% tied to what they're doing. They're visiting similar pages or they have similar behavior on the platform, so they're more than likely going to be tied to what that user ... They're not gonna be that user exactly, but they're gonna be very correlated to them. Where 10% is pretty wide base (laughs) that you're gonna hit.

    6. LR

      I see. So the smaller the percentage, the closer they are, the closer the of a lookalike they are.

    7. TD

      Yep. There you go.

    8. LR

      How do you know if it's the creative that is failing versus it's just, like, never gonna work? Is there something that tells you that's what's not working?

    9. TD

      It's a million-dollar question-

    10. LR

      (laughs)

    11. TD

      ... to be honest with you. (laughs) Uh, because the, uh, the ... You could look at some of the metrics that currently exist in the platform. Click-through rate basically just tells you are the users engaged, because start with the target. You're getting the impressions are there. Is the target right? That should be your first thing. Yes, the target is correct. Users that we want to see are seeing our ads, but they're not engaging with it. That would kind of be the first thing. But the, the part that I think you're really trying to ask is creative versus the content. Because sometimes, depending on the ad unit, you could have, like, a story creative that is literally the creative kind of has to stand on its own, not necessarily the content that's on it. But something like an in-feed creative has a headline, a primary, a description, and a creative. That's where ... Honestly, the only way you're gonna get an answer to that outside of the data that's available to you, like click-through rate, reach frequency, and all those things to kind of pull in more metrics, is a focus group to understand, "Hey, when you saw this image with this content, you know, what was your result from it?" Now, you can build a test within the platform to do a control versus a test where, say, you take the creative with the same primary and the same description versus a creative that's just your logo with the headline and the description, and see what the results are. But more times than not, you're not getting the understanding from the user of, well, why doesn't that creative resonate with you? And they'll give you suggestions in the focus group. Why does that work better for you than not? Um, so that's a really, really hard question to answer without having those dialogue with the user to understand that.

    12. LR

      Yeah. That makes a lot of sense. How long do you recommend these tests run for, these signs of life tests? And I like this term, by the way. I haven't heard this before.

    13. TD

      Ideally, in a perfect world, you're gonna hear the words statistical relevance. Um, mathematicians will, (laughs) will tell you exactly what that number is, and I'm not gonna pretend like a mathematician, but a lot of times it's a budget constraint. So there, there could be a VP finance, anyone like that could just come to you and just say, "Hey, you have $25,000 for this test. You only have $5,000." Uh, you know, just get that information, get that data. You can then also build an op-size off of that. "Hey, guys. We work with our Meta partners, we work with our Google partners. Uh, we put $1,000 behind this test. We know that our impression share was this, our reach was this. If we were to put $10,000 into it, this is the expected return we can get from it, uh, based on the histor-, based on the click-through rate, the conversion rate that we got from the tests that we ran."

    14. LR

      Got it. So basically, you, you have a limi- limited number of dollars to spend. That'll tell you how long you could run one of these for.

    15. TD

      Yeah.

    16. LR

      Any other tips for someone that's trying to experiment with a platform early on for these signs of life tests? So one is, uh, try a lookalike that is v-, like the most targeted version of the lookalike, so 1% you said, and then kind of incrementally grow and see how that ... incrementally increase that percentage as, as, as you spend more. Any other tips, I guess, for people just to experiment with the platform?

    17. TD

      Make sure you're not too hard on yourself if it doesn't work. A lot of times, companies I've either consulted with or worked at, there was always pressure or there was always a need to succeed at whatever experiment we put into market. I believe in creating an environment where it's okay to fail, because we're either winning or we're learning.And more times than not, if we put something new into a platform that we know nothing about, we're learning the bells and whistles. We're learning the functionality of it. Just know there's going to be things you're not going to understand because each platform is unique, each platform is different. The user behavior is different. So give yourself some grace. Understand that this may not work, and as long as you're learning from it, you're going to be okay.

    18. LR

      I love

  7. 18:57 – 20:23

    Choosing the right platforms

    1. LR

      that. In terms of which platforms, hmm, people should explore. Obviously, there's Google, there's Facebook, Instagram. What's kind of like, what are the platforms people should seriously consider at this point?

    2. TD

      Google for sure. And when we say Google, let's make sure everyone understands what G- Google entails. Google is YouTube, Google's Google Search, GDN, Google Display Network. There's a lot of things in Google that you can do. I always say if you have the creative available, you should really, really be looking at doing video, because video is one of those things that I am very bullish on. It is doing really well, um, when you measure it the right way. And as long as you can get that creative and get it consistently, just 'cause you have one piece of creative, if it's doing well, you do need to have that creative refresh, kind of that flywheel going. Make sure you're doing Google Search, again, user-driven, YouTube and Meta, and Meta contains both Facebook and Instagram. And then based on the data available to you, if you do see users on TikTok, definitely go after those. But if you're just kind of starting out and you're, just want to get your feet wet into something, I would say start with Google Search, then get into Meta. Uh, and if you have video available, definitely get onto YouTube.

    3. LR

      Awesome. Okay. So Google Search, YouTube, Facebook, Instagram, TikTok.

  8. 20:23 – 27:56

    LinkedIn for B2B marketing

    1. LR

      What about LinkedIn? Do you see things happening there for, say, B2B companies?

    2. TD

      Yeah. Uh, L- LinkedIn, um, is very expensive in comparison. So, just know if you're going into LinkedIn, it's going to look almost three times more expensive than other channels. The targeting that's available on LinkedIn, knowing job titles, industries, actually targeting people at certain companies is very powerful. The example I always tend to use is that, uh, I worked at a company called SoftLayer. SoftLayer was acquired by IBM because they were trying to build out their cloud portfolio, couldn't do it. They're like, "Hey, let's do the next best thing and buy this company." But when we were just SoftLayer, we were trying to get, uh, Coca-Cola as one of our clients. And, uh, you know, those Coke Freestyles where you can pick whatever drink you want and put the flavor in it?

    3. LR

      No. That's not

    4. NA

      It's delicious, yeah.

    5. TD

      Oh, man, these are great. Uh, may- maybe, maybe it's a South thing, but...

    6. NA

      (laughs)

    7. TD

      Uh, basically, you go up to a Coke Freestyle and say, "I want Dr. Pepper and I want cherry flavor in it." Uh, or you want Coke Zero with cherry and vanilla in it. You can do that. Uh, it's kind of the Freestyle, you get to pick. The reason they wanted, uh, it to be cloud-based is because there's a ton of drinks in there. So you need to be able to efficiently and effectively say, "This store needs more Sprite," or, "This store needs more Sprite Zero," uh, because you have your whole portfolio in this one machine. So they want to be able to update things more readily. Uh, and it was between us, AWS, and Microsoft. And met with the sales team and said, "You know, wha- what do we nee- need to do to win this?" Because it was a big-ticket item for us. And, uh, what we found out was they had two concerns about us. It was recent- recency and private, um, security. So what we did was we found out where the decision-makers were. So A ) In LinkedIn, we took the, anyone who works at Coca-Cola, we want you to see ads about security and recen- recency. Outside of that, we also geo-fenced it, because you would think Coca-Cola, Atlanta, that's where the decision-maker sits, but they were actually in LA. So in LA, we geo-fenced them, and we knew that they worked at Coca-Cola. The next time the salesperson got on the call with the team to say, "Hey, just wanted to check in," and he was like, "Hey, hey, I get it. Security's fine, recency is fine. We hear you." Um, that was great to hear because it was something we were able to do on LinkedIn, and we did it in other platforms as well. But we knew we were able to get not only to that decision-maker, but everyone around that decision-maker to say, "Hey, these, these guys may be the ones we need to go with."

