Using Data Science to Profitably Scale Cross Channel Marketing Campaigns (E135)

  • Jake Cook
  • CEO of Tadpull

Show Notes:

  • RFM
    • Recency
    • Frequency
    • Monetary
  • Master Key
    • Align data
  • Social Media Slot Machine
  • Google Analytics
    • Speed to buy
    • If people buy in the first day, run heavy retargeting
  • Scaling
    • Does not have good Roles
    • Bad cash flow cycles
    • Not knowing margins
    • Quarterly Goals
    • Have a process that’s repeatable



Jake is a forever curious entrepreneur that works across digital marketing, analytics, and AI.

Jake started out as a lost physics major that stumbled into online marketing and eventually design thinking and data science. Helping clients and students use empathy to unlock growth at the intersection of art and science are where he loves to spend his days.


Transcript :

Charles (00:00):

In this episode of the business. E-Commerce I talk with Jay cook about using data science to profitably scale cross channel e-commerce marketing campaigns. This is a business of eCommerce episode one 35.

Charles (00:21):

Welcome to the business. E-Commerce the show that helps eCommerce retailers start launch and grow their eCommerce business. I’m the host Chelsea [inaudible] and I’m here today with Jake cook. Jake is a founder of tadpole company. He started in 2013 where they use data to help e-commerce brands, profitably scale, their marketing campaigns, as joke in the show today. Talk about how you can use data science to scale your marketing campaigns. So, Hey Jake, how are you doing today? I’m wonderful. Thanks for having me on. Yeah, I am awesome to dig into this topic a little bit. We’re talking before the show about, but a few proposed topics and the Xi learning and data science. I feel like this is a new we’re in 2020. There’s a lot of new things going on, but this is kind of the new hotness right on using with we’ve now pulled all this data.

Charles (01:08):

Right. And I feel like everyone’s kind of getting a sense of yeah, you can like get all this stuff and now it’s like, what do we do with this? So, yes. Yeah. So what do you guys, so first you, so your product, right? Usually don’t go into the product the beginning, but it’s just kind of interesting right on you guys basically hook into a lot of different pieces of data and start predicting what customers basically gonna want before they know they want it. It’s that kind of a, yeah, it’s it’s we work a lot with mid market companies. And so, you know, there’s lots of data and we always say, it’s kind of like the kitchen drawer or, you know, the junk drawer you have in your house where you open it up and there’s like birthday candles and bandaids and all sorts of junk.

Charles (01:51):

And that’s kinda how their data is structured. And so a lot of what we do is kind of suck into a, we call it a data pond. I mean, there’s like Lily pads of data, if you will. So you have Google analytics, data or Shopify data inventory date, and we’re sucking all into one place. And when you can kind of get all organized and neatly put together, you start to find some interesting insights. So for example, you might find that customers out of email by, at 30% higher than people that come through Facebook or products that you know, don’t have great margins, but they turn like really quick turn out that their total cash cows. And so, you know, well, how do you find the customers that buy those products that turn quickly? Well, now we’re trying to map the inventory to the customer data.

Charles (02:32):

So a lot of that has kind of just drawn those correlations and help them customers see what’s possible off these assets. They already kind of already own. Yeah. I feel like a lot of this, even I was looking at something that I did with, ah, just like ad attribution and there’s still, it’s really tough, right? Because there’s the same, their social campaigns. Are you doing PPC and then you’re remarketing. And if you kind of just look just basic, if you have Facebook and Google and Google ad words going, they both, they both saying you got a conversion here and then you realize that’s the same conversion. You both like, you both can’t have that conversion. So like, if you do it that way, like I double kind of given you have to add us, but like, how do you know that? Are you doing it on the first click on the last click?

Charles (03:17):

So this is a lot of stuff here. And if you don’t have one platform that kind of brings all together in your head, you would think I got two conversions, but you really got one until you like tie that in. It’s very hard to suss out it is. And I think you hit the nail on the head, you know, you’re using first click or last click, how long are your attribution windows? Are you looking back 30 days? You know, and what’s the path length or how many ways the users come in and, and go through that. And so part of the, the answer to that and, and attribution is something that’s always very, very difficult if not impossible to solve, to solve. So just give up, all right.

