Smarter Marketing Measurement: Your Competitive Edge for Revenue Growth

Revenue Boost: A Marketing Podcast

Smarter Marketing Measurement: Your Competitive Edge for Revenue Growth

"The big ‘aha’ moment for most marketers comes when they cut something they thought was working, wait 30 or 60 days, and see that sales remain exactly the same. That realization—that they were spending money on something with zero impact—can be both eye-opening and unsettling." – That’s a quote from Jeff Greenfield, CEO of Provalytics and a sneak peak at today’s episode.

Today, we’re diving deep into one of the most critical challenges in modern marketing: measurement.

How do you know if your marketing dollars are truly driving revenue? Are you making data-driven decisions—or just guessing?

In today’s episode Smarter Marketing Measurement – Your Competitive Edge for Revenue Growth, I’m joined by Jeff Greenfield, CEO and co-founder of Provalytics.

In this episode, Jeff and I discuss:
✔️ Why most marketing measurement is broken—and how to fix it
✔️ The impact of upper-funnel branding and how to prove its ROI
✔️ How AI and machine learning are transforming attribution
✔️ How to align marketing and finance using a single source of truth

Be sure to listen to the end where Jeff shares actionable steps to improve your measurement strategy today!

Are you ready to take your marketing strategy to the next level! Let’s go!

Kerry Curran (00:01.144)
So welcome, Jeff. Please introduce yourself and share a bit about your background and expertise.

Jeff Greenfield (00:07.758)
I'm Jeff Greenfield. I am the co-founder and CEO of Provalytics, an AI-driven attribution platform. Since 2008, I've been in this space to answer that old question from John Wanamaker: "Half the money I spend on advertising is wasted. The only problem is, I don't know which half." Since 2008, I've been helping marketers and brands determine which half is wasted and how to redeploy those existing funds to increase their return on investment.

Kerry Curran (00:45.678)
Excellent. We're excited to hear everything you know about analytics, data, and attribution. So tell us—when your prospects or brands call you for the first time, what are some of the business challenges they face that make them realize they need your help?

Jeff Greenfield (01:06.432)
I'd say one of the top challenges is the concept of overcounting. Most marketers operate in more than one channel—typically five or six or more. Each channel has its own way of counting. The best way to think about it is that when you're advertising on Meta, they don't know that you're also on TV. They don’t know that you're on Google. Criteo doesn’t know that you're on Amazon.

Kerry Curran (01:17.742)
Mm-hmm.

Jeff Greenfield (01:33.294)
If you have a thousand orders in a day and you're working across five partners, when you add up all their data, it may actually tell you that you had 5,000 orders. So, overcounting is a major issue. Marketers often ask, “How do I figure out all this math?”

Another big challenge is knowing that, as a marketer, you hear anecdotally that channels like connected television (CTV) and podcast advertising work for brands similar to yours. Yet, when you try them, you don’t see results, and you wonder, “What’s the magic? How is it working for them, but not for me?” You don’t see the numbers going up, and you’re trying to figure out why.

Finally, one of the biggest challenges is the constant tension between marketing and finance. Finance teams are heavy on math, and they often talk about marketers under their breath, saying we don’t understand how math works. Meanwhile, marketers think finance doesn’t understand how marketing works. This disconnect is critical because finance controls the budget. If you want more budget, you have to speak their language. Those tend to be the biggest issues.

Kerry Curran (02:57.442)
Yeah, it’s definitely a challenge. I’m nodding and laughing because we all know that CFOs are the hardest to convince of marketing’s value—especially for upper-funnel initiatives. I believe in the rising tide lifting all ships when it comes to marketing, but you're right. If you can’t align investment at the channel level or prove overall impact, it becomes much harder to justify.

You're helping clients identify the sources of traffic and revenue. How do you solve for this? How are you helping them build out a single source of truth?

Jeff Greenfield (03:47.534)
That’s the key—figuring it out. One issue within organizations is that, going back to my earlier example, if a company has five agencies, each agency is using its own methodology. They rely on platform metrics, their own internal metrics, and the marketing team’s metrics. So, if each agency uses three different methods, and then finance has its own, that means the company has 15 or 16 different sources of truth.

Kerry Curran (03:56.077)
Yeah.

Jeff Greenfield (04:17.358)
This becomes a huge issue. We solve it using a statistical, machine-learning, AI-driven approach.

Back in 2008, when I built C3 Metrics, we could collect 100% of the data—all website data, third-party data, and impression data. We could track an end-to-end trail, with date and timestamp, whenever someone converted.

Then, privacy regulations changed everything. Facebook, YouTube, iOS—they all said, “You can’t have impression data anymore.” Now, there are more data gaps than available data. So, we had to ask, “How do we fill these gaps?” That’s where statistics, machine learning, and AI come in.