    8. LR

      Okay, so just to mirror back what you're saying. You're saying you ran ads on LinkedIn, uh, targeting the execs at Coca-Cola, trying to influence them to overcome these barriers they had to buying, to working with you guys. Uh, amazing. (laughs)

    9. TD

      Yeah.

    10. LR

      I've seen people do that on Google Search, but, uh, I've never heard of that. It makes so much sense to do on LinkedIn, and it worked. And did they know that you did this, or they're just like, "Oh, I just kind of changed my mind"?

    11. TD

      Uh, no, uh, w- one of them was like, "We get it. We're seeing it everywhere." We actually tried, because at the time we didn't have an offline team, we actually tried to buy the billboards around the office as well. But I didn't know enough at the time from an offline perspective of like, the right people to talk to, how...

    12. LR

      Wow.

    13. TD

      ... how long it would take to do it.

    14. LR

      Yeah.

    15. TD

      So, uh, we w- we were trying to pour it on hard.

    16. LR

      That's incredible. Like, I could see why LinkedIn is more expensive. That's incredibly powerful. And...

    17. TD

      Yeah.

    18. LR

      ... uh, is the advice then that LinkedIn makes more sense for like higher LTV, higher CV type of products?

    19. TD

      Yeah, exactly. Uh, I would not recommend getting into LinkedIn, and my LinkedIn reps may kill me for this. I would not recommend getting into LinkedIn until you've, uh, tested Google and Meta first. Now, you could be an enterprise level company, and it does make sense to start with LinkedIn, for sure. Uh, but...Depending on your audience, I would say more times than not it would be a Google, Meta discussion before going to LinkedIn.

    20. LR

      Okay. So just generally, start with Google and Meta. Is there one or the other usually? Is it like... I guess, yeah. Do you s- which one would you start with if you had to pick one?

    21. TD

      Yeah. Always Google Search. I'm always gonna say start with Google Search. Uh, but it, it's also creative dependent. If you don't have the right creative for Facebook, it's gonna be really hard to convert, because users are just gonna either be turned off by what you're kind of putting out there, and then also depending on where your user base is. If they're not on Facebook, it, it, it's kind of a moot point. But if they're browsing around on GDN or YouTube, it makes more sense to go there.

    22. LR

      And then you said this insight about video is performing super well right now, and you're recommending people use video. And the key there is you need to be able to make videos, video ads.

    23. TD

      Yeah. You gotta make sure you have that flywheel.

    24. LR

      And the flywheel is just people internally, or some company, agency that can make video ads for you?

    25. TD

      Correct, yep.

    26. LR

      Cool. I guess then, is there any advice there of just how, how to do that, or how, what ads work well or anything there for someone else ?

    27. TD

      Yeah. So, yeah. The, the thing with YouTube is, uh, I always say start with emotion. If you can have an emotional connection with a user, it is gonna be way more impactful and way more powerful than, uh, anything else. And to, to take that further, when I say emotion, that, I'm talking about comedy. I'm talking about, you know, um, hap- happiness. It's not necessarily just, "Oh, we need an emotional connection with the brand." It is just making sure the user feels something after seeing your ad, uh, because then they'll, they're more, more times not going to remember it, which then they will take a favorable action after the fact.

    28. LR

      That is super interesting. One last question about platforms. Are there any other platforms you see working for people that are emerging maybe that people aren't thinking about? Reddit or Snap or X or, I don't know, anything along those lines?

    29. TD

      When, when, when you say emerging channels, my brain kind of goes to connected TV, podcast, VR advertising, uh, audio voice search, and even AI stuff right now. I can tell you podcasts are doing really well for the people that I know are able to measure it correctly. Right? Um, connected TV is also doing really well. You can take the, if you, again, you're doing YouTube right now, being able to take that creative and repurpose it. But the, the VR stuff and the AI stuff, that is stuff that is, I would say very emerging (laughs) right now, because, uh, to, to be completely transparent, I haven't experimented with those things yet. Uh, I kind of like letting other people be the guinea pig and learning from them. And then, uh, if, hey, someone comes to me and they're like, "Hey, we have budget for this. Let's go ahead and test it," great. Let- let's go. Let's see what we can learn.

    30. LR

      Awesome. Uh, podcast ads, I'm glad you, uh, suggested that. I'm a huge fan. Thank you. Uh, they're working really well for a lot of our sponsors. And, uh, I'm very biased, so you don't need to pay attention to me. But, uh, I also think there's a lot of opportunity there. Okay. Let's talk about when to start investing in performance marketing/paid

  9. 27:56 – 33:33

    When to start investing in paid growth

    1. LR

      growth. So say you're like a startup.

    2. TD

      Mm-hmm.

    3. LR

      Do you have any advice for when it's time to start, signs of life test or even... And then also just when is it time to scale, to go like, "Okay, let's go big on this"?

    4. TD

      If you are a startup, um, typically whenever I consulted with them, it was, what are the goals? What are we trying to achieve? And when? Because if you're looking for something quick, paid needs to start immediate, and pages that started yesterday, because SEO takes time. SEO, depending on the market you're getting into, if it's emerging, people may not even know your product exists. Uh, there was a product I was working on years ago, uh, that if you were traveling to a hotel and say it was a romantic getaway with you and your partner, and you wanted the hotel, uh, suite or bed, um, room done up with like flower petals and champagne and stuff like that, uh, they would do it for you. Well, they were working with someone before that was like, "Oh, we should totally be in paid search." And if you look at the keywords that they were doing, it was like booking hotel rooms. It's like, no, that's, that's a complete disconnect. And they're like, "Yeah, but we're trying to build awareness." You don't wanna build awareness through search. You build awareness through display, media-based type media. When we transitioned all of their money over at the time, and this will age me, before Meta was a thing, when we transitioned everything over to GDN, that's when the company started really reaping the rewards, because we were building the awareness beh- around the product. So it depends on how, you know, what is the demand of the prod- of your product and market? If you're doing something that is similar to another, uh, product and market, you could do competitor. Um, uh, I call it coattail riding.

    5. LR

      Hmm.

    6. TD

      Uh, say you have, say, Lenny, you and I create a product that is similar to monday.com. We can go out there and just start bidding on monday.com and say, "See why, you know, Lenny and Tim are better than monday.com," and start getting some people in interested, maybe kicking the tires, starting free trials. But they also have other keywords that they can go after in market. Uh, but if you're going after something new, like hotel room, you know, flowers, and I forget what keywords we were bidding on for that, 'cause it was years ago, people weren't thinking of that. That wasn't something they were searching for. So it just really depends on, A, when do you expect results? (laughs) 'Cause SEO can take time. And B, what is the demand in the market for your product?