Charles (04:00):

It’s kind of like, you know, like trying to like stay in really good shape. It’s a lot easier in your twenties and then it is in your forties or fifties. So like, you should still try, but you know, you may not get, you know, no you’re not ever going to get a hundred percent. Exactly. I think, you know, but versus just throwing your hands up and say, okay, whatever Facebook says, we converted that we’re going to buy, you know, the source of truth, all these things pipe back through Google analytics for almost 85% of websites. And unless you’re like enterprise and you can afford a really top tier analytics package, site analytics package, you know, Google analytics is just fine. You know, you can do a lot with that. Yup. Yup. And you hit the nail on there. Their Facebook is going to tell you something very different.

Charles (04:35):

You know, if you ever compare Google analytics to Facebook, they’ll always be discrepancies bylined by a loss. Like not exactly by orders of magnitude, as you think it would be like 2%. And you’re like, no, it’s off by like triple like it’s. Yeah. It’s very strange. It’s very, very strange. And I think so a lot of what we’ve been trying to work on is giving our clients first party data. So we have our own pixel. So all that data that Amazon or Google or Facebook has our mission is to kind of give that to mid market or smaller the smaller guys. And so, you know, the users as they came through the site, what they did, if they, you know, obviously be compliant with privacy laws in the process, but with that, you start to have this asset, you build it like you really understand your customers and you’re not seeing it through the filtered view of Facebook or, or Google.

Charles (05:22):

And you can, you can do some really amazing things. It takes time to build that data set though, to do the AI on it. And that’s something that you know, can take you two years to build the data set and it takes you two weeks to have the algorithm run against it. So today’s episode is sponsored by drip, drip. It’s a world’s first e-commerce CRM and a tool that I personally use for email marketing and automation. Now, if you’re running an eCommerce store, you need to have drip to try. And here’s why trip offers one-click integrations, Shopify and Magento. There’s robust segmentation, personalization, and revenue dashboards. To give you an overview of how your automation emails are performing. One of my favorite features of drip is the visual workflow builder. It gives you a super easy way to build out your automation world visually and see the entire process.

Charles (06:06):

It lets you get started quickly, but also build very complex automation rules. It’s powerful, but also easy to learn. Unlike a lot of email tools and offer the same type of automation to get a demo of drip today, you can go head over to That’s O E now onto the show. Yeah. So if somebody is kind of listening at home and they want to do something like this, right, they’re running their ad words, they’re running their Facebook, they have their analytics. Maybe even some, they have their Shopify big commerce data. Where do you even get started trying to like put these all together? So one of the easiest things you can do to get started just to kind of like

Jake (06:49):

Get a sense of your customer base is mostly we’ll have acquisition costs. You know, customer acquisition costs, CAC, you know, there’s ratios of your customer acquisition cost versus a lifetime value. So one of the ways to get at lifetime value is a really simple, simple way is just get your customer records and your orders and just, we have some very basic Excel jockeying. You can see which customers buy a lot. And then you can also look at return. So your customers that buy a lot and you know, you see a free shipping, they buy all the time, but they return all the time. You can start to say, okay, we’re going to filter them out and you can start to build these little segments manually. And in the analysis, you’ll come across all sorts of really interesting insights. So you can start to slice the data by geography or zip code, for example, or even state cities, things like that.

Jake (07:31):

But it really starts to what they call recency frequency and monetary RFM is what they call it in customer analytics, analytics. But you’re looking at that as kind of a way to think about what your most customers look like. So how recently they purchased, how frequently they purchase and then what, how much they actually purchased the monetary piece as well. And that can all be done with just, you know, exporting a customer list out of Shopify and doing some very basic you know, Excel jockey and, and some pivot tables and things like that.

Charles (08:01):

Yeah. I feel like being good Excel is like a super power and this many industries, like everyone’s willing to talk to someone they’re like, yeah, I don’t know, Excel that you might want to take a course, like that’s theirs.

Jake (08:13):

And if you’re really hardcore you know, a lot of this stuff can be built with Python and in some, some stuff with these notebooks, a lot of stuff, you know, if you can just do some very basic coding, you can kind of Huck it in there and get them and, you know, get the data clean and get it into a notebook. They call it and start to see what it does.

