George Jagodzinski (00:00):

Today we learned that Amazon isn't nearly as calculated as we all think, and that retail media is going through an inflection point that can't be ignored. I'm joined by Jeff Cohen, chief business development officer at Skai. Jeff spent the last few years as the principal evangelist at Amazon Ads, as that business grew from $37 to over $60 billion. He's lived every chapter of the e-commerce story, and now he sees an inflection point most people are just starting to write about. We dig into why retail media is at a tipping point and how brands can see the full picture across hundreds of channels. Please welcome Jeff.

(00:42):

Welcome to Evolving Industry, a no BS podcast about business leaders who are successfully weaving technology into their company's DNA to forge a better path forward. If you're looking to actually move the ball forward, rather than spinning around in a tornado of buzzwords, you're in the right place. I'm your host, George Jagodzinski. Jeff, thanks so much for being here.

Jeff Cohen (01:09):

Appreciate you having me.

George Jagodzinski (01:11):

Jeff, since we last spoke, I read something that you talked to a recruiter and you said there was only one job you would take, it's evangelist at Amazon, and then somehow that manifested. Are you a Jedi? Walk me through how that happened. What was that like?

Jeff Cohen (01:24):

Yeah, well, she wasn't actually a recruiter. She was a former Amazonian who tried to recruit me. She wanted to recruit me to do evangelist type of work for her company, and I said the only company I would ever go do that for is Amazon. I felt that that is what, in my career, I felt like that is what I had been doing for a number of years. When Amazon Ads first came out in 2016, nobody really knew what it was, how to use it, what the terminology was, what the strategies were. So when we were developing software for that, I was also out on the circuit, if you will, teaching people how to use Amazon advertising.

(02:05):

She happened to know somebody at Amazon and reached out to them and said, "Hey, if you've ever been looking for an evangelist, this is the guy to hire." That led to what they call at Amazon a coffee chat, which is like an informal interview. And at the end of the interview, the hiring manager said, "You scare me." And I said, "Is that good or is that bad?" And he said, "Well, everybody here is very engineering product focused." He goes, "You are product marketing focused, and we don't have a lot of that. So I'd love to bring you on my team to be this connectivity between the developer community that we're trying to build and the market, to try to take what we're building and share it with the market, and then take what the market is looking for and bring it back to the product teams." And so, that was kind of how I got the job as the Amazon Ads tech evangelist.

George Jagodzinski (03:06):

I love it, man. I think I'm going to set a goal for myself this year to hire at least one person that scares me. We'll see how it goes.

Jeff Cohen (03:13):

It's a very Amazonian concept, where it wasn't the job that he had originally designed, but they hire good people. Well, I consider it good people. They hire good people that meet a particular bar or a level, and then they put them into roles to help them be successful. And everybody at Amazon is considered fungible, which I think is very different than how other companies work. And when you're hiring, you're looking for somebody with a very specific skillset. I'm looking for somebody who understands social media, podcast editing, or whatever it is. Amazon doesn't look for that skillset. They look for a skillset across a leveling guideline, and then they believe that you can move into any type of role.

(04:01):

So what my technical title at Amazon was a principal business development tech, something like that. And in that role, I could have had that job under ads. I could have had that job under retail. I could have had that job under shipping if I wanted, because the skills that you use to do those jobs were the same skills, you were just applying different market conditions to them.

George Jagodzinski (04:26):

Yeah, that makes a lot of sense. So then you're within the belly of the beast. What do people on the outside looking in, do you think they get wrong about Amazon?

Jeff Cohen (04:36):

The one thing that I say is that Amazon isn't as calculated as people want to believe that they are. A lot of times what I've heard in all of my time is, "Well, Amazon wants us to do this, or they don't want us to do that, or they're trying to stop this, or they're trying to stop that." Most of the time, what we see as sellers or tech providers or brands are unintended consequences. So Amazon has a goal to increase the number of people using video, and in doing that, it may create unintended consequences that weren't initially thought of when they put that together.

(05:18):

And so, a lot of times sellers and brands see that, when Amazon goes to stop bad actors from posting paid reviews. They write an algorithmic formula to try to identify these reviews, they remove them, and they capture 95% of what they need, but maybe 5% of people get caught up in that, and then you have to kind of clean that up. Those small percentages at Amazon, because of the scale of the business that they have, can impact a lot of people. And so those really become unintended consequences.