    7. LR

      Mm-hmm.Some people use paid ads to drive early growth to bring customers to help them figure out what to build, kind of pre-product market fit growth. Is that something you recommend? Is that something you advise against? Is that a strategy that you've seen work?

    8. TD

      Yeah, product market fit is a huge thing. Um, we've run into some of this at, uh, Shopify, and we definitely ran into this at, um, IBM. So for the longest time, you know, IBM is, "We're a global company, we should be everywhere." Uh, but when we started looking at the data, it was, "Should we be everywhere?" For example, Africa. We were in the whole continent. We weren't just in, say, South Africa or Egypt or something like that. And the more we dug into it, the more we realized the biggest issue wasn't necessarily that users weren't interested in our product and weren't able to... and weren't purchasing it. It was because we didn't have a product market fit, mainly from an operation standpoint. In Africa, there are multitudes of different types of, uh, currency. There's the franc, there's the rand, there's the shill, and we, we were just, "Hey, USD. Thank you." (laughs)

    9. LR

      Mm-hmm.

    10. TD

      So of course they weren't converting, unless they were going out of their way to make a way to, uh, convert that. So when we s- took a step back and we said, "Okay, where do we want to start?" it was South Africa. That's really where we were trying to get, quote unquote, stronghold in. Uh, so we had to make sure we had the rand available. Once we did that, we were, uh, we started doing tremendously better 'cause then we had a product market fit. They, they had demand for our product, and now we were able to serve them, meet them where they were.

    11. LR

      So essentially you're saying probably not smart to run a bunch of ads if it's not working, if you don't have something people actually want yet in that produ- in that market.

    12. TD

      Right.

    13. LR

      Okay.

    14. TD

      Or able to convert.

    15. LR

      It... Because conversion is gonna end up being really low. No one actually cares about what you're doing.

    16. TD

      All you're gonna do is really annoy the users, 'cause say in the future they are, "Hey, I'm, I'm still interested, but I don't wanna use that product 'cause I already tried." It's like, "No, no, you can totally use it now." "I already tried. It's, it's, it... I had a bad experience with them," and some users will hold that, you know, you accountable to that. "One bad experience, and I'm just never giving you my business."

    17. LR

      Interesting. So is that, is that generally your advice? If you're a startup, you're not feeling like you actually have product market fit yet, should you even experiment with and do signs of life tests? Or should you hold off until it's like, "Okay, it's actually working, let's go"?

    18. TD

      From an operation standpoint, I al- I always wanna make sure those things are tied off. Yeah, it i- it doesn't make sense to... A- again, if you have a... if you have a major budget and you're trying to get awareness into a market, great. Yeah, you can go after a market. Uh, but just know that it may... it's probably not gonna convert very well.

    19. LR

      Got it. So-

    20. TD

      I wouldn't recommend it.

    21. LR

      Okay, great. Awesome. Very, very

  10. 33:33 – 37:41

    Common mistakes in performance marketing

    1. LR

      clear answer. Let's talk about the mistakes that you see companies make when they're investing in performance marketing. So, uh, we got introduced through Casey Winter's, an illustrious former podcast guest, two-time podcast guest, and asked him about you. And he told me that when you take over for an agency at a company that you work with, you crush their performance within a month. I'm curious to what you see and find they are doing wrong that allows you to be such a hero when you come in and take over.

    2. TD

      I think agencies have playbooks. Uh, now, I've worked at agencies. I, I did consulting for a long period of time, and it was never an approach I took. I looked at each account and each company as its own thing. But I think a lot of agencies just come in and they go, "Oh, this is like A, B, and C company, and this is what we're doing. Copy, paste, done, move on." And they're also not willing to get deep, deep, deep into the weeds of stuff. Um, I may get a little too far into data, (laughs) uh, than, uh, some other people. For example, like I have, uh, ops cadence that, you know, myself and my team follow. So within that ops cadence, we have things like finance, performance, structure, keywords, ad copy, quality score, targeting, et cetera, et cetera. And then withi- within each of those, we have specifics. For example, the keyword subsection is keyword granularity, brand versus non-brand, search query reports, negatives, so on and so forth. I feel like agencies don't necessarily get into all of those things every single month, where some of those we're doing weekly, some of those we're doing bi-weekly, and some of those we're doing monthly. But they touch the things that they think they need to touch. They turn on automated bidding, they're doing their search query reports, and then they just kinda move on with their day, because they have 50 other clients they have to get to. "Well, what about looking at your, your conversions? Where are we converting? Have we tested different landing pages? What is a better user experience that we can be getting users? Right now, it's a five-step process. Can we get it to a three?" Partnering with, like, PMMs to say, "Hey, here's something that I think could help improve our, you know, lead to conversion." Uh, just stuff like that, that they're, you know, too busy with too many other things to focus really, really deep into those. And typically, whenever I managed an agency in the past and I consulted, I made sure that you didn't... you weren't stretched thin enough that you couldn't do those things, because I'm a firm believer in hiring smart people and then getting out of their way, but each week having one-on-ones with them, just, you know, spot-checking things. Just saying, "Hey, I looked at this. This doesn't look right. What's going on there?" "Oh yeah, I do that on this day." "Okay, great. Just making sure you're covered." Because sometimes people won't scream uncle when they should be screaming uncle, 'cause they think it's a sign of weakness. But...... let me know when you're overwhelmed. Maybe I have a solution for you. Maybe there's a way I can coach you to be better at something. Or we just need to hire more resources, because the client portfolio you have is five, and when we took them on, they were all at 100K but now they're all spending two million plus. We need to offload some of that from you, because you're doing such a good job that you've scaled them up, let's give you three clients instead of five and hire someone to take those two.

    3. LR

      So it sounds like you, basically they just don't have the time to care and spend on all the things that they need to be doing.

    4. TD

      Yeah.

    5. LR

      And when you've helped companies, you actually go deep and you don't have a lot of ... you, you have the time to do it well. I guess when someone's trying to find an agency or someone like you, I know you don't do this much anymore, any advice for just like h- how to know if they're gonna be great? Is it just like, agencies in general are probably not a good choice? Is it hi- hire someone in-house? I guess, what, what advice do you have for people that are like, "Oh man,

  11. 37:41 – 40:36

    Working with agencies and consultants

    1. LR

      I don't ... I want to avoid this."

    2. TD

      Agencies are a good place to get things started. Even when I was consulting, I would honestly ... I'm a big believer in forward thinking, backwards planning. So if I'm taking this contract on with this new client, it's wh- what's the end goal for you? If your end goal is to have this at, you know, uh, performant, $100 million spend, there's no way one person can do that. So let's create milestones along the way of, uh, making sure we're checking in and saying, "When is the right time to hire?" Uh, whether that's a data scientist, whether that's a creative person in-house, or replacing me full-time. There's no reason that you should be holding a company back. If anything, you should be helping them get to that milestone. And I think that's why I've been able to do so well for myself, because I show that, hey, I have the best interest for you and your company, uh, as opposed to how much money, how much more money can I squeeze from you by holding onto you as a client? I would say agency/consultant, um, is a good way to start, because if you, as a business owner, you shouldn't have to log in ... oh, I haven't logged into Google in like three weeks to look at stuff because I'm doing payroll, I'm doing HR stuff, and I'm meeting with clients and I'm doing sales. Start there, get it to a good place, create that milestone of, hey, when we're spending 50K a month, I need to hire somebody full-time to take this over. And just have that conversation with your agency and your consultant. Uh, I think more times than not you will have a positive reaction to that, but if you don't, I would say that's a major red flag and probably someone you shouldn't be partnering with.