Charles (08:30):

So yeah, that’s, that’s, that’s where gets really fun. Okay. So, but you would say at least try to get as much data as you can out of these, like figuring out how to suck it out of Facebook, suck it out of Google. Like just get it out of their platform basically. Yeah. I think getting it back and the nice thing is

Jake (08:47):

You kind of need a way to like, they call like a master key, if you will, or way to kind of like align everything and an email address. If you can get an email address, you can kind of match that to your customer record, maybe match that to a lead gen ad in Facebook, for example, and, and kind of pivot that again, like this is, it’s not trivial to do this stuff, so it can be kind of complicated, but if you grind on it, you’ll be real all of a sudden, and it takes time, but you’ll start to see your acquisition costs fall. Cause you’re really honing in on the right target audience. Are those, those lookalike audiences, for example, can be really well done. And you run like a more profitable business at the end of the day, cause you’re not spending that money on customers that

Charles (09:22):

They’re going to buy in return. So if you have to

Jake (09:26):

Capital, it can be a really powerful asset. If you really understand that kind of mechanisms and the lifetime value off actual your own individual data, you can go and say like, Hey, we have an L a lifetime value of three years and it’s about, you know, this type of revenue. And so it costs to acquire these customers, investors really like to hear that. And even if you’re wrong, maybe there’s slop in the system, but it, the fact that you even think in that way, they really, really like to see that kind of sophistication we’ve seen

Charles (09:52):

Well, and just having that data kind of gives you the, the guts, right. To say, I’m willing to like 10 X down on this. Like I’m willing to really take a big bet on X because I believe in it, like we have the data, we did the analysis because without that, you’re basically just gambling at that point. Right. It was just trying to like, you’re trying random things unless you really know

Jake (10:11):

Exactly. And it’s kind of that we call it the, you know, the social media slot machine, we’re just kind of throw it on the slot machine, hoping you get three cherries and you cash out. And you know, there’s also that I think Andrew Chan, he was growth at Uber. He’s an Andreessen Horowitz. They had this thing cause it kind of a lot of bad click throughs. Basically he uses a little more profanity,

Charles (10:30):


Jake (10:31):

Basically over time, if you identify an acquisition channel over time, you know, economics that tells us that, you know, people will flock to it. And so if you we’ve seen a lot of businesses that will get built in two or three years on like, say Instagram stories or some sort,

Charles (10:46):

But you know, Snapchat or some sort of hack like that. But over time it eventually kind of starts to AqueSys and costs, rise, and the lifetime value isn’t there. Then it can really kind of put you in a bad spin if you can’t exit the business. Yeah. I was talking to the last show on the very last, at one 34. And I’m old enough to remember where an ad words ad words was cheap. So it’s like a great deal, like 10 cents per click. And now it’s like, you can’t even, you couldn’t, you couldn’t get 10 words, a 10 word term for 10 cents at this point. So it’s exactly different. But like when it first came up, when these channels were new, that was the thing. And like, if you look at new emerging channels right now, like you said, like when Instagram stories was a new thing that was possible. So it’s just, it’s a matter of the platform and a time. And if you’re reading old, you know, eBooks and blog articles, they’re already past days, right. You need to look at what’s coming up now. Yeah.

Jake (11:45):

And I think on, and then, and then testing with your customer base because you know, there, that’s, that’s the whole trick of it, you know, if you can jump ahead and get a piece of it, I’m always a little bit nervous to build on platforms that like a Snapchat, for example, like, it’s great, but be thinking about ways with a cross channel approach where like, okay, we’re going to, we’re going to harvest customers on Snapchat, but we’re gonna move them into an email sequence or we’re gonna put them into email. Cause that’s a channel we can own, we can do automations and flows that way, but we’re not just purely relying on Snapchat to make revenue for us. So kind of managing the risk a little bit with a cross channel approach is something we really try and work with clients on.

Charles (12:22):

So when you were saying cross channel and the whole master key, right. The part that kinda, I see as a little tricky, right. Once you get that customer to get their email address, like that’s a mastic. Yeah. I get it. That’s like your email address. I can email you. You’re a unique person. What do I do higher up in the funnel? Right? Like when that first, when that person just happens to click on the ad for the first time just happens to view a blog posts. You haven’t pixeled them, but you don’t, it’s five steps before the email. What do you do there to kind of link that altogether? Yeah.