(05:55):

So, that's one. It's not as calculated. I think the second one is that Amazon systems, their ability to build rigor, to build mechanisms, and to lead with the leadership principles are not just things that were written into books. They are actually done and used and managed on a daily basis, and they really do drive every conversation that occurs within the Amazon system.

George Jagodzinski (06:25):

And then when you were there, you saw tremendous growth, right? And now you're at this point where you decide to leave. And some people might say, "Jeff, you're crazy. What the hell are you doing leaving Amazon?" They're the big elephant in the room. And I think, I talk to a lot of people these days, especially with AI, they think that they're really questioning what they're doing. They think it's going to fundamentally change. But you and I have been around a while. We've seen patterns where the timing's just not right. I remember when early video days, when Brightcove was around, we were talking about in-video purchasing for e-commerce. Never really happened. But you see an inflection point happening right now, obviously, that you believe in. I'm curious, maybe first starting with the human journey to do that. You're talking to your family, you're like, "I'm seeing this thing." What's that look like?

Jeff Cohen (07:10):

Yeah. So, listen, I think there's the mathematical equation behind things. What am I making at Amazon? What do I have the projection to make? Where's the stock going? What type of upward mobility do I have? I think you have to constantly question those type of things. And that was something that I did almost on an annual basis. I would kind of set goals for myself for the next year of where I wanted to be, so that I could look backwards and I could see if I would get there. And then I'd take a predictive model of, do I believe I can get to where I want to be? So I think that that's one side of it.

(07:47):

I think the second side of it is, from an individual learning perspective, do I have capacity to continue to learn within my current role? Do I have upward mobility? Can I get increased scope? Quite frankly, when I was at Amazon, I started to become comfortable in my job. I had been doing the same thing for two or three years. I was successful in doing it, in terms of my reviews and the processes.

(08:13):

And at Amazon, when that happens, you really have two options. One is to move into another department, like I was talking about, or to change your job and your scope of work within your current role. So I knew some kind of change was going to happen. Either I was going to change and take on a different responsibility, or I was going to leave to go to a different department, or I was going to leave to go outside of Amazon. I decided at that point that I really wanted to kind of look at the market.

(08:43):

Being in the role that I was in, I had this really interesting view where I was able to work with hundreds of technology companies, thousands of agencies, and really get this sense of where the market was going. I had spent a personal journey of about a year really diving into AI, trying to understand fact from fiction, trying to understand how it can drive me individually and how it could drive businesses. And when I met with the team at Skai, they actually originally came at me with a product management job, and I talked to the head of their product team. It just wasn't the kind of work that I wanted to be doing, but I was really intrigued by what they were doing, because I got this really deep under the hood look at where their business was at, but more importantly, where they wanted their business to go.

(09:37):

And when I sat back at the end of the day, I said to myself, if I was going to go out and build a company, these are the things I would be building. And so, when I looked at the opportunity to come into an enterprise organization, be on the C-suite of the business, help drive the business forward, that led me to go back to the CEO, and effectively what I said was, "I don't think this is the right role for me, but I think this is the right company for me. So if there's a position that fits what I'm looking for in the future, I'd love to figure it out." And then a few more months passed, and that's when they had an opening on their business development team, which fit more in line with where I had interest and where they had interest. And here we are today.

George Jagodzinski (10:28):

I'm seeing a trend here, where you're very clear about what the role is that you want, regardless of what they're telling you they need at that moment. I love that.

Jeff Cohen (10:36):

When I was on the executive team at Seller Labs, I grew an individual confidence that either one, I wasn't going to get fired, right? I had grown past in my career to know that I could be successful in a role and that I wouldn't necessarily be fired, but if anything ever happened at the business and I was let go, because there's other reasons you could be let go besides performance, I knew that I had a really nice safety net within the market through the connections that I made that would help me find another job.

(11:10):

And that is an unbelievable level of confidence that you have when you go into your next conversation, because you don't need what they're offering, you don't need what they're paying, and you can start to kind of dictate those rules of what you're looking for. And I understand that's a maturity that you get to within your career, but I've even tried to instill that into my kids, who are very early in their careers, to say, when you can speak with that confidence, it comes across to the people that you're speaking to, and you can really, truly sit down with somebody and talk about, how do we build something great together?

(11:49):

Chalk up a lot of it to the confidence that I built back at Seller Labs, and I proved myself when I was at Amazon. I definitely had imposture syndrome. I didn't believe that I belonged there when I first joined. And so I was proud of myself, if you will, for being able to not just succeed, but thrive within the Amazon culture, and then parlay that into what I'm doing today.