    3. LR

      Awesome. Okay. So your advice is to get started on paid growth, find an agency or like a small shop consultant type person to get you started, have a conversation, and when we reach a certain scale, we're gonna hire someone internally to run this for us. And then potentially they'll keep working with you, potentially they'll start things on their own.

    4. TD

      Yeah.

    5. LR

      Or, or is that ... Or is it ... Is it, is the assumption they'll start working with you, they'll become the owner of this thing? Or is it like, we may transition you out?

    6. TD

      That's a really good point, actually. So, uh, there have been times where, uh, a client's like, "Hey, we've reached our milestone, I want to bring someone in." That person comes in and say they're an expert in, uh, social media. Like, "Hey, I still want you to execute on Google and Bing, because that is just not my bailiwick. I'll take over the social stuff, but we do need to talk about your fee. Maybe it does need to go down."

    7. LR

      Mm-hmm.

    8. TD

      Completely acceptable. Let's have that discussion, make sure we are both aligned to what that should be. Uh, but yeah, th- there have been times where they're like, "Hey, we've reached our milestone, we gotta bring someone on," but that person that they bring on is like, "Actually, I see expansion in this direction, but I can't do this. Are you willing to stay on and help me with this?" Great. If not, I'll just have to find

  12. 40:36 – 47:33

    Hiring for performance marketing

    1. TD

      someone else who can.

    2. LR

      I want to talk about the team that you s- build over time, but later. But specifically for this first person that you hire, what sort of person is this person? Is it like a data person? Is it a person that's just done specific performance marketing on channel? Or what do you look for, ideally?

    3. TD

      I'm a big believer in, uh, th- there was a, a book written by, uh, Nate Silver called The Signal and the Noise. Are you familiar with it?

    4. LR

      No.

    5. TD

      Um, so, n- uh, are y- uh, so Nate Silver is the guy-

    6. LR

      Yeah, FiveThirtyEight, FiveThirtySeven.

    7. TD

      ... who created ... Yeah, yeah, yeah. Yeah. Um, and in the book, it, he basically, uh ... I'm gonna use the word art. I'm gonna use the word art. Detailed the art of like probability, statistics, um, uh, and applied them to like real world, um, circumstances. It included case studies of, with baseball, which of course I loved, um, elections, climate change, poker, stuff like that. And the ... I like that book. It's kind of dense, but the thing I liked the most about it was the title, and just changing the title ever so slightly to Signal Not Noise. So typically, whenever I hire people, I want to hire smart people and kind of get out of their way, but the biggest thing I want to focus on is, what is your thought process when it comes to data? Because I can teach anyone how to do Google Ads, I can teach anyone how to do Meta ads. That, that is not the hard part. It is the data part that is the hard part, because there is so much noise going on in those accounts. They give you everything, which is great. (laughs) It's great that they give you all this information, but you could have someone that's, "Oh, but look at the reach, look at the frequency, look at the CPMs, look at the CPC, look at the conversion rate, look at the cost per lead, look at the cost per MQL, look at this." Hold on. (laughs) That's a lot of noise you just said. So let's ... What is the signal and what is the noise? And let's make sure we're focusing on the right signal versus the right noise, and that has to be a data person because there's a lot of data in these platforms.

    8. LR

      This episode is brought to you by Eppo. Eppo is a next-generation A/B testing and feature management platform built by alums of Airbnb and Snowflake for modern growth teams. Companies like Twitch, Miro, ClickUp, and DraftKings rely on Eppo to power their experiments. Experimentation is increasingly essential for driving growth and for understanding the performance of new features, and Eppo helps you increase experimentation velocity while unlocking rigorous deep analysis in a way that no other commercial tool does. When I was at Airbnb, one of the things that I loved most was our experimentation platform, where I could set up experiments easily, troubleshoot issues, and analyze performance all on my own. Eppo does all that and more, with advanced statistical methods that can help you shave weeks off experiment time, an accessible UI for diving deeper into performance, and out-of-the-box reporting that helps you avoid annoying prolonged analytic cycles. Eppo also makes it easy for you to share experiment insights with your team, sparking new ideas for the A/B testing flywheel. Eppo powers experimentation across every use case, including product, growth, machine learning, monetization, and email marketing. Check out Eppo at geteppo.com/lenny and 10X your experiment velocity. That's geteppo.com/lenny.

    9. Everyone I talk to that's worked with you is just, like, Tim- Timothy is incredible at working with tons of data and finding the things that matter. Uh, someone told me that it actually comes from your love of baseball, that you just go crazy with stats and spreadsheets of baseball games and players and things like that. Is there something there that you could share that is interesting?

    10. TD

      Yeah, I... So, I played baseball my whole life, uh, but when you, um, blow out your arm and you're as fast as a sea otter, it's, it's not very conducive for you to continue facing, uh, you know, chasing that dream. But the one thing I always liked was the stats. Even as a kid, when I would watch, like, baseball games, uh, on TV, it's like, "What's this guy batting? How often is he getting on base? Uh, you know, how many RBIs? Like, oh, this guy has terrible average, but he has a lot of RBIs. That means if there's a runner on base, he may not get a hit, but he's gonna get that guy in." Um, so you want to make sure you put them in the right spot in the batting order, so on and so forth. I could talk about that for hours. But, uh, again, that, that could be another thing where there's a lot of data available. What... Which one should you focus on and, uh, ultimately where... what is the optimization you're going to make to the lineup or to the, uh, where you place fielders on the field to ensure that we're, we're doing the right thing and becoming as efficient and effective as possible? As far as dealing with all that data, that is just, I would say, a skill I've learned over time. Uh, early in my career, uh, whenever someone, uh, you know, was like, "Oh, we have this new project, uh, who wants to work on..." "I'll do it." I, I was always just eager, you know, like a sponge, and just I wanted to get in as much stuff as possible. I did SEO for a portion of my career, I did email marketing, I did affiliates, and then eventually, I remember the first time I saw paid, because, again, I was doing, like, SEO where it's like, "Yeah, we're gonna do this and we think it's gonna work." (laughs) But, uh, we won't know for about, you know, six months. Someone showed me a paid platform and I was like, "Wait. You know this is the keyword you're betting on. This is how many times the ad was shown, how many people clicked on it. How many... Wow, this is amazing. I need to, I need to get more into this." And that's just kind of where it started. And, uh, when I usually hire people and we go through the interview process, I make sure that they can be data-focused. Really, it's just ensuring that we're focusing on that signal and not the noise. And one of the interview questions I usually ask is, like, I, I throw a bunch of, you know, data points at them and then say, "O- out of all that, uh, w- what would you do to optimize the account?" And more times than not, that they're just like overwhelmed. Uh, but the people that I like are like, "Well, w- what was the purpose of the cam... What's the purpose of the campaign?"

