IN CLEAR FOCUS: Jeff Greenfield, co-founder and CEO of Provalytics, discusses AI-driven marketing attribution in a privacy-first world. As third-party cookies and user-level tracking decline, Jeff explains how his platform, Provalytics, combines marketing mix modeling with the granularity of multi-touch attribution. Learn how marketers can measure effectiveness across channels, including “no-click” media like CTV and podcasts, while maintaining privacy compliance and gaining actionable insights.
Episode Transcript
Adrian Tennant: Coming up in this episode of IN CLEAR FOCUS
Jeff Greenfield: With Provalytics, what we’ve done is we have taken the statistical techniques that are proven from marketing mix modeling, and we have taken those and merged them with the granularity that we had and multi-touch attribution.
Adrian Tennant: You’re listening to IN CLEAR FOCUS, fresh perspectives on marketing and advertising produced weekly by Bigeye, a strategy-led full-service creative agency growing brands for clients globally. Hello, I’m your host, Adrian Tennant, Chief Strategy Officer. Thank you for joining us. Measuring advertising effectiveness while respecting consumer privacy has become increasingly challenging. Respecting consumer privacy has become increasingly challenging. The recent announcement that Google will no longer phase out third-party cookies in Chrome highlights the ongoing uncertainty that marketers face. As the industry continues its shift toward privacy-first approaches, businesses need new solutions for attribution and measurement. Our guest today is an entrepreneur and innovator who’s been at the forefront of marketing attribution technology for over 15 years. Jeff Greenfield is the CEO and co-founder of Provalytics, an AI-driven attribution and planning platform designed for a privacy-first world. Previously, Jeff co-founded C3 Metrics in 2008, one of the first enterprise-scale multi-touch attribution platforms. His insights have been featured in leading publications, including the New York Times, The Wall Street Journal, and Investors Business Daily. To discuss the future of marketing attribution, I’m delighted that Jeff is joining us today from Portsmouth, New Hampshire. Jeff, welcome to IN CLEAR FOCUS.
Jeff Greenfield: Thank you so much, Adrian. It’s a pleasure to be here today.
Adrian Tennant: Well, you’ve been involved in marketing attribution technology since 2008 when, as I mentioned, you co-founded C3 Metrics. So, what inspired you to create Provalytics in 2022?
Jeff Greenfield: That’s a great question because, you know, attribution to those of us in the space and measurement, it’s a great space to be in. But as a marketer, it’s not very sexy because it it involves numbers, which typically, as creative folks, we tend to shy away from. So I had exited my former company right before COVID, thought I would never go back to measurement. I’d already done 12 years in the space and, honestly, was looking for something else. But then what happened is it was a series of steps that occurred where I saw the whole industry was shifting. Changes were occurring and how the advertising industry was working online that was going to change the shape of measurement. Everything from privacy to the GDPR in the EU. And all of these things came about. And even on top of it, as a result, Google Analytics required folks to actually update their code. So there were all of these things all happening at the same time. And I looked at this, and I said, “There’s never been a point in marketing where so much attention has been on measurement, where people from the C-level on down are focused on it because they have to be, because they have to update that code.” And I said, “This is a great time to get back in because what I had built before at C3 Metrics was no longer going to function. It wasn’t going to work anymore. So former customers, colleagues, clients of mine were gonna be left without a viable solution for them.” So that’s what led me to kind of reenter the same marketplace, if you will.
Adrian Tennant: In today’s privacy-focused environment, what are the main limitations of traditional attribution models?