Jake (12:54):

Kind of top of the funnel, middle of the funnel and bottom of the funnel, I think are tofu, MOFU really interesting. Yeah, exactly. Yeah. The little fun acronyms. I would say really depends on kind of like we could make a better idea. I think of this might be let’s run through an example. Do you want to give me like an example that we can kind of riff on of like a company cause like B to B or B to C price point, all of this I’ve seen really influenced it in radical ways. So why don’t we come up? We could do a, we can do a Yeti cooler coffee mug, the coffee mugs. Good. Actually it’s a cheaper product. I feel like there’s an easier

Charles (13:30):

You know, it’s usually a shorter window right of time, so excellent. Okay.

Jake (13:35):

So we’ll say we’re going to sell the Yeti coffee mug and they’re like 30 bucks at retail. So top of the funnel, I mean seeing it, you know, let’s say they came in from a paid ad on social. We can start to see kind of what they’re doing, you know, into your point. We can pixel them. What we’re looking for at that point is we might go look at Google analytics, get started, say for this product category, how fast do people purchase? And if it’s a coffee mug, we know the most 80, 90% by, within the first day. So we know that that purchase decision is going to happen pretty quick. And then you can feel something

Charles (14:07):

Cool can pull out. Right? You can say people that start like first time we saw them to check out and you can run that on athletics, dumped that out.

Jake (14:16):

Yeah. And that’ll give us, and again, we’ll get they’ll have a customer ID or, you know, we don’t know who they are necessarily, but we’ll get a sense of like for this product, this is how these users buy. And if you have enhanced eCommerce with Google analytics set up on your side, it’s pretty straightforward to do this. So let’s say we have that data. We’ll look and say, okay, most of you will buy a mug. And at 30 bucks they’re going to buy in the first day. We’d want to look for any like lurking variables in that. Do we offer free shipping, no shipping? You know, what does that let you know, is it, are there products sold with it? You know, we may look through our order day, a little bit on that. And knowing that people buy within the first day, we would say, okay, we’re going to run some pretty heavy retargeting based on maybe time of day.

Jake (14:56):

And see what that’s gonna look like. And then the next step I would say we could kind of test a little bit is usually what you see as like a lot of people buy in the first day. And then, you know, out at day 21 or day 30, you’ll see another kind of bump where people purchase. So we might to run some tests to say, you know what we’re going to start to do. We’re going to fire retargeting two weeks later. Cause we see a nice, you know, farther out, some customers start to buy then too. And then we can look at our acquisition cost. Is it cheaper than versus like trying to hit them really quick within 24 hours and kind of test it that way.

Charles (15:26):

So, okay. So you’re really looking down at, down to a product or product category or just find that level and really try to map it out on like, what’s the not like what’s the buying journey at the top? Like just from my site, like I sell Yeti pro, like let’s say I’m Yeti and I saw coolers and mugs and chairs, not like what’s that, but what’s the, what is the journey for this coffee mug? And what’s that look like?

Jake (15:50):

Yes. And I love, and let’s say it’s a brand new customer. They’re not, you know, they’re not on an email list or something like that is pure customer acquisition. I think then yeah. Looking at the path length is, is in Google analytics is a really, I think good place to start. Cause you know, if it’s a higher price point or more, there’s more research involved in the decision. That’s where like, you know, search and forums and Reddit can come in, you know, customer reviews, all that kind of stuff that can kind of influence it. But I always say like in our experience, the lower, the risk, the quicker, the conversion. So it was like, it’s a coffee mug. It’s got, you know, 4.8 out of five stars, 3000 people reviewed it, the brand of Yeti’s behind it. It’s a, you know, we took the risk out of it. I’ll probably purchase.

Charles (16:31):

Yep. Yeah,

Jake (16:33):

Yeah. If it’s a new emerging medicine for say you have an allergy or something.

Charles (16:38):

Yeah. And there’s side effects,

Jake (16:40):

Mike and Chris, you know, I might be like profuse bleeding from the nose or something then yeah. You’re going to probably split that path link to convert. It will be a long time for you, everyone purchased that or something.

Charles (16:50):

So would you say, start with kind of the simple products that you don’t, you’re not trying to map out like a 90 day journey trying to map out like a two hour journey. That’s the beginning.