    11. LR

      Mm-hmm.

    12. TD

      Uh, "The purpose of the campaign is to drive conversions." "Okay, well, I would focus here, like, focus on how many clicks we're getting and how many conversions. Are we targeting the right people?" Because who cares about how many impressions you're getting? Who cares about the reach? Who cares about the frequency? If the goal of the campaign is to get the conversions, that's what we should be doing. If it was awareness, it's how many people saw it. Uh, if it was consideration, how many people are we getting into the funnel and are considering us as from a white paper download or a demo, as the product solution that

  13. 47:33 – 54:43

    Metrics that matter

    1. TD

      they're looking for?

    2. LR

      That's a great segue to the next question I wanted to ask you, which is just metrics that you love to focus on, pay attention to when you're helping people with paid ads. Like, a lot of people think about CAC. A lot of people think about return on ad spend, LTV to CAC. I know it really depends on the goals, as you said, and, but I guess is there anything that you find is like, "Just forget these metrics," these are kind of the bucket that you focus on mo- most?

    3. TD

      Yeah. So, so for those metrics, I would say, I really hope you have a, a great rev... uh, finance partner. Shout out to Courtney and Nick, uh, the, the fantastic finance partners I have at Shopify. Well, Nick's not there anymore. Sad cry emoji. But, uh, Courtney, you know, we're, we're able to collaborate on those things, understand this is the investment we're putting in, this is the expected return, this is the CAC that we can, uh, basically put a guardrail on, you have to be within. So, hope- hopefully you have a really strong financial partner you can use for those. But as far as the stuff that I focus on, there are things that whenever we're, we're looking at accounts, uh, we'll, we'll kind of hyper-focus on. For example, Google will give you information about what is going on with your ad copy and how to ultimately show as often as possible. Uh, what I'm showing here is a visual of brand versus non-brand when it comes to expected click-through rate, landing page experience, and ad relevance. The reason I like...... building reports like this is because it will kind of show you where your hole in the ship is. If you look on the left side, expected click-through rate, 75% above average. Landing page experience, 84% above average. Then you get down to ad relevance. (laughs) Lot more, uh, colors than there are green. Only 35% above average. So for your brand ad copy, the clear direction you have from the data that's available to you, is jump into the account and make it more relevant. Uh, that will improve your ad strength, and once you improve your ad strength, uh, that will impro- prove your quality score and more times than... I think it's 12 per- a 12% increase in the number of impressions you can get if you increase from, uh, a below average to an above average for the data that's available here. And then on the, the non-brand side, clear expected click-through rate, uh, you really need to be working on. And that data is also available here, that you can kind of see. I put in red, (laughs) uh, the, the one for brand, your quality score being at a nine, but your average CPC is $8.00. That is very interesting, 'cause it should be the lower your CPC is... I mean, the higher your CPC is, the lower your CPC should be. So that is an investigation that it's like, "Hey, here's a signal, let's go do a deep dive." So there's further reports, uh, that you could look at. Uh, looking at the ad strength again, average, excellent, good or poor, you can see clearly on the, uh, non-brand side, doing really well. They're (laughs) doing really well here. Most of their ads, 46 of their ads are excellent, and a bulk of their spend is going there. Um, but if we look on the brand side, you have one ad, one ad that you could go in and just click pause on. That is a poor. And look at that CPC, $7.00. So, just turning that ad off by itself, not only will pro- I'm a firm believer in that there's a, an account level qu- uh, qualities score, not just necessarily a keyword or an ad qualities score. So just turning this one off and removing it from the account, this one ad could dramatically improve your performance. And then Google actually tells you what you need to do to increase those strength things. So this report is ultimately powered by this one. So for your ad strength improvements, uh, they'll tell you, "Try adding a few more unique headlines, or unpinning some of the ad sets. Try including more keywords in your descriptions or your headlines." So they're actually giving you this information, but more times than not, I just feel like people are just missing this and just not taking action on it. Um, so definitely some stuff available, uh, within the account that will help you focus on that signal versus the noise that you could be just getting from the clicks, the impressions, and all those things that are available to you.

    4. LR

      That was amazing. For folks not watching on YouTube, you pulled up, uh, actual reports from, I think, Imagen past clients. Is that what the, uh, the data was from?

    5. TD

      Yep.

    6. LR

      Okay. So actual reports from past clients of how you evaluate... And this is Google Ads, basically. Uh, Google Ads data.

    7. TD

      Correct.

    8. LR

      Incredible. Okay. And so within the, those reports, you had poor, excellent. How, how do you determine when something is going great? Is that something you set? Like if it's above this threshold, is that a benchmark you have? Or is that Google telling you this is poor, this-

    9. TD

      Yeah. So that's Google telling you.

    10. LR

      Okay.

    11. TD

      Um, there are some times where you will work really, really hard (laughs) uh, to take that, uh, poor to an average, that average to, uh, an excellent, and there, there's just not a whole lot you can do, uh, sometimes. It, it kind of is what it is, but I would argue more times than not, people are not doing those things to improve their ad strength. Uh, so ju- just if you have the change history available in every account, in Meta, in Google, if people are doing the tests to try and improve and get better, great, you have the right people working. But if not, that, that's a clear indication that there's definitely room for improvement.

    12. LR

      Something that someone asked on Twitter is, is for benchmarks around any of this stuff, like how do you know when your CPC is good? I guess you just look at... Do you look at like Google telling you it's excellent or it's poor? How do you know if your conversion rate is good? How do you know if your CAC is good?

    13. TD

      Yeah. Every industry and every company is gonna be different. Healthcare, lawyers, those are some of the highest cost and average CPCs I've seen. If I was to compare that industry, say, to a B2B industry, it would not be very good at all. Honestly, what I would tell the user to do, every single platform has partners available. More times than not, they're kind of like salespeople. I actually really like the partners, uh, that we have at Shopify, uh, because I do feel that they are partners. Shout out to Fr- Francisco at Google, and, uh, Sammy and Ilana at, uh, Meta, and Nick and Sam and Bryan at LinkedIn, in case they all hear this. But, you know, I think they're really great partners, and

  14. 54:43 – 56:15

    Competitor analysis and CPC insights

    1. TD

      they will give you that information. They will actually say, "If you give me your... the five people you consider to be your top competitors, I will tell you if you are above or below a certain threshold." Now, they can't tell you CAC 'cause they would have to have transparency into, like, conversions. Uh, but click-through rate, conversion rate, cost per click, things like that, they will give you that information 'cause they, they, they will anonymize it so you won't know who is who. Like if you say, "My competitor is monday.com," they won't say, "Monday.com has a click-through rate of 5%."They, they, they won't tell you that. But they will tell you, "Of the competitors you gave us..." And we added three more in just to, you know, say, "Hey, we found some people." And also, to where if you only give them two, you could be like, "Oh, it's one or the other." They- they'll give you that information, 'cause they want you to succeed. Because if you succeed, you spend more money on their platform. Uh, so if you find out your average CPC is really high, take the actions that are in the account that tell you what you can do to lower it. But if you find out that you have a lower CPC, you're like, "Oh, great, I thought that was worse." And then, you can communicate that internally to your team and say, "Hey, our average CPC may seem high, but in reality, we got with our partners and we are, uh, you know, on par with the industry."