Jeff Greenfield: Yeah. I mean, it’s night and day. If you go back to 2010, let’s say, so 15 years ago, marketing attribution, and you could say the whole advertising industry online, it was kind of like the Wild Wild West. There was no privacy regulation, so you could collect pretty much anything that you wanted. And so back then, we were able to build models not only off of your first-party data, and first-party data is the data you have on your website. So, you know, knowing how someone clicked through to your website. But we can also collect data from any time an ad of yours was shown. So if someone was on the New York Times and they happened to look at an ad, we could see that they looked at the ad. So we could stitch all of that data together. So that was the first thing is that you could collect 100% of all of the impression data, the ad viewing, and also all of the ad clicks. So that’s the first piece, and that’s gone away. There’s pockets where you can get that data, but you used to be able to know when someone actually an individual viewed a YouTube ad. That’s gone. You used to know when someone viewed your Facebook ads. That’s gone as well too. And so when you think about YouTube and some of the Google properties and you think about Meta, that’s about 80 percent of most of the advertising. And then we also have these new platforms like TikTok. You can’t get the view data from there. You can’t get it from Snap. You can’t get it from Twitter, which is now X. So there’s a lot of holes that’s going on in all of this kind of viewing data. So now you can only see the endpoint. Because if we step back, when we think about how marketing actually works, as a marketer, I invest dollars to buy eyeballs, to get people to be aware, if you will, and you’re buying what we call ad impressions. And the idea is that these ad impressions, if the message is right and resonates, that will lead to awareness. And when awareness builds up to a point where someone will say, “Hey, you know, I’m kind of interested in it,” they’ll walk into the store. So if you’re a retailer and your advertising is hitting properly, then you’ll get more foot traffic. But in the online world, that’s translated as someone coming to your website, so that’s a click. In order to do proper measurement before all these privacy concerns, we would wanna see all of those impressions at a user-level and see all of those clicks. Now all we can see is the end result, which are those clicks. That’s probably the biggest shift that’s occurred in the privacy world. The other big shift has to do with the app environment and the mobile environment, which is where most people live today. When iOS did that big update a number of years ago, people had to opt in. They had to say, “Yes. I wanna allow this app to track me.” And so most people, of course, did not opt-in. And so what that means is apps like Meta and others are kinda blind to how people are making their way around the app world. So that piece is kinda missing out as well too. So there’s a lot of holes in what we used to do, which makes it kind of impossible to do it the way we used to do it from a measurement perspective.
Adrian Tennant: So, Jeff, could you explain what makes Provalytics different from other attribution solutions, particularly in terms of methodology and approach?
Jeff Greenfield: When we saw that these changes were occurring in the marketplace, we started to say, “Okay. Well, this world is starting to remind us of how things used to be before there was digital.” Before digital marketing, marketers never had user-level data. We would run TV ads, radio spots, newspaper, direct mail. We’d run a campaign, and then we would sit and wait 3 or 4 months and see how it worked out. And then we would do, like, a post-campaign analysis with the statistical technique called marketing mix modeling or media mix modeling, where we would look at the relationship between how many impressions we would put in market at a channel level. So how many TV impressions were there on a weekly basis, and then we would compare that with what sales were like. And so from that, we could tell what worked, what didn’t work, and then we could recommend budgets on how much you should increase or decrease your budget and what channel it should go in in order to get the results you wanted. So that’s how things worked before digital. Now the reason that attribution became so popular with digital is because marketing mix modeling was like this it was a project. It wasn’t a product. It was a heavy lift. You had to get 3 years’ worth of weekly data, and it would sometimes take folks 6 to 9 months to put this together. So you can imagine that the turnaround on this was, like, maybe once a year, and the output was like a 100 to 200-page PowerPoint presentation. So think about, as a digital market today, a PowerPoint presentation, that’s not very compelling at all. And if the output says, “This year, 2025, spend 2% more in search,” well, how does that help me as a search marketer? What campaign? What ad group? What keyword? What match type? It doesn’t really tell me. Or if I’m a TV person, it’s like, what creative is working better? What outlets are working better? And what combination is working better? So multi-touch attribution worked at the granular level of that campaign ad group, a creative level, which is where we like to make decisions. It also worked with daily data as well. So at Provalytics, what we’ve done is we have merged the two together. We have taken the statistical techniques that are proven from marketing mix modeling, and we have taken those and merged them with the granularity that we had and multi-touch attribution. So we don’t need 3 years’ worth of data on a weekly basis. We take in daily data, but the output is at a level that a digital marketer absolutely loves. And so what that does is it gives us actionable output. So not only can we look at what happened yesterday, but we can also recommend because we have the forecasting capabilities for marketing and mix to tell you where to spend more the next day. And what’s really great that we do is that since we’ve been in this space since 2008 in the world of attribution, we know the battle that folks are dealing with. And what I mean by that is that every company right now has some measurement. Even if it’s a spreadsheet, it’s a measurement that everyone agrees to. The marketers who are making the decisions, the partners they’re working with, all the way up to finance, budgets are set based on something. And so whenever you bring in new measurement, even though it’s better, you now have to go through this conversion process where you have to get everybody on the same wavelength. It’s kind of like a new religion! And so we really work with our clients hand in hand, working through that process of change management to get everyone on board so that everyone agrees, “Okay, this is” – what I like to say – “it’s less wrong than what we were doing before.” We don’t wanna say “better” because measurement is one of these things that, yes, it’s better, but we just have to come to the agreement that what we’re doing is wrong, and what we wanna do as marketers is be less wrong. And that’s really what it’s about.