The great thing is that we no longer need user-level first-party data. AI has become so advanced that all we need is daily aggregated marketing data from platforms and separate conversion data. We can link them together.

This allows us to connect digital and traditional channels to digital KPIs—whether on a company’s website, Amazon, or other marketplaces. We can even connect marketing impressions to individual scripts written each day.

We’re now in a privacy-centric world. We’re not tracking at the user level, but because of stronger math and faster computers, we can achieve insights that were previously impossible.

Kerry Curran (06:26.286)
That’s incredible. You bring up so many examples of how difficult it is to track conversions and touchpoints, and to demonstrate a channel’s benefit and halo effect. So, break it down—how do you help brands, as you’ve said before, measure the unmeasurable?

Jeff Greenfield (06:54.636)
It’s really about understanding how different channels impact one another.

I was talking earlier today with a TV agency for one of our clients, and I reminded them how much things have shifted. Years ago, direct response TV ads would say, “This product is only available through this 800 number—limited supplies!” People would stop what they were doing and call.

Now, consumers know they have options. They can visit the website, check Amazon, or walk into Walmart. The challenge is understanding how media in one channel influences conversions in another.

For example, a brand might run TV ads directing viewers to their website, but most people actually go to Amazon instead.

The biggest way we help brands is by taking data through a step-by-step process. First, we align the internal marketing team, because this is a new way of looking at data. Insights may feel uncomfortable at first—because they challenge assumptions.

Then, we work with agencies. Brands hire search agencies to follow Google's guidance. But when you're advertising in 20 different places, you need to shift focus. Convincing agencies to adopt a new methodology takes time.

Once everyone is aligned, we integrate the data into internal dashboards. This is where things get exciting—the CMO or VP of Marketing can go to finance and say, “Look at the dashboard. The numbers add up. Overcounting is fixed. The halo effect is accounted for.”

And that’s how you, as a marketer, get a bigger budget to grow the brand.

Kerry Curran (10:34.094)
That’s so smart. Change management is one of the hardest parts of implementing new strategies, especially in marketing. How do you convince marketers, agencies, and CFOs to trust your data?

Jeff Greenfield (11:04.142)
Great question. Unlike old attribution models, which weren’t incremental, our data is fully incremental.

To build trust, we back-test all data. We validate models using a method called K-fold testing. Instead of withholding a full month of data, we train the model with a month’s data but hold back different portions across multiple tests. This lets us validate the model while keeping recent data.

But the real proof comes when marketers act on our insights. The moment they cut a campaign they thought was working, and 30–60 days later, sales remain unchanged—that’s the aha moment.

Here’s the transcript with only grammar corrections, ensuring clarity while maintaining the original tone and intent:

Jeff Greenfield (11:04.142)
Well, that's a great question. Unlike the days of attribution—where the big complaint was that it was never incremental—our data is entirely based on incrementality. Everything we do is incremental. One of the ways we convince people of this is by back-testing all the data to validate the models.

Kerry Curran (11:05.688)
You.

Kerry Curran (11:11.054)
Mm-hmm.

Jeff Greenfield (11:33.986)
What I mean by that is, if you go back to the old days of marketing mix modeling, you would use about three years’ worth of data. The last month of data would be held back, and then you would ask the model to predict the revenue for that most recent month. You could then compare the prediction with actual revenue to assess how well the model worked, which helped build confidence in the results. However, those results were based on a three-year period and were primarily used for planning the next year.

Kerry Curran (12:03.832)
Mm-hmm.

Jeff Greenfield (12:04.158)
But marketers today are most interested in what happened in the last month or even the last week. We don’t want to hold back that data. There’s been a lot of work in machine learning and AI to validate models while still providing the most recent insights.

A technique called K-fold testing was developed for this purpose. It involves training the model using a month’s worth of data while holding back a portion of the days. For example, we might hold back the revenue, leads, or add-to-cart data for 20% of the days and ask the model to predict those values. Then we repeat the process, holding back a different 20%, and do this five times. By the end, we’ve held back 100% of the data at different points, allowing us to fully validate the model’s accuracy.

Even though we can show a chart demonstrating that the model predicts outcomes with, say, 93% accuracy, nothing beats real-world testing. If the model suggests that a campaign isn’t producing the expected results and recommends cutting it by 50%, we can test that recommendation by actually reducing the spend and observing what happens.

Kerry Curran (13:11.758)
Mm-hmm.

Jeff Greenfield (13:26.816)
The big “aha” moment for most marketers comes when they cut something they thought was working, wait 30 or 60 days, and see that sales remain exactly the same. That realization—that they were spending money on something with zero impact—can be both eye-opening and unsettling.