Jake (16:58):

Yeah. And again, I always like to come back to like, what is margin turns or what, where we make, you know, if we don’t make, maybe we make, say 20% of that $30 mug, but when we just sell like crazy the margin, you know, the terms may make up the margins, whereas say the cooler make 50% margins. If I’m Yeti, we might think about, you know, you gotta run some math to figure out where to kind of focus and what the topline prof top line revenue and profit can be from it. So it gets a little bit, a little bit more into the finance side of it, but yeah, low risk decisions at purchasing the first day. Those are, those are great. The more complicated ones that could be a couple of weeks. One of our clients sells really high in backpacks and that can be like eight to nine visits before they purchase. And these are extremely, well-made super high quality, like four or five

Charles (17:44):

Log backpacks. That’s a very

Jake (17:46):

Different customer journey where you might need to drop them into a, you might, you know, get them to sign up for a, a three part series and eat, you know, on Facebook and an email series. And you kind of use an email on Facebook back and forth, some retargeting, all that to orchestrate it.

Charles (18:00):

So if something like that, how do you, once you have that analytics data, is there a way from going on to actually linking that pixel data with the email address or like when right. Cause I’m trying to think, when do, can you see that full, how much I spent on an ad to acquire a customer versus that same customer bought once maybe they buy 10 times? Like how do you kind of, how do you link, where does that get ever yet linked? Or does it, is that because I know Google tries to obscure a file that, you know, the pixel like the visitor down to the actual physical human. Yeah.

Jake (18:35):

And there’s like, for example, if you’re going to Lytics or sessions versus users so you have cross-device challenges. There were, you know, I go, I see something on my phone, I researched it on my tablet, but then I check out on my, on my laptop. So, you know, that can be potentially three users. If I wanted to different IP addresses, it could be three users to Google or, you know, in different sessions. So that’s, that’s all very, gets really messy, pretty,

Charles (18:58):

Pretty quickly.

Jake (19:00):

You know, I would say, you know, you can look at return on ad spend and that’s a, that’s a good metric to start with and that’s kind of what the industry standards are. But when you look at your margin structure in there, you know, you may be able to, to handle, you know, maybe a 252 to 250% return on ad spend because your margin structures are so much better if you’re selling direct. So you maybe make a 70% margin, you can take away lower row, row as number because you’re just making, you know, you just make more cash on the back end

Charles (19:28):

Trying to say that that return on ad spent on that first. Like you’re not even looking at, Hey, we can get them on the email marketing list and sell two, three, four times. You’re just saying, what’s the return on ad spend on that number one sale at that point.

Jake (19:42):

Yeah. And if we’re running a 4th of July campaign for Yeti to sell tumbler drinking cups or whatever, like we might just, you know, time bound it by a holiday or something like that and run it. If you’re running more of an evergreen campaign, that’s a longer term piece, you know, you can segment by the campaign by dates and geography and all sorts of interesting things like that. But to your point, and that’s kinda the Holy grail, right? The, where that user came from. I actually, I know them, you know, there’s a lot of permission based marketing that needs to be built into that where the user can sign up or get some value in exchange for that. And that’s, I think where the next phase is really going as loyalty, Amazon prime is, is built around getting that really good visibility to the user and the loyalty around it. So I think a lot of mid market companies can do some fantastic stuff with loyalty and I will, you know, I’m going to log in to earn my points or get exclusive access or whatever. Now I really have, I can, as a data scientist, I can really match you to all your attributions where you came from device, all that good stuff. And in exchange for that, you know, you’re, you’re getting hopefully rewards through loyalty.

Charles (20:42):

Yeah. Because I’ve always come started at like loyalty rice. Like that user that you acquired maybe through was just kind of, even if they purchasing product, maybe they, they would just happen to be scrolling through some, let’s just say Instagram on Facebook. Right. They were just scrolling. They weren’t thinking of buying, but they happen to see the mug. They liked this full July spot. They just hit buy done. But that user, that may be acquired through PPC. They were searching best coffee mugs they were doing. And you have a whole different campaign. Maybe pay more to acquire that user. But it turns out now they like a coffee mug, efficient auto, and they order nine of them and they just kept buying more coffee mugs. So I actually own like a Getty. So maybe me

Jake (21:24):

I’m the same way. Yeah.