Adrian Tennant: Jeff, how does Provalytics handle attribution for what I think you call “no-click” channels like connected TV and podcasts?
Jeff Greenfield: That’s a great question because when we start to think about the world of attribution, most advertisers today, and we’re talking some of the largest in the world, are in what we call last-click attribution, meaning they are measuring based on that last action. Who did that last click? And, yes, that is also, wrong as everything is, and it’s definitely more wrong than what most people are doing, but it’s definitely absolutely wrong when you start to spend money on channels that don’t have clicks associated with them. For example, we’re on a podcast right now. Podcast advertising is one of the number one growing channels for marketers today, and when you’re listening, there’s nothing to click on. You have to remember the ad, kind of like television. And, by the way, the other new no-click digital channel is digitized television, connected TV. And then we’ve got out-of-home billboards, but now we have digital out-of-home. So everything’s being digitized, but these traditional channels, if you will, or premium channels, is what they call them, these don’t have anything to click on. So, how do you measure them? So the way that we measure them is the way that we measured them back in the day, and it comes down to what do you use as your centralized metrics. So if your metric that you’re using, that you’re basing all your measurement on, is clicks, then what that means is is that you should only be advertising in click-based channels because you wanna be able to measure it. But once you move to something that’s no-click, you now need to be able to look at a central metric that you can look at equally across all the channels. And that metric is, as we’ve talked about already, impressions, which is what the job of advertising is, which is to get more eyeballs, to get more people to listen to your ads so you can build awareness. So we do what we did before digital, which is we look at impressions. And you can say, “How can you look at impressions for pay-per-click like Google?” Well, there are impressions in your Google Ads reports. It’s in there. But the problem is that when most people download those reports, they just ignore that column. They never look at it. But impressions are the centralized metric across all forms of advertising, including every single digital click channel. So all you need to do is move further up the funnel if you will, and that’s what we do at Provalytics, is we measure everything based upon impressions that are in market. And the reason we do that, besides the fact that we can include all channels, is because when we think about what are we gonna do with our next dollar, we wanna be able to go and say, spend a $100,000 more on this channel. Well, it’s based upon how many more impressions you can buy, and that’s really what it’s about at the end of the day.
Adrian Tennant: Now that Google has reversed its decision to deprecate third-party cookies in Chrome, how does this impact the value proposition of Provalytics’ privacy-first attribution platform?
