00:05:50 – Welcome: Why Measurement for 2027

Chris: Welcome, and thank you for joining. Today's roundtable is all about measurement in 2027, and we have three stellar panelists. I'll be facilitating. My name is Chris Kervinen, and I lead Sellforte's go-to-market in the US.

Before introducing the panel, let's take one step back and discuss why we're here. There's an elephant in the room we're carrying into 2027. Marketing measurement gets a lot of attention, and there are newer technologies to help us with the tough questions as we enter the age of AI. But there's often a big gap between how marketing thinks it drives the business and how finance sees it. Google's effectiveness equation research shows that 56% of marketers feel rising pressure from finance to prove that marketing drives impact, yet only 15% give their CFO a top score as a partner. So there's a lot of pressure, and a mismatch between those two worlds.

Here's how the next 54 minutes will run. We'll meet the panel, then go through six questions, from what happened in 2026 to what's coming in 2027, and wrap up with Q&A. Please drop your questions in the Q&A section. The session is recorded, and we'll share it afterwards.

00:10:47 – Meet the Panel

Danielle: I'm Danielle Mathews, Senior Director of Integrated Marketing at Pura, a US smart fragrance brand that's growing on the DTC side, into retail, and is looking at international expansion in the next couple of years. Before Pura, I was agency side, where I ran digital and stood up media divisions at various agencies and publishers.

Thaynna: Thank you for having me. I work in corporate media at Grupo Azzas 2154, a kind of in-house media consultancy for a portfolio of fashion brands including FARM, Animale, Maria Filó, Reserva, and Cris Barros, some of which also operate in the US, Europe, and other countries. My role spans paid media strategy, MMM, and attribution, and we focus on translating measurement into decisions leadership trusts. We run MMM with Sellforte to layer on counterfactual analysis, and we maintain solid GA4 and GTM governance.

Kate: Hi everyone. I'm Kate, and I lead the campaign analytics team at AMP, a full-funnel media agency. I work across a portfolio of national brands in different verticals and at different stages of measurement maturity. It all comes down to one question: what's the real incremental value paid media drives for our clients' businesses, and how can we use data to make it work harder and drive growth?

Chris: We have a growing DTC brand, a house of fashion brands from Latin America, and the agency angle. It will be a good discussion.

00:13:08 – 2026 Recap: Pura's Year of Retail Growth

Chris: It feels like one of those years where you say "what a year," and it's still only September. Before discussing 2027, let's recap 2026. Danielle, what were your main learnings, and what does your setup look like going into 2027?

Danielle: 2026 was truly a year of growth for Pura. We were a DTC-only company. In 2025, we tested some smaller specialty retailers and launched on Amazon, but 2026 is when retail really bloomed. We're in every Target store, in the wellness section of most Ultas, and in Nordstrom, Best Buy, Costco, and more.

Coming from a DTC world, where that's all finance, leadership, and marketing knew, it pushed us toward integrated marketing and understanding what media has to do. Like Kate said, we had to prove incremental value, because our CFO won't want us to invest in anything we can't prove.

We kept our performance budget pretty much flat to protect .com, and we made major investments in brand and product marketing to build the mid and upper funnel. Then, as the brand grew, funding our digital shelf on Google and with our retailers became more expensive. That leads into the fun conversations about CAC and new customers for .com, the never-ending conversation with finance. For us, it's about protecting our core. Meta and Google work extremely hard for us.

We also looked at our measurement setup. We have a standard MTA tool, but we wanted an MMM as well to have the full-funnel conversation. That let us work with finance on why we needed to protect the brand budget instead of cutting it, and on the impact that has down the funnel.

Chris: When did it become important to link the performance channels to retail sales?

Danielle: We saw the signals in the platforms earlier than finance and the wider organization did. The platforms tell you when they're being starved in the funnel. It was very evident we were farming the same people over and over again, and we needed to bring net-new people to Pura to keep growing.

Investing in an MMM showed us which channels we were over- and under-investing in. You don't take it as the Bible, but it let us have the hard conversations with finance and say: we're under-investing in brand. And the mid funnel always gets cut the most. I don't understand why, because that's what connects the lower and upper funnel for us.

00:18:34 – 2026 Recap: Grupo Azzas 2154 and the Limits of Last-Click

Chris: Thaynna, you've also had a full year. How would you characterize 2026?

Thaynna: 2026 was a challenging year for the business. Media keeps growing, but our challenge now is making each channel more profitable, just like Danielle said. We have strong confidence in our data and tools like GA4, backed by data tracking and constant improvements to data quality. But we all know that last-click is reliable for what it measures, and it's still structurally blind to the top, middle, and offline channels that are essential to brand health. Positive ROI only shows up for the bottom of the funnel. MMM corrects that myopia and shows the real impact of every channel and every decision. That's why we can now have a brand budget and make sure everything is balanced.

Chris: Profitability has been one of the key terms for 2026 across industries. Everybody wants to squeeze more out of the budget. But as Danielle pointed out, you have to protect the core. You can't keep taking and expect it to stay the same. Spoiler alert: that's question three.

00:21:11 – 2026 Recap: AMP's Triangulation Approach

Chris: Kate, how does 2026 look from the agency side?

Kate: We serve a broad portfolio of clients at different points of measurement maturity, with different media scales and channel mixes. Some clients are sophisticated: we've stood up MMMs for them and actively use MMM to optimize investment decisions. Others still rely mostly on platform-reported metrics and last-click attribution.

What worked well in 2026 was when our measurement approach was holistic and based on triangulation. First, MMM gives us the big picture for cross-channel allocation and how budget should be spent across funnel stages. That's where we lean on the incrementality readout. Second, attribution is used for real-time optimization while campaigns are in flight. Third, geo lift or conversion lift tests help us test a new channel or tactic, or confirm a channel is still contributing the way we saw last year. That builds trust in our findings and lets us keep calibrating the MMMs. The strongest measurement programs lean on all three.

Chris: You're like a school principal, seeing every class from those learning to walk to those already running.

Kate: It makes it challenging and interesting at the same time.

00:24:07 – Data vs. Gut Feel: The Pinterest Example

Chris: 2026 could be summarized as growth: in sales channels, in the test-and-learn and incrementality mindset, and in pulling more of a client portfolio into advanced measurement. Next, data versus gut feel. We have more advanced tools and models, but we can't outsource our brains and decision-making to a model. Where do you trust the data, and where do you trust your own expertise? Were there letdowns in 2026? Danielle?

Danielle: Everyone says there's no silver bullet in marketing or media, and it's true. What works for one brand doesn't work for another. A platform that worked two months ago may not work today because it changed an algorithm or got bought. We run a lot of tests to scale into new channels. We look at platform data, our MTA, and Google Analytics, and we triangulate to get some source of truth. We all know the platforms, including the retail media networks, love to show how great they are in-platform.

When you talk to your CFO, they want to see direct return right away, and that's not always the case. Pinterest, for example, doesn't drive a ton of sales for us. It's not where people click and purchase. But Pura is a fragrance experience company. Our job is to help customers make their home a sanctuary. Pinterest is about discovery and inspiration. I'm building a house right now, so I'm constantly on there looking at kitchens and living rooms. And people think about scent as part of the design. They want guests to walk in and say, "Your home smells so good."

So we want to show up when people look for inspiration. Pinterest not driving the sale in the moment doesn't mean it isn't keeping us top of mind for when they search later. If a channel isn't driving direct conversion, maybe we pull it further up the funnel and rethink the content and audience strategy. We can't rinse and repeat what we do on Meta and put it on Pinterest. It has to be curated content. That was a real example from this year: we had to pause, rethink, and focus on that channel to show up the right way.

Chris: Data never tells the whole story. You have to know the context and your customers.

Thaynna: Like Danielle says, we can't fully let go of last-click yet. But with MMM, we can test new channels without feeling alone, because now we have data we trust. We also tested Pinterest, and now we see it as a strategic partner rather than a test channel. We have fashion brands, and our customers go to Pinterest looking for new looks or shopping by occasion. At the beginning it was difficult because we couldn't see a good ROI. But working with the partner and with MMM, we now see a good ROI in GA4 and a better ROI in MMM. We also saw Pinterest drive impact at physical stores, not only ecommerce. We now have a strong relationship with Pinterest and are taking this approach to other partners, like TikTok.

Chris: So you have a gut feeling that some channels last-click pushes down might actually be working, and that's where you need better experimentation. And to be clear, this roundtable is not sponsored by Pinterest. Kate?

Kate: Marketing has always been part science, part art. Even with all the data available to us, we still rely on intuition. Where I trust the data is in well-built MMMs, especially those continuously recalibrated with geo lift or conversion lift tests. When we make investment decisions based on what the MMM tells us, we see the results materialize after the campaign ends.

But we still rely on instinct, for example when the MMM shows weak payback for some upper-funnel tactics and we still want to prioritize them because showing up there for our audience matters and will pay off over time. We simply can't afford not to be in that channel, because it gives an edge to our competitors and could do lasting damage to the brand.

It's getting harder to win those discussions, though. A recent NielsenIQ study shows that the majority of CMOs are prioritizing short-term payback and looking at hard data. So we have to come to the table with hard data plus informed intuition.

Chris: Do clients with last-click only and clients with advanced measurement weigh gut versus data differently?

Kate: Conversations with last-click data only are much harder. You have no data showing what the upper funnel, say premium streaming, contributes. When you're limited to last-click and ROAS, it's natural to lean into the channels showing hard revenue numbers. With an MMM, you have the complete picture of business contribution, including short-term payback for upper- and mid-funnel tactics, and those discussions with CMOs are a lot easier.

00:36:45 – Brand vs. Performance: Finding the Right Split

Chris: Now the good old wrestle between brand and performance. How do you split activities and budgets, and has that changed this year? Kate, would you like to start?

Kate: This is a constant discussion with our clients, almost daily. Historically, research has shown that a 60/40 brand-to-performance split maximizes both short-term performance and long-term growth. Today there's so much pressure to perform this week, this month, this quarter that everyone chases immediate payback and leans into performance. We still have clients who understand the importance of brand and are closer to 60/40.

But leaning too hard into performance doesn't stay cheap. When performance has to absorb all that spend, you quickly hit diminishing returns. The cost to acquire a new customer goes up, and you don't necessarily see incremental revenue or brand growth. We're also seeing CPM increases in the low funnel. This year alone, CPMs for low-funnel Meta, the workhorse for many brands, went up over 20% year over year, and we expect that to keep creeping up. That puts a lot of pressure on ROAS.

Chasing the immediate sale can be dangerous for the brand. You become hostage to ever-increasing costs, competing in the same low-funnel channels as everyone else, mostly capturing existing demand instead of growing new demand. Going into 2027, remember that performance is easy for competitors to copy. Anyone with budget can copy your targeting, creative, audiences, and low-funnel tactics. What's hard to copy is the brand, and investing in the brand is what pays off long term.

Chris: Budgets have shifted toward performance, costs are rising, and we're all competing for the same-sized cake. The winner of that race is the big ad platforms. They get more money for the same outcome.

Kate: Exactly. We need to grow the pie to grow the business.

00:41:42 – How Display Awareness Became Animale's Top Incremental Initiative

Chris: Thaynna, have you seen this at Grupo Azzas 2154?

Thaynna: Yes. We have different budgets for branding and performance, but sometimes performance needs to help the brand, and we're always thinking about how to support brand goals without hurting profit. A concrete example from Animale: in the second half of the year, we had an awareness target that required more investment at the top of the funnel. The challenge was doing that without sacrificing revenue.

We looked at the MMM, measured the incrementality of those campaigns, and proved that Google Display awareness can drive bottom-line sales too. The campaign was optimized for CPM efficiency, and it became Animale's highest incremental impact initiative. It generated nearly half of the brand's projected incremental revenue with only a small piece of the budget, while delivering the account's lowest CPM. That gave us data-based confidence to scale the campaign and hit our impression goals without compromising performance. Sellforte helped us with that.

00:43:32 – How Pura Won Incremental Brand Budget from Finance

Chris: I loved how you put it: sometimes performance needs to help the brand. That's a collaboration we all want to see in 2027. If you can show finance that brand investment also lowers CPMs and other metrics they track, it builds the case for balancing budgets around real return, not short-lived uplift. Danielle, is brand vs. performance a big discussion at Pura?

Danielle: What Kate said resonates. On the agency side, I had this conversation five to eight times a day with clients. On the brand side, I have it with my finance team, chief brand officer, and chief revenue officer, informing them with as much data as we have.

As a DTC brand, you invest heavily in performance. Everyone follows the same model: social, influencers, affiliates, and Google. You grow that baseline as much as you can and scale from there. For Pura to grow, we knew we had to invest in brand and mid funnel so people would know Pura exists. Home fragrance is a large category, but it's fairly stagnant, even declining, as people get more careful about what they breathe in. Pura entered as one of the first smart diffusers, going up against candles and wax melts. We partner with fragrance brands like Nest, Lafco, Brooklyn Candle, and Capri Blue, and we add the technology side and make sure everything is clean and safe. Smart diffusion wasn't something people understood quickly, so there was a lot of education to do. That helped us own the smart home fragrance diffuser category in the US.