George Jagodzinski (12:16):

Love it. I love that story. And so now, moving past the human aspect of it, you're seeing this inflection point as you make this decision. What is that inflection point that you're seeing?

Jeff Cohen (12:26):

Yeah. So what I was seeing in the market was that, and it's kind of funny, because what I saw a year ago, I'm starting to see people write about today, and it's along the same lines, which is that there's a maturity that's happened within performance advertising, right? There's only so many keyword impressions that Amazon can serve, and there's only so many brands that can bid on those impressions. And so, advertising was very simple in the early days. You bought a keyword, it was matched to a search term, and using latch touch attribution, they would tell you whether a sale occurred.

(13:03):

But as that market has evolved and we've gotten into multi-touch attribution, it's opened up a whole nother world to say, well, what other factors are impacting the buy? And as brands have grown in the number of retail media networks that they participate in, they need to understand how that inflection is happening when they go to advertise on Amazon, but it drives a sale at a Walmart store, or they go to place a television ad on ESPN and it drives a sale to their website, or they place an ad on Prime Video, it creates discovery at Amazon, but then it creates a purchase at Home Depot.

(13:49):

And all those things are part of the non-linear journey of the shopper today. There's nothing wrong with what Amazon's doing, and Amazon's great at what they do it, but it's just one piece of it. If you look at e-commerce as a pie of total commerce, it's still a relatively small portion of the overall commerce pie. And so, while Amazon is the large e-commerce giant, once you start to put Walmart and in-store into that equation, all of a sudden they're equal in size, right? And now you start to add these other omnichannels, and that's how enterprise brands are growing brands.

(14:29):

And so, that was really interesting to me, and I saw how brands were starting to shift their mentality of upper funnel advertising to its impact on performance marketing. And so, I wanted to put myself into the technology side of that, where we're connecting all of the dots of what's happening and helping brands and agencies drive that omnichannel growth.

George Jagodzinski (14:56):

I feel like I've heard 25 years of people struggling with attribution, and no one really seems to feel like they've solved it.

Jeff Cohen (15:02):

Yeah. Listen, I don't think you ever solve it. I don't think that we're solving anything, a problem today that didn't exist 25 years ago. Measurement requires you have uniform data that allows you to be able to have a foundational base, and then you have to create the data measurement points and then track them with consistency. And so, I always like to say, nobody really knows if your ROAS is 3.2 or 3.4, but if you're tracking with consistency and it changes from 3.2 to 3.4, that you know that you've incrementally improved.

George Jagodzinski (15:41):

I feel like a lot of people are listening right now when you said data and they're like, "Crap, my data sucks." What am I supposed to do?

Jeff Cohen (15:47):

Yeah. I mean, listen, that's where all brands have to start. You have to start with that foundational layer of the data cleansing. How do you bring in different definitions of data from different retail media sources to have some common definition that allows you to look across those channels? Those are all the challenges that you're facing, whether you're an enterprise brand or whether you're a digitally native brand or whether you're an emerging brand, right?

George Jagodzinski (16:19):

Yeah. Yeah. Totally. Jeff, I've heard you speak about in retail media bubbles. I'd like you to talk me through, what are the bubbles and where are we at on them?

Jeff Cohen (16:30):

Yeah. Bubbles are constantly forming and popping. I think you kind of started this conversation off by bringing up one of the ideas, which is like, do people have the right idea, but is it the right time? And a lot of times when you see bubbles burst, it's not necessarily because anything is bad, it's because the market wasn't ready for it. And for those of us that are aged, I had more dark hair a number of years ago than I have today, we know that from the bubble burst of the internet one, the internet two, there were a lot of great companies doing great things, they just weren't there at the right time. And there are companies today doing what those companies did because the market finally caught up.

(17:16):

So obviously, one of the biggest bubbles that's out there is AI, and I think that one of the revelations that we've come to have with AI is that we all know it's going to be game-changing, right? There's no doubt about it, but I think businesses are just now really getting into the, how does this impact my organizational structure, my organizational design, my organizational operational efficiency?

(17:46):

And so, we've developed some really cool in-house tools inside Skai that have allowed us to do things that used to take BI teams weeks to come to the data. So we created an internal data lake that allows us to ask questions, allows us to see who are the movers and the shakers, how is business going? And so, we're starting to see a lot of that impact internally on our business. And then I think there's the secondary side of the AI, which is like, how do customers use our tool in an AI world, which is their personal evolution into using the AI. So I think AI is one.