Charles (21:27):

So like, you might have paid less to acquire that person’s or social, but they just bought it as a gift real quick versus that other person PVC is Googling of a coffee mugs. They’re a nerd. They love research STEM, but then now they just keep buying. Is there any way to kind of link up all that data to really understand, Oh, the PPC one is really like, it’s not just convert cause it’s the ROAS is worse. Right? Like it could be significantly worse with UC, but maybe because of, so intent-based, you know, I’m just thinking of a scenario like this where like a very deep and that’s big. Right. And that insight could really change, like how you advertise. Yeah. You’re asking us,

Jake (22:05):

I ask questions. I think that for a lot of ease, it’s easier than ever to get into e-comm. Right. I mean, it’s, it’s, it’s so easy to get into it. It’s also incredibly hard to be profitable or run a good, you know, and you, you can see companies that can get to 20 million and like basically breaking even, or barely producing any profit for the owners because of all this stuff. And so to your point, yeah. Intent versus inspiration, right? Inspiration on social. Oh, my friends are at the beach. Yes. I need that coffee mug that looks like it’d be fun, blah, blah, blah, or, you know, intent of like best coffee mug for commuting. And that’s, I think where PPC starts to weave into like SEO. So if you can look at your, your keyword campaigns and what phrases are kind of working by campaign, you can start thinking about writing blog content and email series is in welcome series and things like that off that.

Jake (22:54):

Whereas and this is where AI and machine learning gets a little bit black box where you can hook all this stuff into Facebook and Instagram and it’ll go find you customers, you know, here’s how much I’ll pay to acquire them. This is what it kind of performs, but you’re not really getting at the deeper, the deeper economics or what’s going to build a profitable business side of it. And that’s okay by that. I’m not saying that’s right or wrong. Depends on your overall goals of the business. But I want to really understand as much as possible what’s driving my not only my top line, but my profit. And if I’m acquiring a lot of customers on Instagram and they’re not that profitable because they buy and they return all the time. Cause it was an impulse buy. You know, I’d much rather like punch a little bit on top on revenue.

Jake (23:32):

But knowing PPC has given me consistent customers because their intent was there. They bought what they were really looking for. They don’t return. You know, and this gets in the weeds a little bit, but I think the way you can tell you can kind of start to track that to track that is with really good UTM parameters are really tagged the links really cleanly. And really this is whiteboard. Sit the whiteboard, really think your campaigns through cleanly and then tag all those links. As I just in 4th of July campaign, you know, 4th of July intent, dah, dah, dah, dah, you know, very, very clean tagging is a good way to get at it. Cause that’ll all flow through Google analytics and then you can start to like have the file folders if you will, on how traffic’s coming. And everything’s kind of neatly organized in your file cabinet.

Charles (24:12):

Okay. Yeah. It seems like what you’re, it seems like the way you’re approaching this, you’re zooming very like deepen on a very specific product campaign. Like, like an analyzing, just like one very particular flow, like trying to isolate us, analyze us and then move to the next one is kind of what it sounds like your approaches,

Jake (24:32):

I think. Yeah. And especially in mid-market econ, there’s so much to get done. And I would S I would say analyzing off profit or whatever you’re trying to do with the business is super important because again, over and over and over again, I’ve just seen so many businesses that are doing five to 50 million, and nobody has any idea kind of how they’ve done it, or they built it on some things that were a little bit shaky and, you know, like Amazon, for example, you can look up a store on Amazon. That’s great. And that can be good for cashflowing it. But if you build the whole thing on Amazon, like all of a sudden you get a letter from Amazon saying, Hey, we’re going to take two more margin points this quarter, you know, what are you going to do about it? Well, you don’t, you know, you build something on land, you don’t own even were renting it. And so being very thoughtful on that piece of it, and then back, you know, then once you understand one floor or one product or one economic part of the lever, you can expand out into the other parts of the biz.