    2. LR

      Got it. So the advice is, don't seek generic benchmarks for any of these metrics. You can talk to your rep at Google, Facebook, LinkedIn, et- et cetera, and they will give you, essentially, the numbers -

    3. TD

      Great summary, yes. Yes.

    4. LR

      Amazing.

  15. 56:15 – 56:38

    Custom reports and data visualization

    1. LR

      Okay, back, going back to the report you just showed, is that, is that a report you developed custom that gives you like, "Here's the most important stuff to pay attention to," or is it basically an export from Google Ad Manager and...

    2. TD

      So these are all e- exports.

    3. LR

      Okay.

    4. TD

      So th- this is just the visualization of the data that's available. So this, all data is available in every account. This is data available in the account, this is

  16. 56:38 – 59:10

    Understanding impression and click share

    1. TD

      data available in the account. Now, in saying that, there are more reports that we get into, like impression share and frequency. This is something I tend to look at that not a lot of other people do. It's not about showing as often as possible, because that's what impression share tells you. And that's kind of like ego marketing, "I want to be number one, I want to be there all the time." It's about showing to the right person as often as possible. Instead of thinking about being on top of the page, you should think about serving the right user, which can be measured with our click share. And then Google will put us on top of the page, which we can see through our top impression share. That's what this whole visualization is here for, where I feel like a lot of people don't do this type of visualization. And this can be done by leveraging all the user data available in Google, such as, you know, demographics, locations, affinity audiences, in-market audiences, testing different bid strategies and more. So, the data's readily available, but it's just making the visualization a little bit more digestible. And the reason that I actually have two here to show is because this is an example of when I was working on an account back in 2023, uh, where we started making some changes. And you can see that clearly our impression share, you know, I would say was baseline. You know, it didn't really change too much. But look at our click share going up. Because now, again, we're talking more to the right people, not necessarily just everyone, generic people. Versus this campaign that was, this was the test and this was the control. And you can clearly see that we were talking to a lot of people, but no one cared about what we were talking about. So this was an easy test to call and just say, "This one's performing a lot better." And then, you know, verifying the other metrics still look good, conversion rate, number of conversions, stuff like that. This one was the clear winner, so we went in this direction.

    2. LR

      This is super cool. I love that you're showing this. And again, plug for YouTube to actually see the charts (laughs) you're talking about. I see one more slide on this deck. I'm curious what's there.

    3. TD

      Yeah, so this one, I'm hoping you have show notes, uh, that we can, uh, give a shout-out to somebody. Um, so...

    4. LR

      Yeah, absolutely.

    5. TD

      Yep, uh, there's, uh, a website called PPC Hero. Um, (laughs) what I usually tell people is, "Terrible name, great content." The actual website used to be a little bit more cartoonish with like superheroes on it, but now it's a l- you know, they've

  17. 59:10 – 1:02:14

    True competition metrics

    1. TD

      refined it. But, uh, the writer there, his name's Jacob Brown, this is what he calls the true competition metric. And again, this is why I highly recommend we give people the, uh, link to the article, 'cause this does get a little dense. So, what he does is that he creates four new metrics with the auction insights that are available. And two of those metrics are position above rate when we show and amount of times they show and we don't, added together, and it will show how often a competitor ranks above you in all auctions you are available for. Using this type of lens is an effective way to identify genuine threats. Not that just this data can be used to gain comprehensive insights, like determining your position and impression share. But by creating a baseline that you can see how this data changes over time through different bid strategies, keyword ad copy optimizations and more. This is a very, very powerful one, 'cause I do know a lot of people are always concerned about, "What are my competitors doing? How are they doing it? Why are they doing it?" And with all the smart bidding currently available, you know, back in, back in my day, you know, we were... It was all manual bid. So you could change the bid ever so slightly and then be like, "Oh, now my competitor's showing above me." And this is even before Google gave us auction insights data. Uh, "This is now available to us in the platform." Tell... Using this report will actually identify a true threat, as opposed to making an assumption based on personalization that maybe you're seeing when you Google a keyword, or using a tool like Semrush that ultimately is, uh, saying, "Oh, this, this person's appearing above you." Well, let's run this report and see if it's a genuine threat, as opposed to just maybe ego marketing going on.

    2. LR

      Super cool. And this comes from PPC Hero, is, is what you're saying?

    3. TD

      Yep.

    4. LR

      Cool. We'll definitely link to that. Are there any other tools or workflows that you find really helpful in analyzing this endless amount of data?

    5. TD

      Please, audience, if you, if, if you are using AI in this way, let me know. Uh, because a lot of the stuff I do is can be a little manual, and can be a little time-consuming. If they've found ways to automate through AI, uh, that does ... That ... Where I can provide the information that isn't, you know, sensitive, uh, that they will, you know, make it a lot cleaner and a lot more digestible, let me know. But a lot of the things, I have templates that you can load the data into and it will just automatically give you, uh, the result from it. But yeah. It's, uh ... Right, right now it is a little manual, but if there's a way to automate this through AI, please users, uh, contact me on LinkedIn and let me know.

    6. LR

      Sounds like a startup opportunity right there.

    7. TD

      Yeah.

    8. LR

      Sweet. Okay. Let's move in a

  18. 1:02:14 – 1:08:52

    Attribution and incrementality

    1. LR

      slightly different direction. Attribution. What a, what a sexy, uh, exciting topic. So attribution, basically it's how do you assign credit to a channel and to a campaign, uh, so that you know where growth is coming from, what's working. What's the state of attribution today? What do you find is, is helpful? How do people do attribution well in today's world?

    2. TD

      Yeah. I'm a big believer and proponent in multi-touch. I fall into the camp of more time decay. Historically, the research I've done, users tend to forget very quickly where they first found your brand. Yes, you should get credit because this is the first time they saw your brand or the first time they interacted with it. Uh, for sure. Let's give you some credit for that. But, uh, definitely not the reason that they ultimately converted. Um, linear is fine as well, but, uh, you know, overall, I do think, you know, attribution by itself is biased. It does not answer the question of whether the person clicking on or seeing an ad would have converted anyways, even in the abs- absence of that ad. This is why numerous companies like Netflix, uh, and eBay have done studies to get an understanding of the incrementality of paid advertising campaigns, whether that's conducting GeoX or a conversion lift test and important slices of, uh, the marketing channel. I know eBay was very popular when they did that, uh, experiment. I think it was like 2012. Uh, and I'll have a link for you for that as well, uh, where they ultimately decided to cut almost all of their brand spend because they found that users were already going to find them through organic anyway, so they were just kind of throwing money out the window, and it was really hard for competitors to get their ads at the top of the page because their quality scores were so low.

    3. LR

      Mm-hmm. Okay. So multi-touch is the way you like to think about it. Basically give credits to all of the channels that you detected the user saw your ad on, and less credit if it was further back in time.

    4. TD

      Right. Yep.

    5. LR

      And then in terms of tooling, is there tools you love? Is, is it stuff you build in-house? Generally, how do, how do people go about doing this sort of stuff?