Jeff Greenfield: Well, I think what’s interesting is that they were talking about getting rid of cookies for a number of years. And we’re talking about 3rd party cookies, the so-called “tracking cookies.” They’re gone from Firefox. They’re gone from Safari. They’re gone from iOS, so they don’t work on your iPhone, if you will. They’re gone from most places, but they’re still there in Chrome and will be for a certain time frame. But, you know, cookies was one of these things that and, again, I’ll emphasize, 3rd party cookies was one of these things that nobody really sat down and looked at to say “how accurate are these?” This was the backbone of the Internet. Everyone just accepted it, that this was the best that we could do. And what’s happened over the last couple of years is platforms have been preparing for cookies to be completely deprecated is they’ve come up with alternative methodologies for stitching together user IDs across the web. And what we’ve found is that cookies are kinda like the emperor wearing no clothes. Not everyone’s been embarrassed to point their finger to say, “Hey. These things aren’t really as good as we once thought they were.” But they’re there. They’ll still be there. You can still utilize them. But from the proposition standpoint of view of Provalytics, when it comes to privacy-first, privacy in the end is going to prevail because that’s what consumers want. That’s the safest way for folks to move forward. And the vantage point that we have is that by focusing on nonuser data that’s aggregated, the fact that we’re still able to deliver actionable, granular data means that whenever the market shifts again, which it will, it always does, we’ll be in a position that we have future-proofed measurement for our customers. Because the data that we get, which is the number of impressions, the number of clicks, not at a user level but at a daily level, and how much money they spent per day, that’s always going to be available because that doesn’t go against any privacy laws anywhere whatsoever. So we wanted to make certain that when we built something, that whenever the next shift came, for whatever reason, our customers weren’t going to be left with having holes in their measurements. So we built this out in such a way to have it future-proofed.
Adrian Tennant: Let’s take a short break. We’ll be right back after this message.
 | Nick Bennett: Hi. I’m Nick Benett, author of “B2B Influencer Marketing: Work with Creators to Generate Authentic and Effective Marketing.” My book provides a practical guide for business-to-business brands looking to partner with creators who align with their brand values and can help deliver real business results. I break down key concepts and explain how to develop effective Influencer marketing strategies, from identifying the right partners to measuring campaign success. Whether you’re new to B2B influencer marketing or looking to enhance your existing programs, this book offers actionable insights and frameworks to help you succeed. I hope my book helps you transform your B2B marketing strategy with the power of authentic creator partnerships. Thank you! |
Adrian Tennant: Welcome back. I’m talking with Jeff Greenfield, CEO and co-founder of Provalytics. Can you explain how Provalytics works with both direct clients and advertising agencies? And specifically, are there different approaches for each?
Jeff Greenfield: There are different approaches. You know, when you’re working directly with the brand, the brand is focused on, obviously, getting the best results for them. The long-term success of a new measurement platform, as we talked about a little bit earlier, is creating a single source of truth for that brand. Then there are several steps to doing that. First is getting the marketing team to say, “Hey! This new approach is less wrong than what we’ve been doing before. We like it.” So that’s number one is getting the team, however many there are, who are making those decisions. The second is getting the vendors, the partners, the individual agencies on board saying, “Okay. Yeah. You know, even though I’m buying on behalf of them, let’s say, paid search, and even though paid search says that we’re doing a better job than what Provalytics says,” the reality is Google doesn’t see everything, isn’t aware of the TV and the podcast advertising, and I just have to accept this as the new truth. So that’s kind of that second step. The third step is getting this data to live internally at the brand as part of an internal dashboard. All brands now have dashboards, and you want this to live internally. Because once it lives internally, then you can start to get finance and marketing looking at the same data. Finance is not going to log in to an external dashboard to help with their decisions on setting budgets. And so that’s the advantage of working directly with the brand. When we’re working with an agency, that agency is typically one of several vendors. Now sometimes the agency for smaller clients, they may be handling everything. And in that case, that’s wonderful because then we can accomplish all of that together. When the agency is one of, let’s say, five vendors, that job of getting the other vendors on board accepting this is a little more difficult because even though we’re independent, they see us as coming from that agency. So there’s a little bit of extra work that’s involved, but we’ve been able to do it, because even though the agency is bringing us forward, we’re being brought in as an independent source. And we also have some agencies that we work with where we actually white label the product, and they bring it in as their own product. We help them along in the background, help them go through that same process because everyone wins when you have an independent perspective. You know, this is the thing with the statistical platform. There’s no way to juice it. There’s no way to shift it. Someone may say, “Hey. You know, I really feel good about this. These ads that we’re running, I feel great about it.” And it’s like, it doesn’t matter how you feel about it. The numbers are the numbers. And sometimes, folks love us, and sometimes, they absolutely hate us. But that’s the game that we’re in is the numbers are the numbers.