The truth is, if you’re not using analytics at this scale, much of what you’re doing may have little to no impact. That’s the first thing to recognize. But it’s also important to understand that you didn’t know any better before. The focus should always be on improving and moving forward. The best way to build trust in the model is to first show how well it predicts outcomes, and then implement the recommendations to see the results in action.

Kerry Curran (14:18.946)
Yeah, that’s so smart. I love how you’re able to prove the impact of your model and show how it works. It’s a challenge to truly understand what’s working in marketing.

One of the things we’ve discussed before is the impact of branding initiatives and how different channels influence the bottom line. How are you uncovering those insights for marketers, especially in channels where there’s less of a direct click path?

Jeff Greenfield (14:54.636)
First off, I think many marketers who have only worked in digital marketing have a warped view of how marketing actually functions. I blame Google Analytics for this because it’s entirely click-based.

Many marketers believe that we invest dollars to buy clicks, and clicks lead to sales—that’s how marketing works. But that’s actually not how marketing works.

The click is the last thing that happens. What we do as marketers is invest dollars to buy eyeballs, which we call impressions. We buy impressions to capture attention. The job of those impressions is to build awareness, and when awareness is built up enough, people will take action—whether that’s visiting a store or, in today’s world, clicking on a website.

For most brands today, their "store" is online, meaning clicks lead to conversions. But the hyper-focus on clicks—driven by Google, Meta, and other digital platforms—has pushed marketing dollars toward the lower funnel, at the expense of brand-building efforts.

Kerry Curran (16:22.126)
Mm-hmm.

Jeff Greenfield (16:22.242)
And that’s a problem because the lower funnel is the most competitive space. It’s a bidding war. If you spend the same budget this year as last year on a particular channel, you’ll likely get fewer clicks because the cost per click keeps rising. Just look at Meta’s and Google’s earnings reports—they keep increasing because advertisers are stuck in this lower-funnel trap.

Kerry Curran (16:42.232)
[Laughs] Mm-hmm.

Jeff Greenfield (16:50.102)
Larger brands are catching on. They’re moving up the funnel. Investing in upper-funnel marketing is the gift that keeps on giving because your funnel stays full. It delivers returns at twice the rate of lower-funnel tactics.

We measure this by focusing on how marketing actually works—tracking impressions rather than just clicks. Our impression-centric model allows us to compare different channels—linear TV, CTV, direct mail, paid social, and more—on an apples-to-apples basis.

Branding efforts often take longer to show impact, but we track multiple KPIs, not just revenue. We incorporate leading indicators, such as website traffic, call center volume, and other engagement metrics, to capture branding’s long-term effects.

Branding has always been critical, but now it's finally being recognized as the key to long-term growth.

Kerry Curran (18:40.856)
Mm-hmm.

Kerry Curran (18:44.812)
Yes, I completely agree. I’ve seen this play out across multiple brands. There’s been such a race to the bottom—just focusing on immediate conversions without building awareness or customer relationships.

I hope more marketers and CFOs are listening to this. Branding is the growth lane, and making smarter investments across channels is what truly drives long-term revenue growth.

Jeff Greenfield (19:18.614)
A thousand percent. Most marketing today is focused on offers, benefits, and limited-time deals. But brands that differentiate themselves with emotional messaging—connecting with their audience on a deeper level—win in the long run.

Marketers obsessed with lower-funnel performance often forget that consumers form emotional connections with brands, and those connections drive purchasing decisions. The complexity of digital marketing has caused many to lose sight of fundamental marketing principles.

Kerry Curran (20:14.53)
Yes, I agree! That’s exactly why we’re here—to help educate people on marketing strategies and foundations.

One key thing you’ve pointed out is that you can tie ad creative and messaging to performance. Going back to that emotional connection, how are you testing and measuring it?

Jeff Greenfield (20:43.694)
Absolutely. We incorporate ad creative as a dimension in our model. This works well for video, TV, and radio advertising. Even for search and social, brands can extract key ad attributes and integrate them into their marketing hierarchy.

Once you categorize creative elements, you can analyze which components are driving higher sales or leading indicators. This data informs future creative strategies, ensuring continuous improvement. That’s what makes this so exciting.

Kerry Curran (21:32.62)
I love that. Insights like these help brands become smarter, more efficient, and more effective with their marketing investments.

Jeff, thank you so much for your expertise. For marketers who want to improve their measurement approach, where should they start?

Here’s your transcript with only grammar corrections, ensuring clarity while maintaining the original tone and intent:

Jeff Greenfield (20:43.694)
Absolutely, because that becomes one of the dimensions of the model. What’s really exciting is that when brands actually take the time, they can easily analyze this for video advertising, TV, or radio. However, it becomes a bit more challenging when dealing with search and social ads.

That said, it doesn’t take much effort for marketers to go through their ads, identify key attributes, and integrate them into their marketing hierarchy. Once they do that, they can start seeing which ad components drive more sales or leading indicators. This, in turn, helps shape future creative decisions. That’s what makes this so exciting.

Kerry Curran (21:32.62)
Yeah, I love that. I love the level of insight, and anything that helps brands become smarter, more effective, and more efficient with their investments is incredibly valuable.

Jeff, I appreciate all of your insights. For the people listening who are thinking, I need to get smarter about my measurement, what are some foundational steps they should take to get ready?

Jeff Greenfield (21:59.128)
Well, the first thing I’d say is that most marketers running campaigns typically have a Google Sheet sitting on their desktop or laptop. It tracks daily spend, clicks, cost per click, and cost per sale. But what’s often missing is the impression number.

And chances are, when they downloaded the reports to build this sheet, impressions were included in the data—they just ignored the column.

Kerry Curran (22:09.422)
You.

Jeff Greenfield (22:28.096)
So, I’d recommend repulling all of that data for the last 12 months on a daily basis. Add an impressions column right after the date, then start graphing your daily impression volume alongside your daily clicks and daily sales. Look for relationships in the data.

This is a DIY approach to what we do at Provalytics.

Kerry Curran (22:40.204)
You.

Jeff Greenfield (22:54.302)
As you analyze these relationships, look for a time delay between impressions rising and an increase in clicks and conversions. When you identify days where impressions spiked and led to a later uptick in sales, dig into those specific days. What did you do differently? That’s the type of activity you want to do more of.

This is the first step in preparing for a paradigm shift—understanding that we buy impressions, and that’s where marketing analysis should begin.

Kerry Curran (23:17.166)
I'm sorry.

Jeff Greenfield (23:22.964)
The second step is education. At Provalytics, we’ve put a lot of thought into this, especially with all the privacy changes and how the industry is evolving.

We created an Attribution Certification Course that covers the past, present, and what we see as the future of attribution. Because marketing will continue to change, the best way to prepare is by strengthening your foundational knowledge.

The course is completely free. It takes about an hour and a half to complete, and there’s a quiz at the end. If you pass, you get a certification you can showcase on LinkedIn. It’s a great resource to deepen your understanding of how we got to where we are today.

Kerry Curran (24:11.278)
Excellent, Jeff! This is incredibly valuable. I’m definitely going to check out the Attribution Certification myself.

Tell us—how can people find you? Where can they get in touch with you and learn more about Provalytics?

Jeff Greenfield (24:25.634)
People can always find me on LinkedIn if they want to connect. They can also visit the Provalytics website, where we offer an on-demand demo.

We also host regular live demos, where we walk through the platform in detail and explain exactly how the modeling works. If anyone is interested, they can sign up, watch the demo, and schedule a time to chat with us.

I’m always happy to speak with marketers—or anyone interested in this space. I know that, to most marketers, this is just math, but to me, it’s kind of sexy.

Kerry Curran (25:07.382)
Awesome! Well, I’m glad we’re making data and attribution sexy again, right, Jeff?

Thank you so much for sharing your expertise, insights, and free resources with the audience. This has been fantastic.

Jeff Greenfield (25:13.506)
That’s right.

Jeff Greenfield (25:27.064)
My pleasure, Kerry. Thank you so much for having me.

🎙️ Thanks for tuning in! Struggling with flat or slowing revenue growth? You’re not alone. That’s why Revenue Boost: A Marketing Podcast brings you expert insights, actionable strategies, and real-world success stories to help you scale faster. 🚀

🔥 If you’re serious about growth, hit follow, subscribe, and drop a ⭐⭐⭐⭐⭐ rating—it helps us keep the game-changing content coming!

New episodes drop regularly—don’t miss out! 🎧

Follow us on Apple

Follow us on Spotify

Watch on YouTube

Read Revenue Boost: A Marketing Podcast transcripts.

To learn more about Kerry Curran and the RBMA: Revenue Based Marketing Advisors, go to www.revenuebasedmarketing.com and be sure to follow us on Kerry's LinkedIn Profile and The RBMA: Revenue Based Marketing Advisors Profile.

RBMA specializes in business transformation to drive revenue growth. We lead companies move from an ABM strategy to a company wide Go-to-Market program that sets you up for sustainable, year-over-year revenue growth.

If you're in the market for a Fractional Chief Marketing Officer or Fractional Chief Revenue Officer be sure to reach out to Kerry. Kerry is also available for speaking, panel moderation, and other professional presentation services. For services and contact information check out the RBMA: Revenue Based Marketing Advisors website here.