    6. TD

      Yeah. I don't know if I want to say I have been fortunate or not. Uh, but most of the companies I've worked with, like the, the, the big companies of the world, the Pinterest, the Shopifys, the IBMs of the world tend to build things in-house. So from a, a third party tool, unfortunately I don't have a whole lot of, uh, go-tos for that. So again, either that's a, I've been fortunate or I've been sheltered.

    7. LR

      And then are, are there tools that you see people use? Or is it just like really nothing amazing out there?

    8. TD

      Yeah. No- no- none that I've seen that's like, "Oh wow, that's amazing." Uh, I will always say make sure you're trying to leverage the in-platform tools, uh, because the biggest thing with performance marketing is the signal you're sending the platform, because if you are telling the platform, "Oh, I want to get more of the users that are doing this action," it's going to do a really good job of giving you that. But if that action doesn't equal business results, you're, you're, you're not helping yourself at the end of the day. You're, you're giving it a- again, a lot of noise instead of the right signal. So, uh, try and leverage the in-platform tools, or if there is a tool you're using, make sure it does, you know, uh, uh, allow third party integration through the Google, Meta, TikToks of the world.

    9. LR

      Then let's talk about incrementality. You mentioned this idea of how do you know if the money you spent led to incremental growth that wouldn't have happened if you didn't run, run the ad. Is there any advice there? Anything you've seen about just like how to think about incrementality correctly and not just give yourself all this credit for stuff that would have happened anyway?

    10. TD

      Yeah. I mean, there are many ways to judge the effectiveness of, uh, growth overall. Like some of the stuff we talked about before, brand metrics, awareness, recall, things along those lines. There are also, I know some companies look at like leading indicators like visits and clicks or attribute which of the efforts are linked to or perceived, uh, drivers of actions, like leads, conversions, to prospect, prospect to scale. But, you know, really the, the, the results should be coming from either that GeoX or geo experiment or conversion lift. Uh, the results of those experiments should ultimately fuel the plans for your, uh, what we call IAF, incrementality adjusted factor, and will allow you to be more precise in how efficient each channel is, ideally by region. Sometimes you just have a holistic of like this is how Meta is, this is how YouTube is. But if you on- if you don't have it by region, don't fret. It's fine. Just have it by platform. And kind of like what does that look in practice? Uh, when you're running a conversion lift test, you intentionally do not show your ad to some users when you win an auction, and instead Google shows the next bidder's ad.This is your control group and adds up to some opportunity cost that is estimated in terms of impression share per- uh, percentage, as well as spend holdback. Spend holdback is like how much money you will not spend because of this test. And all of the platforms are, uh, willing to partner with you on this because, you know, Facebook knows this, LinkedIn knows this. All of them know that they are very visually based creative assets that are not getting as much credit as they deserve. So if you go to any of them, you if- say you don't have a dedicated rep, you can call. And if you ask for this, more times than not, they are willing to partner with you. Uh, in saying that, if you're not spending enough, they probably will not, uh, help you with this 'cause there are certain spend thresholds you do need to meet. But, uh, I would say at least start there. Uh, and also if you're not spending more than I would say 50K a month in the platform, doing this is going to be a lot more work, uh, than you're gonna get result from. Y- you're just not gonna have enough signal there.

    11. LR

      So basically don't, don't worry about running incrementality tests when you're just, when you're-

    12. TD

      Yeah, when you're starting out.

    13. LR

      ... not spending a ton.

    14. TD

      Yep.

    15. LR

      Got it. And okay. And so basically to understand actual incrementality, every platform has a way for you to actually test it on the platform, and the team there can help you run it.

    16. TD

      Yep.

  19. 1:08:52 – 1:10:53

    Building a performance marketing team

    1. TD

    2. LR

      Awesome. Okay. Let's go back to talking about team structure and how to kind of build your own performance marketing team. So we talked about the first person that you hire and the advice there was someone that's very... Understands how to find signal in noise. So there's that one person, and you... Your advice there was maybe around like 50K. Was that like an actual threshold they usually recommend or is that just like a run example?

    3. TD

      Uh, every business is slightly different.

    4. LR

      Mm-hmm.

    5. TD

      I mean, if they're, uh, well funded, 50K may not... Uh, the, the threshold may be higher. But, uh, yeah, everyone's different. 50K may be... Like if somebody said... If you were to just say, "Hey, give me a number," I would say, "50K to, to start having those conversations." Because if you're at 50K, say for the month of June, great. It's gonna take us three months to hire someone anyway, so at least start the conversations now.

    6. LR

      Awesome. Okay. What are like the first three to five hires look like generally that you recommend for scaling internally performance marketing?

    7. TD

      So the first thing, uh, like we said, someone data-driven that can get into the platforms. The next is, uh, going to be creative, uh, because I need those two now working hand in hand, um, making sure the creative is matching the tone and also the performance that we're trying to achieve as a business. Uh, and then third would be a dedicated data scientist.

    8. LR

      Mm-hmm.

    9. TD

      A fully dedicated person, 'cause they can help you with things like the incrementality testing. They can help create reports that will ultimately make everyone's lives better. They will be able to build analyses that you will not... As a generalist will not be able to do yourself. You know, th- there's a saying of like, uh, I'm not a data scientist but I like to play one online, because th- what they do, they ultimately make us look really good, because we're ultimately the ones reporting on the performance of it. But they were the ones that helped build that environment and build all those things for us

  20. 1:10:53 – 1:15:48

    Creative and ad copy collaboration

    1. TD

      to succeed.

    2. LR

      And in terms of the creative person, is that like a graphic designer? Is it like a marketing person? What's like the actual skill set there?

    3. TD

      It would be more, uh, graphic design/branding.

    4. LR

      Mm-hmm.

    5. TD

      Uh, the reason for that is because they're... You're... If you have a good marketing mix, you're gonna have... If we keep it to a three-step funnel of awareness, consideration, purchase, you're going to need to build some brand equity in a specific direction. You're going to need to, uh, make sure you're communicating value, which now you're not being as creative. You are being more directional. And then ultimately the purchase is like, "Hey, click here. Convert now." So you do want to give them the ability to still be creative. "Hey, I hired you because you have a good creative eye and you're good at what you do." But now we need to focus on getting that person to convert. So you give them a little free rein to be creative, but then you also need them to be able to execute against that creative. You need to get those users to convert.

    6. LR

      And they're also writing the copy, I imagine, for like the Google Ads or...

    7. TD

      I usually say that should be collaborative.

    8. LR

      Mm-hmm.

    9. TD

      Um, the performance marketer should be able to write most of the ads, but I can't tell you how m- how many times in my career where I've written an ad and I'm like, "This is the greatest ad ever written known to man." People will write stories about this ad. It is amazing. And it flops.

    10. LR

      Mm-hmm.