Charles (25:24):

Yeah. I think I’ve talked to, I was talking to our founder once and kind of said, how did you grow? And because I tell him this marketing channel, they’re like, yeah, but like, they like astronomical growth within 12 months. And I said, yeah, but when I like kill that challenge with doing this all the time, I was thinking, wait, why, why are you doing that? They’re like, that’s not, ultimately, we just, we knew we could do that to like, you know, like that was like the steroids right on, it was bad. It was like the crack, right. They were going to do it. They knew they King of beg, but they knew it was their building on sand. And they knew as soon as you get to a certain point, they could hire, you know, X number of people, these very specific roles. Then they could do everything internal and go from this like shaky foundation to now we’re going to really build it the way we want to. I always find that’s like the most insightful, but they did it with such like they knew we had to get to this sort of revenue goal so that we could hire XYZ so that we could really run these campaigns. And now we can really do it because they couldn’t do it. They couldn’t start from zero. That was the thing. And they knew us. That’s awesome. I think that’s a huge piece of it. And

Jake (26:27):

Actually economics. I mean, a lot of this stuff you’ve seen set some implosions with direct to consumer companies because they’ve raised money from investors and all that money basically goes to run paid ads. And they’re not thinking about some of the unit economics. There’s some really good professors at Wharton and Peter fader and other guys, Daniel McCarthy, he’s an Emory and they do some really good breakdowns on like pub IPO is like [inaudible] as companies go public and they’ve looked at like Slack and Peloton and Wayfair, and they go after Wayfair pretty hard off the numbers. And it’s, it’s really interesting cause their acquisition costs are so high and when you figure and all these other things, but the lifetime value doesn’t make sense. So they’re kind of propped up with you know, investor dollars and, you know, the unit economics sooner or later we’ll come home to roost. So yeah, if you know what you’re going to do and like, look, we got to your point, we’ve got to do some, do the steroids to get the thing going to then transition out. I think just being very thoughtful at the outset of what that is.

Charles (27:23):

Yup. What are some of the things that you see as people’s side to scale have come up or either that they’ve thought of, or haven’t thought of as a scale? Cause it was always kind of surprising. It’s almost like like a game of golf, right. Where you have to, like, you have to do, they have to like get good at one thing, but then as soon as you do that, you’re basically now doing something else. It’s like, not that thing anymore. Yeah.

Jake (27:46):

Oh man. It’s the, what got you here? Won’t get you there. Yeah. All the time. It’s like, I think what scales. Yeah.

Charles (27:52):

Yeah. I think

Jake (27:55):

A couple of different things, one I’ve seen a lot is they don’t have good roles to find on the team like, Oh, so-and-so, you know, we have a digital marketing team. Well, what does that really? What does digital marketing really mean? So, you know, good roles, accountability, like here’s, you know, this person is doing SEO and this how we’re going to hold them accountable. The other thing I’ve seen a lot is, is the greatest scale is it’ll take a lot more capital than I realize growth takes cash. And I think that’s something that can trip people up and go really, really fast, but say you’re on Amazon and you don’t get paid for 90 days. Like your cashflow cycles will each alive on that. So that’s a, that’s another one I’ve seen. Honestly the other one too is just like not knowing margins or being super tight on not understanding their margins and all the costs that go into that.

Jake (28:38):

And then and then like, what are the, how do we, you know, what are our goals for this year and this quarter, and then breaking them down by month and week and like every week, just grinding and looking at the data and making tweaks and modifications off that. A lot of folks don’t even really know what they want to do on every 90 days or, you know, why do we want to grow 30%? Okay. Well, exactly. How do you grow 30%? Well, we’re going to do it on Facebook ads. Okay. What kind of Facebook ads on mobile? I, you know, like you got to get very specific cause until you get it down to that level, you can’t really measure it. If you can’t measure it, you’re just kind of throwing stuff at the wall and you can get lucky. You can get lucky that way too, but I’d much rather not use luck. I’d much rather use science as much as possible.

Charles (29:18):

Defining if you’re saying for your quarterly goal, we’re gonna run 10 experiments that we don’t know which one’s going to, like we’re going to run some mobile free mobile ones, two desktop on two, on Facebook for an ad or whatever. And you kind of just define these, my experiments. That’s one thing you’re doing. You’re still doing science and you don’t. Yeah. And that’s a measurable thing. Did you run 10 or did you only, you only actually get five, right. Done, whatever. So at least that you can actually measure the experiments and out of that, the next quarter, when you set your goals, you might come back and say, okay, we found out eight of them were dogs, but these two, now our quarterly goal is to Tenex down on these two experiments around it. So that’s, as long as you kind of know, you know, you can still try different things. You just need to know that you’re trying things like in separate the mouse.

Jake (30:05):

Yeah. And I would just, I was a, I was a physics major in school. So one thing I did when I wasn’t like the most brilliant math talent and for sure, but one thing physics taught me is like, what are our givens? And like, what are we trying to solve for one of our givens? You know? And what are we solving for? It’s very simple, but you can kind of use that logic. Yeah. We’re going to try user generated content on Instagram and we’re going to three ads around these types of value propositions. That’s awesome. Like totally go for it. You know? And more often than not with that stuff, you’ll be shocked. You know, number three, which you thought was awful, outperforming everything by 10 X and you know, that’s, that’s how you unlock this really good customer acquisition funnel by testing some of that kind of stuff. But a lot of times, as they scale, as they get to scale, you know, they don’t really even have some of these pieces in place. You know, they’ve done some of that stuff. What’s got them success and is awesome. But you can’t repeat that in a repeatable way or have a process for making repeatable experiments. So maybe that’s kind of the biggest lesson I’ve seen is be very thoughtful on,

Charles (31:04):

On that part of it. Yeah. I’ve definitely, I’ve seen folks tickets and notes out, but they kind of see something that’s working. And instead of doing the experiment in themselves, there was copying someone else’s experiment and it gets them further, faster. But then like you said, as a scale, they’re not seeing why, like they don’t have all the failed experiments. Right. So they don’t know why there’s other things that don’t work and why this one did. So when they try to, and then whatever Instagram stops working and they need to do that same thing, another channel, but they can’t figure how to replicate that on Snapchat anymore. They just dead in the water. So I’ve definitely seen that. And a lot of times it’s just because people are just reading something and copying something, but they’re not really understanding why they did the thing in the first place.

Jake (31:48):

Yeah. Hacks are great. Right? Yeah. We all want, we all want a shortcut. Right? The problem with the hack though, is you oftentimes you don’t understand why it worked as well as you probably need to. And if it works for a while, you start to build this false confidence in it. You know, the whole joke that wants you to have a hammer, everything looks like a nail. And then all of a sudden you walk in and everything’s a screw. It’s like, Oh, I can’t really pound a screw into the wall without a lot of effort and mess. So yeah, I think that’s a very common thing. And especially with online, cause you can go on Twitter or follow along with some, these really smart PPC or paid media experts. And you know, you know, you can copy some of their methods and get it working.

Jake (32:25):

But again, like I’ve seen over again where the algorithm changes, right. And then all of a sudden it doesn’t, you know, your RO ass blows up and our associates with horrible lives. Cause they’re on these massive, like emotional swings or ads are killing it one day and then they’re not the next and then they’re back and you can see them on Twitter, just like self-destructing cause they’re so like freaked out all the time. And I always think of like, when it comes to customer acquisition and retention, you really want to think of it. Like your stuff like your 401k, right. You’re going to want maybe a few risks stocks, but you also want some really good blue chip. You know, you want some bonds, you want a little bit of diversity. Like you want some email humming, you want some SEO going, you know, some, some forums or referral traffic coming in versus just betting the whole farm on a, on Instagram ads or something.

Jake (33:07):

Oh, I guess. All right. I think this is a good place to wrap it up. Don’t put the phone on Instagram ads that the garden has always just say, you bet the garden, not the farm. This is a quote of the day, but the garden, not the farm on Instagram. All right. Jake, if people want to find, you kind of learn about what you’re working on, what can they do that? Yeah. I have two websites. I teach digital marketing and eCommerce. It’s one is that on digital There’s a bunch of stuff. I just kind of open source for free. If people are interested in how you can use like user empathy and digital marketing cross-channel so that’s there. And then the company we run is, is And for folks that are interested, we have a thing called and it’s alive 24 hour index.

Jake (33:48):

We’re grabbing all sorts of data, normalizing it. And you can see how your site compares to chefs, by geography, for revenue and transactions and all sorts of things, your average order value, conversion rate, all that stuff. So we’re building that out and kind of opening it up to the community. So people have a data set they can compare to. And if they’re way off, they kind of know where to start. If, Hey, we have good traffic, but conversion rates was really down and we should focus on onsite conversions. It’s nice when you can compare yourself to some sort of a benchmark. So yeah, that’s a good place to start. I like that. I will definitely turn a show notes so people should check that out. Awesome. That’s super helpful. Thanks for coming on today. My pleasure. I very much appreciate the invite and thanks for putting content like this out there. It takes a lot of work to put together a podcast. So tip of the hat for doing the work

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.