(18:29):

I think one of the other, maybe it's a bubble, I don't know, but there's definitely been a massive growth around full funnel advertising, the expansion of DSPs. We used to think of DSPs as simple retargeting, remarketing type of functionality, but now they're true growth channels for finding audiences and expanding the reach of your brand. And so, I don't know if it's a bubble or if it's more of an evolution of how brands are thinking today.

George Jagodzinski (19:00):

Yeah, and it can be so overwhelming for people, especially with small teams, maybe a mid-market brand. With all the many years of experience you have, if you were coaching someone at a mid-market brand right now on what they needed to do with their small team, what do you think the advice is?

Jeff Cohen (19:16):

Well, I think the first is, from a cultural perspective, you need to be driving AI as something that needs to be ingrained into everything that they're doing. I manage a team and I tell my team this all the time. They need to be spending the extra time to figure out where AI can create efficiencies and where it can and cannot be used.

(19:38):

So for example, the data lake. Our company came out with this data lake. I've asked our team for this quarter, for Q1, to do everything the old way while also asking the data lake the question. And what I'm looking for is to understand, how often is the data lake right? How often is the data lake wrong? And if we can get to a statistically relevant number to know that the data lake is right 95% of the time, then we can stop doing it the old way and just be dependent on the new way. So that's one way that I have kind of coached my team to think about it and to do it.

(20:17):

I think the second thing for brands to be thinking about is, what's going to move the needle for your business? There's so much noise in the market, right? Go on LinkedIn and just fly through your feed. There's so many people telling you what's right, telling you what's wrong, telling you where you should put your focus. Have a very specific strategy for your ICP, so know who your ideal customer profile is, and make sure that the actions that you're taking are actually going to impact that business. Don't get caught up in all the other hype and noise that's out there that can't actually drive what you are trying to do.

(20:57):

I had a boss give me this recommendation in the past, which is that 25% of your time should be on your administrative stuff, 50% on your time should be driving to what you have and what you're doing for the here and now, and 25% of your time should be working on kind of the future. Because if you're not planting the seeds for the future revenue or the future growth, then you're never going to get around to actually getting to them.

George Jagodzinski (21:26):

Yeah, and it could be horrifying if you just allow yourself to get sucked into the LinkedIn feed, and you try to go sleep at night, you're like, "I'm not doing enough. Everyone's doing it better than I am. I'm a failure." What I do find though, that point of the parallel path thing, there's a lot of that going on right now. I know even with some of our clients that are resistant to things, we might just on our own time do the same thing that they're doing in parallel, but just do it with AI, and then you can hand it to them and say, "Hey, look what we did," and now all of a sudden the hesitation's gone right now.

(21:56):

But I don't know about you, but I'm finding a little bit frustrating in that I pop back and forth between every different model I possibly can every day. And this week, for whatever reason, Gemini is junk and all of a sudden Claude is way better, and I don't know what's going to happen next week. How do you personally manage keeping your finger on that?

Jeff Cohen (22:17):

Skai is a Google shop, so therefore I've learned to use Gemini because it's ingrained into everything that we do, whether it's-

George Jagodzinski (22:26):

I take back what I said, then, Jeff. I'm sorry.

Jeff Cohen (22:28):

Well, listen, I don't disagree with you. There are at times that you'll copy and paste something out of Gemini, put it into Claude and get a different response. I personally have narrowed it down to two. I use Claude, where I'm learning to kind of build, I don't remember what they call them, but you can build projects, bring in connectors. You can do all that in Google as well, I just kind of feel like Claude is a little bit of a better system for me.

(22:53):

And Claude is kind of attached to all my personal stuff, and then the Google is attached to all my work stuff. So when I'm... I don't write emails anymore. I go into Gemini, I click that little icon, I throw in all of my notes of what I want it to say, and then I let it generate the first version for me, and then using that, I then tweak it.

(23:18):

Now, I wish that it would have some memory behind it, so I didn't always have to tweak some of the things that I feel like I'm always tweaking, which is what you do get with Claude because I can build out more of a memory-based system. But that's how I've tended to use it in terms of those larger systems for personal use.

(23:39):

And then, for what it's worth, I do it for all my personal stuff, too. So I mentioned I have kids. My kids are in college. I had my kids send me all of their syllabus and I created... This is the first way I learned how to do some of this stuff. I took all of their syllabus and I put them into one project on Claude, where I was then able to ask it questions like, when are teachers available for office hours? When are my tests? When can I come home for winter break? All those types of things, where I now have this database that I can ask questions to, versus these eight different PDF sheets that I had before.

(24:18):

And you just kind of start to learn on some of those personal things, that you're like, oh, wow, that's really easy. That was really nice. And now, how do I apply that into my work world?

George Jagodzinski (24:29):

Yeah, the personal tool. Well, that's a fun project. The personal to work makes a lot of sense. I've been using it for nutrition and fitness, and I think I've over-prompted it, because it's now really passive-aggressive and snippy towards me. But getting back to the future, I think we're going to see a lot of normalization of all these models over the next year or so. And if we look at retail media, fast-forward three years from today, where do you see things... How different are things in three years?

Jeff Cohen (24:53):

Yeah. I mean, most of the retail media providers that are out there are all, I like to call it they're pushing their plus series. So whether it's PMax or BrandPlus or Performance Plus or their AI Max, they all kind of have their own versions of this, where they want their technology to add the context to what you're trying to do to determine where your ads go.

(25:19):

Now, we as marketers have shied away from wanting that to be our primary driver for our advertising, and the reason why is because we feel like it doesn't give us control, right? A lot of advertisers have worked under the confines of control. And so, that's where a tool like Skai starts to come in, where Skai is saying, "Hey, we're going to work with these tools."

(25:42):

And you have to work with these tools, because these tools are what are driving Rufus ads and Sparky ads and Gemini ads, right? That's what's driving them. So you have to participate in them, but can you add some layer of intelligence to it, where you're still maintaining your budgets, you're managing the return on investment that you're looking for, you're taking the cross-functional output of what's happening and bringing that in to give it additional context.

George Jagodzinski (26:11):

Yeah, that makes a lot of sense. I'm excited for our future. I'm optimistic. Jeff, I could talk to you all day about this. I've got two fun questions for you to finish. I know you're a whiskey guy. So if Amazon is a whiskey and Skai is a whiskey, which one's which?

Jeff Cohen (26:30):

Oh, wow. That's a really great question. I would probably say Amazon is probably more like 16, right? It's a deeper aged, a nicely finished. I think Skai is probably... I don't know if you follow the Bardstown line, but where you can have this really great blended product, but then you can have it finished in the prisoner or the silver oak, where Skai's kind of giving you these different flavors that would give you different reasons to want to enjoy it in different ways.

George Jagodzinski (27:05):

Oh, I love that. That makes a lot of sense. And then the last is, through your many years of experience in life and work, what's the best advice you've ever received?

Jeff Cohen (27:13):

I'm going to say that it was a lesson that I learned in my first job as an intern. Always be truthful and honest about the mistakes that you make. As an intern, I had to do a mail merge. For those people that are too old, that's where you used to take your CRM and then Excel and you put it together so that it put everybody into the letter correctly and into the envelope. And I was doing a mailing campaign.

(27:38):

About a week or two after the mailing campaign went out, I got some return mail and I realized I had never printed the letter on company letterhead. I did not sign the owner's name on the letter. And I went to my boss and I brought this up to her and she goes, "It's okay. Don't worry about it." And I felt really guilty, so I went into the owner's office and I told him what I did, and he said to me, he goes, "Listen, you're going to always make mistakes. Mistakes are part of how you're going to grow. It's more about learning from the mistake that you made. And by the way, if I had known that you knew that you made that mistake and you lied to me and you covered it up, I'd fire you."

(28:19):

And that has stuck with me my whole career, that... We were talking about having confidence and things like that. Be open, be honest. Failure is not a bad word, right? Own mistakes when you make them learn and be able to demonstrate how you're going to not make that mistake in the future. Because at the end of the day, baseball players make millions of dollars for hitting two and a half out of 10 times, right? So we're not supposed to be perfect.

George Jagodzinski (28:53):

Yeah. Oh, I love that. Yeah. Builds more confidence in yourself and people around you have more confidence. Love it, Jeff. Thanks so much for being here. Really appreciate it.

Jeff Cohen (29:02):

Appreciate you having me.

George Jagodzinski (29:05):

Thanks for listening to Evolving Industry. For more, subscribe and follow us on your favorite podcast platform, and pretty please, drop us a review. We'd really appreciate it. If you're watching or listening on YouTube, hit that subscribe button and smash the bell button for notifications.

(29:18):

If you know someone who's pushing the limits to evolve their business, reach out to the show at evolvingindustry@intevity.com. Reach out to me, George Jagodzinski on LinkedIn. I love speaking with people getting the hard work done. The business environment's always changing, and you're either keeping up or going extinct. We'll catch you next time, and until then, keep evolving.