    11. TD

      Because I'm not the target audience more times than not. Uh, so I think it should be collaborative and no idea should be left on the cutting room floor 'cause... Perfect example. I was working with ADT, you know, the security company. We wrote the most perfect ad when Google Ads only allowed a headline and two descriptions of 35 and 35. We got every single value prop in there somehow. It was amazing. And the ad that it was going against was dollar sign, zero, setup fee, dollar sign, zero, install fee. That ad won. It's like that is... No. How did that... (laughs) H- it, it barely uses any of the characters and it almost tells... I mean, tells you almost nothing, but it, it won. Um, had more conversions, a higher click-through rate. So we took that and we applied that with the value props and it did better. So it should never be like... Unless the idea is don't buy our product, which hopefully someone is not writing that ad copy, it shouldn't be left on the cutting room floor. Always test it. Always be willing to learn what, what works, what doesn't.

    12. LR

      For this first hire, what's the, what's like the title of this person usually in your experience?

    13. TD

      Lately it's been, uh, growth marketing specialist, growth marketing manager, uh, because they're going to wear multiple hats. You know, like at any startup that you're at, you, you're, you're gonna be asked one day to, "Hey, I want you to do performance ads," and then tomorrow it's like, "Hey, I need you to help me build out this spreadsheet for a spec sheet," that you, you have no idea what you're doing. So you're always gonna wear multiple hats, so just having a general title like that to start out with. And then if that per- person matures into a role, you can make them more of a specialist, or if they start showing signs of like, "Hey, I really like doing the social stuff," and we've scaled enough, "Okay, let me hire a paid search person." So yeah. I always start with a general, uh, and then as the team grows, we get more into specialties.

    14. LR

      So like growth marketing person, it's kind of like the-

    15. TD

      Yeah.

    16. LR

      ... broad umbrella.

    17. TD

      Yeah.

    18. LR

      And then are these people sitting in Google Ad Manager and Meta Ads and just like running ads manually?

    19. TD

      I'm a firm believer, at Shopify we call it GSD, getting shit done. I'm a firm believer in getting shit done. Uh, you should be in the account, like I mentioned earlier with that ops cadence. We have stuff we need to be doing weekly, biweekly, monthly. The bigger the company gets, the tend to m- you wind up in more and more meetings talking about the things you want to do and how you're going to do it and stuff like that. Uh, but keeping those people kind of sheltered away from that and focused on those things are gonna drive the best results for you you possibly can get. And that means hands on keyboards, in the Ads Manager tweaking things, setting up a calendar. I'm a firm believer in setting up a calendar. We did this test, we started this test on this day. That means this test will end on, a month from now. Put a notification so you have a cool down period and you report out to the org what you did, how you did it, why you did it, and then the results from it. And then, all right, what we learned from this, uh, is this, and we will be applying that to our next test and this is how. So yeah. Hands on keyboards, doing all of those things. So again, hiring those smart people and just getting out of their way.

    20. LR

      I love

  21. 1:15:48 – 1:20:52

    Managing workload and hiring strategy

    1. LR

      that. In terms of how this team grows, you mentioned s- when we were chatting that you wait for someone to cry uncle, to hire more and to add to the team to kind of avoid bloat. Talk about that.

    2. TD

      You know, unfortunately we've seen a lot in the news lately with a lot of tech companies, um, letting go of some really talented people. And that is, I feel like just created bloated organizations. We, every, uh, month, uh, my current manager Dean created this, uh, calculator that we look at that says, "How much time are you spending in meetings?" Uh, if you have any PTO coming up, put that in there. Optimizations, reporting, stuff like that to basically add up to how many days are in the quarter. Uh, 'cause you know, every quarter, well not every quarter, but most quarters you'll have like vacation or you'll have say what we call a Shopify burst where it's, uh, we meet in real, in real life to, you know, get shit done in real life as opposed to remotely. You know, put all that in there and then what does the number equate to? Oh, we're in the red right now for these, you know, two to five people. How many quarters has it been that way? Okay, it's only this quarter. This quarter we have a summit coming up or we have a burst coming up, or we have a lot of travel because, uh, we're meeting with partners, so on and so forth. So maybe this is an isolated thing. Let's go ahead and wait till next quarter. All right, next quarter it's red again. All right, now maybe we need to start having the conversation of what this new hire will take over, uh, what they will be responsible for and, uh, how much work they'll be taking on and doing to replace, uh, some of this red, uh, that is going on. And if it equates to a full head, great, we can move forward. We've made our business case. Uh, but sometimes it doesn't. Sometimes we're just red and, uh, we need to do a better job of making sure, "Hey, we need to step out of these meetings. These meetings are sucking out our time and we don't need to be a part of it." Or, uh, "This launch we don't need to be a part of, we just need to be consulted on it. We don't need to be in every single meeting every single time or every single communication." So just making sure we're looking at the right things before we decide to hire someone and making sure that we have stuff for them to do.

    3. LR

      And, uh, red means, uh, they have more work.

    4. TD

      Yeah. More days than there are in the mo- in the quarter.

    5. LR

      And so they basically estimate here's how much I, how many days I need to do the things I've committed to for the quarter. And then it's like-

    6. TD

      Yep.

    7. LR

      ... "How many actual days do you have this quarter?"

    8. TD

      Yep.

    9. LR

      That is super cool. And so step one is okay, if you're in the red, let's cut some stuff and then if they're still in the red and you've cut stuff, then okay, we need to start hiring.

    10. TD

      Yep.

    11. LR

      That is very cool. Is that just a, is that a Shopify thing or is that something you d- you do on your team?

    12. TD

      I've done stuff like that at other companies before, but uh, kind of bringing it forward again, I don't want to take the credit for it. Dean was kind of the one that brought it back up and was like, "Oh yeah, I used to do this. I don't know why we, I stopped doing it." So it's definitely something I've used at other companies for sure.

    13. LR

      That is super cool. You mentioned this ops cadence, is that something you can describe just like what this cadence looks like of how you run?

    14. TD

      Yeah. So I lo- I love me a spreadsheet, so it's just a spreadsheet and, uh, visually, uh, I'll do my best to describe it. Let's say column A has those buckets I was talking about, like a finance, performance, structure, keywords, so on and so forth. And then within those buckets, or let's call those everyone loves rocks and pebbles right now, right? So that, that's your big rock. Your big rock is keywords. Then within that you have pebbles, uh, keyword granularity, brand versus non-brand, search query reports, negatives, so on and so forth. And then within that we say how often we're doing it, are we doing it weekly, biweekly, monthly? And then that allows us to, if anyone in the organization's like, "Hey, how often are you guys updating ad copy?"... easy answer. How often are you guys, uh, doing search query reports? Easy answer. And it allows us to make sure we hold ourselves accountable to those things 'cause a lot of times we, we have a lot we're doing. We're working in Google, we're working in Meta, we're working in YouTube. You could easily forget, "Oh, I didn't do that. Ah, I, I got to make sure I do that again." So it hold you, it, it's a way to hold yourself accountable, but it's also a way for me as a manager to go in and kind of spot check that and make sure that they're doing the things that they, that need to be done in the account.

    15. LR

      The core of this, essentially, there's a spreadsheet that everyone aligns on of here's how, when and how often we do certain activity to operate this performance marketing machine that you've built.

    16. TD

      Yep.

    17. LR

      Awesome. And it's, it's both internally so that everyone knows and then also when people ask, "Hey, when are you gonna do this?" Say, "Okay, here, here's the date."

Episode duration: 1:42:03

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode zNJyb3R_Pnc

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome