RevelOne’s Spotlight Series regularly features insights from top experts in our Interim Expert Network. We cover a broad range of topics at the intersection of marketing, growth, and talent. If you’re interested in exploring these topics further and engaging with one of our 250+ executive or mid-level experts, please contact our team at experts@revel-one.com.

Marketing measurement has always been challenging. Since the dawn of time, marketers have tirelessly sought methodologies to measure and maximize returns on their marketing budgets. Traditional tools like Media Mix Models (MMM), pioneered over 60 years ago by companies like Procter & Gamble, and Multi-Touch Attribution models (MTA) from the 2000s, have seen their effectiveness fluctuate due to technological and privacy changes. Despite these shifts, marketers continue to hunt for a measurement edge to maximize ROI and elevate their brands.

So what’s next for marketing measurement? What’s coming in 2025 and beyond? 

Long Live Cookies? And APIs.

Cookies have become integrated into nearly all marketing channels in some way over the last 25 years. They store information in customer browsers and pass information back to marketing tools that enable tracking, optimization, and more. Since 2020, marketers have been preparing for “the death of the cookie” as browsers have vowed to be more and more privacy and consumer-centric. There are lots of technical details around first-party cookies, third-party cookies, privacy controls, etc. that I won’t get into here. The punchline is that while Google was on track to follow Safari/Apple and kill third-party cookies entirely, they have since backed off that stance and canceled their cookie crackdown. In parallel, Google Chrome has maintained its position as the dominant browser by market share, so the end of cookies does not appear to be as near as it once was.

That said, while everyone was preparing for cookies to go away, advertising platforms like Google and Meta invested heavily in backend conversion APIs that allow advertisers to share data with the platforms without using cookies at all. The advertising platforms also got much more creative, and at times sneaky, in how and where they use modeled data to plug gaps left behind by cookies when representing results to advertisers, particularly when using longer click-and-view attribution windows. Marketing tool companies have also gotten very creative in using first-party cookies for internal analytics tools like MTAs, etc. More on walled gardens and MTAs below.

Privacy Banners on Privacy Banners

Once new and scary to both consumers and advertisers alike, everyone is now swimming in a sea of privacy banners that are largely never read. Consumers have grown numb to the privacy banners driven by CCPA and GDPR and marketers have grown savvy to using conversion rate optimization strategies to elevate opt-in rates. Advertisers need to follow the rules, but they also need to not punch themselves in the face by accident with their privacy banner UI, because their privacy banner opt-in rates drive the amount of data they can share with their advertising platforms which drives their campaign success which drives their marketing spend ROI. 

The Walled Gardens Are Growing

As mentioned above, the mega advertising platforms Google, Meta, and Bing have been heavily investing in conversion APIs that allow for backend communications between advertising platforms and advertisers. This has reduced their reliance on cookies (which have been under attack, see above) and allows them to enrich their data and target more new customers without privacy concerns (except for that pesky privacy banner opt-in, see above). These platforms don’t need anyone else to tell them who they showed an ad, or who clicked on an ad, and if advertisers tell them via API call who ended up purchasing, these platforms have everything they need to keep driving more new customers to advertiser websites and printing money for their shareholders.

This is why advertising dollars continue to flow to these platforms and their flagship, macro, “AI-enabled” products like Google’s Performance Max, Meta’s Advantage+ Shopping Campaigns, and Microsoft/Bing’s…Performance Max. Yes, they named it that too.

The challenge marketers will face going forward is to determine 1) how to feed these beasts with enough creative to maximize ROI and 2) how much to trust these tools. They will need to test and balance budgets across these campaign types and platforms as they are changing very, very fast.

Platform Reporting: Stable Apples + Oranges

A constant within the marketing world is the general stability and usefulness of “platform reporting,” aka the things that the advertising platforms tell advertisers are true. The strength of the Meta and Google walled gardens continues to grow (see above) and the ad-serving algorithms used by the large advertising platforms are some of the largest deployments of machine learning and AI in the history of humankind.

The only problem is that, while their reporting is generally stable and consistent, advertisers can only safely use today’s results to compare them on a relative basis to yesterday’s results or last month’s results on the same platform (e.g. is today higher or lower than yesterday). You can’t trust the actual number itself (e.g. a platform-reported $38 CPA is very likely not actually a $38 CPA), and you can’t compare numbers from one platform to the results from another unless you have run incrementality tests to create discount rates for each before doing so (see below).

So platform reporting is an essential tool for managing a channel relative to itself, but it’s comparing apples to oranges if you try to use it to compare one channel to another.

MTAs: Living to Fight Another Day

Once a marketer's dream, the drawbacks of MTAs in a post-consumer privacy-based world have been widely documented. Since browser-based privacy changes started heating up in 2017, MTA efficacy has been on a steady decline. I have spent years of my life and millions of dollars on tools and teams to build internal MTAs, and I am very sure I’ll never build one internally again. There are, however, several 3rd party tools available like RockerBox, Northbeam, and TripleWhale. At their peak, these tools promised advertisers cross-site view-through (“impression level”) data that would in theory tell a clear story of the location and frequency of ad impressions (without clicks) that later drove on-site purchases.

More recently, post-browser privacy revolution, these tools can only deliver a picture of where customers came from just prior to arriving on your site (via UTM), what the customer does while on-site (which is available via GA4 and other tools), and attempt to link users cross-session and tell a complete picture of the on-site customer journey. These tools rely on their cookies being placed as first-party cookies, while under attack and auto-deleted by Apple/Safari, Google has canceled their crackdown. This could mean that MTA efficacy has stabilized, at least temporarily, and paired with the generally disliked rollout of GA4, third-party MTA tools could see a resurgence. However, they are still pretty expensive, they are limited to click-based activity, they will use modeled data to fill Safari-sized holes in their data sets, and they require maniacal attention to detail to implement and maintain across a marketing team to be somewhat accurate. 

Similar to the app advertising platforms, the companies that make these tools and the advertisers who use them are living in fear of the next browser privacy change that might knock their measurement approach out of commission. Some marketers love these tools. I am not one of them. But given Google backing off the death to cookies bandwagon, it looks like MTAs will live to fight another day.

Fast & Lean MMMs Continue to Emerge

MMMs are magical tools for the likes of the companies that invented them…mega CPG brands with deep historical sales and marketing data and massive distribution. Brands like Gatorade, Clorox, Cheerios, Pepsi, and Coke can get massive value out of MMMs and have done so for decades.

The MMM market has historically been dominated by players like Neustar (now TransUnion) and Analytic Partners which generally require annual contracts, cost many hundreds of thousands of dollars, and take many months to deploy. These providers are generally too big and heavy for fast-moving growth-stage companies with lean teams. There is, however, a cohort of leaner, faster MMM providers like Recast, Prescient, and Paramark which cost $5k - $20k/month, deploy pretty quickly, and generally harness open-source models and SaaS playbooks & best practices. These providers make spinning up an MMM faster and cheaper…but they are still MMMs.

Again they are great for generally stable businesses, but a major problem remains for the fast-growing cohort of companies: if you are relying on the organic fluctuations of your business and advertising spend to fuel an MMM, the result will be something that appears to be highly precise but might be a massive hallucination. These models rely on large historical datasets (12-24 months) and can easily confuse cause and effect, e.g. are the results from a Black Friday sales bump due to the uptick in spending on Meta during that period or the increased consumer proclivity to buy during that window? Additionally, the historicals of growth-stage businesses are often just too volatile and the advertising changes are not aggressive enough for MMMs to be useful by themselves. An MMM fueled by aggressive incrementality testing is another story, though. More on this below.

The Rise of Incrementality Testing

The biggest trend in the last ~18 months in the marketing world that will certainly continue in 2025 and beyond is the explosion of incrementality testing. Platforms like Haus and Measured have created tools for advertisers that enable them to run lots and lots of geographic holdout tests in their campaigns. When these tests are run properly, they tell marketers very quickly what results their spend is delivering to their business without needing any support from analytics or data science teams. These tests don’t rely on cookies and are very privacy-friendly. They give marketers rare slivers of truth in a sea of apples and oranges (see above). Well-structured and analyzed incrementality tests tell marketers, at a moment in time, what results are actually coming from a campaign or platform. They do this by simply turning the tested marketing effort off in a variety of locations and seeing what happens. If you do this and no KPIs change, whatever you were testing wasn’t doing very much work for you.

Some challenges with incrementality testing are that you can only run a couple of tests at any given time, and keeping track of test results gets pretty complicated. More on this below. Disclaimer: I am a Haus investor and advisor.

On the Horizon: Incrementality-Driven MMMs

A new class of tools that is on the cusp of being brought to market and widely adopted is the incrementality-driven MMM (“iMMM”? You heard it here first?).

First and foremost, the most valuable MMM is one with a massive amount of aggressive incrementality tests in the historical data that the model is ingesting. I am sure that I will never use an MMM again without a rich history of incrementality testing, and ongoing incrementality testing run in parallel to the MMM.

Additionally, while the existing MMM tools can ingest incrementality tests into their datasets, advertisers need these tools to put the results of the incrementality tests as the hero within the MMM readouts. The incrementality tests are the strongest, and sometimes the only, “facts” in the data while everything else is based on modeled correlations. An MMM used in this manner can be a very helpful repository for housing and visualizing incrementality test results and surrounding modeled data over time as performance data and test results build up within a company’s historical data.

I am sure that there will be several providers who bring this to market in 2025 and foresee them being the standard operating for marketers very quickly. 

Conclusion

As we look toward 2025 and beyond, the landscape of marketing measurement is rapidly evolving. The challenges posed by privacy regulations, the uncertain future of cookies, and the expanding walled gardens of major advertising platforms necessitate a shift in how marketers approach measurement. Incrementality testing has emerged as a vital tool, offering rare snapshots of truth in an environment filled with inconsistent data. The likely rise of incrementality-driven MMMs promises to combine the strengths of traditional models with the actionable insights of rigorous testing. By focusing on accurate measurement and being adaptable to technological and regulatory changes, Marketers can maximize the ROI of their marketing dollars and elevate their brands to new heights. The future holds many uncertainties, but with the right tools and strategies, marketers can navigate the complexities and thrive.

About the Author

Rob Webb is a marketing consultant passionate about building scalable growth systems, particularly for growth-stage consumer & prosumer software businesses. He’s helped multiple businesses grow 500x, managed $100M+ media budgets across a variety of channels, and hired and led hundreds of amazing people. Teams he’s managed have driven $1B+ in gross sales. He has built and scaled organizations across Growth, Marketing, Product, Engineering, and Design. You can find him on LinkedIn or check out Growthish where he shares marketing & growth insights.

About RevelOne
RevelOne is a leading go-to-market advisory and recruiting firm. We help hundreds of VC/PE-backed companies each year leverage the right resources to achieve more profitable growth. We do 250+ retained searches a year in Marketing and Sales roles from C-level on down for some of the most recognized names in tech. In addition to our Search Practice, our Interim Expert Network includes 250+ vetted expert contractors – executive-level leaders and head-of/director-level functional experts – available for interim or fractional engagements. For help in any of these areas, contact us.