Adrian Tennant: Don’t shoot the messenger.
Jeff Greenfield: That’s right. Please!
Adrian Tennant: Well, your platform emphasizes the importance of measuring incremental impact. Jeff, could you explain what this means and why it matters for marketers?
Jeff Greenfield: Yeah. Incrementality is one of the most important aspects. You know, the idea being is that if you were to run an ad, let’s say, on Facebook, and let’s say you’re a local flower shop, and you have typically, on average, let’s say, 10 customers per day. That’s what you’ve typically averaged for the last couple of months. And you run your first ad on Facebook. Whatever amount of money you spend on it, you spend on it. And all of a sudden, you notice within a couple of days, you’re now averaging 12 customers per day. So you went up from 10 to 12. So now the question is, how many customers per day is Facebook responsible for? Well, it’s pretty straightforward. You know what you did before. You did 10. Facebook has added 2, so you would say that the incremental new customers per day that you’ve received is 2. So you’ve had 2 extra sales per day. And then you can go and figure out the math, you know, what your average sale is, and you can see, okay, what is my return on my ad spend based upon that, knowing that my baseline was 10? And if you’re only advertising on Facebook on just one place, it’s pretty simple to figure that out. So your incrementality is the sales that you wouldn’t have gotten if you hadn’t have done that advertising. Where it gets very confusing is when you’re already advertising. You’re advertising across 7 or 8 different channels. You’re spending lots of money, and now you’re trying to figure out what is my base? What would I have if I didn’t have any sales whatsoever? And then how do you divide this all up? And it gets very confusing because what happens is that once you start advertising on a bunch of places let’s say you’re now a larger advertiser. You don’t have 10 sales per day. You have a 1,000. And let’s say you’re on five different channels, and you look at your books at the end of the day, and you have a 1,000 sales, and you log in to Facebook, and they say that they’re responsible for all 1,000 sales. And you’re like, “Okay. Well, great.” Now you go to Google, and Google says they’re also responsible for 1,000. And you log in to your three other platforms, and you add it all up, and it ends up saying that you actually got 5,000 sales, but you know you only got 1,000. This is that duplication issue that Facebook is not talking to Google, who is not talking to Snap, who is not talking to TikTok, who’s not talking to CTV. Nobody knows what else is going on, and so you have to figure out a way to deduplicate these things. And so at Provalytics, our data is based upon the actual contribution, which is that incremental amount. And the way we do that is we do that through a sophisticated process of what we call simulations, where we take that granular data down to that creative level. And for every single day, we actually remove every single line item, and then we run a simulation, and we ask the machine learning AI and the platform, how many sales would we have had if this wasn’t there? And then it gives us an answer, and then we do a whole validation process and to see how predictive it is and how well it works. And that’s how we get to those incremental numbers. Because if you’re only advertising in one place, simple to figure out. Once you start working across multiple channels, that’s where it gets very difficult.
Adrian Tennant: What advice would you give to marketers who feel overwhelmed by the current complexity of attribution and measurement?
Jeff Greenfield: Well, the first thing I would say is if you’re living in the world of clicks where most marketers are, because Google has done a very good job through Google Analytics of training everyone that everything is about clicks. And, also, as a marketer, we all love success, and we know that a click is closer to whatever our key performance indicators are, our conversions, whether it’s a lead, a download, or a sale. We all have a Google Sheet on our desktop that’s got every single thing that we’ve spent money on. It’s lined up for every single day of the year, and it’s got our cost per click, our cost per lead, our cost per sale. It’s all there. And what I would say is that if you’re overwhelmed by that, meaning that you’re finding that it’s not doing a good job of showing you where you should spend your next dollar, you need to take a step back because you’re too deep into the forest to see what’s really going on. So you need to broaden your perspective. And the way to do that, first, is to use that same Google Sheet. And what I would say is go back the last 12 months, and this is gonna be like a project. This is like a long weekend, but you have to add impressions to your Google Sheet. Maybe it’s there already. Maybe it is. And if it is, great. But you need to go back to the last year on a daily basis and add in impressions. And what you wanna do is create graphs of daily impressions, total daily impressions across wherever you’re spending your dollars, and then correlate that with your clicks and correlate that with your sales or your leads. Because remember, the thing you can buy today, you can’t buy clicks. Clicks are the impact from what you’re doing. You can buy impressions. And what you wanna look for is the relationship between impressions that you buy today and that impact typically several days or maybe a week or so later. And you’re gonna find where things start to jump up. You’ll find where you all of a sudden, you’ll see clicks go up. And if you look into the source, you’ll find that the source is typically organic or brand. They just showed up. So now you wanna go back and see when impressions were increased maybe a week before, and now you wanna dig into those days to find out what campaigns, what creative was running. And whatever that was, you wanna do that again. And when you get sophisticated enough with it, now you can start to look at it on a seasonality basis where you can start to say, “Okay. This is what I ran last year, Black Friday. These are the campaigns that work or these types of campaigns work.” So that would be the first thing is to is to take a step back away from clicks. You kinda wanna say “no” to clicks. That’s really the first thing is understanding that, really, we’re buying eyeballs. We’re buying impressions. And because when you focus on clicks, you’re so far down the funnel, you’re chasing your tail. You wanna live in that world that’s more up funnel. That would be my probably my number one recommendation across the board.
Adrian Tennant: Great insights, Jeff. Thank you. If listeners would like to learn more about you or your work at Provalytics, what’s the best way to do so?
Jeff Greenfield: Well, I’d say the best thing is to go to the website, Provalytics.com. You can check it out. There’s a lot of articles to read up on, especially all about, you know, these challenges that are around privacy. There’s actually a new research study that showed that the impacts of these privacy changes, especially for marketers that are hyper targeted, really lower funnel, was about a 37 percent decrease in revenue. So pretty significant. Other articles as well about all of these deduplication and the impacts of retail media and all about single source of truth. But also in there, you can download our 2025 Playbook that talks about all these challenges. We also have on there a free attribution certification. You can take a free course online and take a quiz. You pass it. You get a great certificate for LinkedIn. And then we also have an on-demand demo as well too. So everything there is available at your fingertips to check us out and learn more.
Adrian Tennant: Jeff, thank you very much for being our guest this week on In Clear Focus.
Jeff Greenfield: Thank you, Adrian. It’s been a pleasure.
Adrian Tennant: Thanks again to my guest this week, Jeff Greenfield, CEO and co-founder of Provalytics. As always, you’ll find a complete transcript of our conversation with timestamps and links to the resources we discussed on the IN CLEAR FOCUS page at bigeyeagency.com. Just select ‘Insights’ from the menu. Thank you for listening to IN CLEAR FOCUS, produced by Bigeye. I’ve been your host, Adrian Tennant. Until next week, goodbye.
TIMESTAMPS
00:00: Introduction to IN CLEAR FOCUS
00:41: Challenges in Advertising Effectiveness and Privacy
01:15: Guest Introduction: Jeff Greenfield
02:08: Inspiration Behind Provalytics
03:55: Limitations of Traditional Attribution Models
07:20: Provalytics’ Unique Methodology
11:38: Attribution for No-Click Channels
14:33: Impact of Google’s Cookie Decision on Provalytics
18:18: Working with Direct Clients vs. Advertising Agencies
21:48: Understanding Incremental Impact
25:17: Advice for Overwhelmed Marketers
28:27: Learning More About Provalytics
29:27: Conclusion and Thanks