Now that we're in retail, we're not up against our traditional .com competitors. We're up against plug-ins and candles, and we're more expensive because of the tech. We have to show why it's worth it, for example that it lasts longer than a candle. So we knew we needed those investments.

The MMM really helped. This is our first year using one. We're using it to plan October to March, our fiscal back half. About four months ago, finance asked us for proactive cuts in upper and mid funnel. The team obliged and said we'd wait and see how business was doing. Business is doing great. So last week we asked for an incremental $3 million. Did we get all three million? No. Did we get more than half? Yes. To me, that's a win with finance. We could do that because the data showed that if we invest, we'll receive incremental revenue we wouldn't otherwise achieve. Maybe not in the next month or two, but in the next year. Another unlock for finance was seeing in the MMM that what we spent in 2025 is still driving sales in 2026.

One more thing. Brands that own their identity and stick to it, instead of using every data point and audience to tweak themselves to fit someone new, are going to win. We've seen a very large brand this year not stay true to itself, and we've seen it feel the impact. Knowing your brand identity and investing in it only makes the performance side better.

And there's organic. We're investing heavily in organic social, PR, and experiential events so media doesn't have to work so hard. We're fragrance: you want to smell it. Same with beauty: people want to try the products. So we run events and capture organic content to bring that experience to the screen, because smell-o-vision isn't a thing. That also lets us show finance the full brand investment and how much of it media carries.

Chris: Getting incremental budget is a big win for marketing measurement. Increasing budgets feels like a blast from the past given where profitability pressure is. Kudos.

Danielle: We've got to celebrate the small wins.

00:50:10 – From Numbers to Decisions: Rebalancing Google and Meta at Animale

Chris: From numbers to decisions. When you present the numbers to finance and leadership, how does that work? Thaynna?

Thaynna: We still look hard at last-click, and we're introducing an MMM culture across the teams. Many of our brands now run MMM, but sometimes we go back to last-click to check. We have a case from Animale that is a clear example of the incremental view versus last-click.

At the beginning of the year, over half our budget was allocated to Google. Ad costs were running above target, and last-click revenue was missing our goals. MMM showed that Google was mostly capturing existing demand, while Meta was generating new demand. So we rebalanced the mix: we reduced Google's share by about 50% and increased the Meta budget by a similar margin. The result: overall revenue nearly doubled, both in last-click and in the incremental view, and ad costs dropped well below target. Google maintained its revenue and efficiency levels despite the smaller budget, which confirmed it was capturing existing demand. We're rolling this out to other brands and other channels, including TikTok and retargeting partners.

00:52:52 – Why Leaders Don't Act on the Numbers

Chris: It's encouraging to see this with a concrete case. As a house of fashion brands, that can be really powerful internally and give you the courage to try new things. Kate, what makes a leader act on a number instead of filing it away?

Kate: A few things lead to not acting. First, contradictory numbers. We have to do our due diligence across all available data sources and agree on the number we present to leadership.

Second, numbers that arrive too late to change anything. It's better to have a directionally right number when budget decisions are being made than a precise number months later when you can't influence anything. Speed of insight is critical for today's leadership.

Third, vanity metrics. Impressions, reach, followers, and engagement rate matter for media optimization, but if they aren't tied to business value, today's CMOs have no patience for them. They need to see the impact paid media drives on the bottom line. When we can say "when we spend X, this is what it drives for the business, and this is how the next dollar should be spent," that's what leaders act on.

00:55:26 – Test, Cut, Scale: How Long Should a Test Run?

Chris: Danielle, how has numbers to decisions worked for you?

Danielle: I live by test fast, cut, and scale. Clear numbers make leadership move faster. We test in a way that lets us walk into a room and say: here's what we tested, here's the hypothesis and objectives, here's what we learned. It didn't work, so we cut it. It worked, so we scaled it. Keep it black and white.

As Pura has grown, we've discussed every test with everyone up to the CEO, and we've learned who the core decision group needs to be. We have a revenue org and a brand org that together form the marketing function, so it matters who makes which decisions. It comes down to the metrics and KPIs: if we do X, we expect Y, and if we don't see Y, why not?

In 2026 we expanded into truly integrated, omnichannel media, and we made mistakes. That's fine as long as it doesn't hurt the bottom line and you're structured about it, burning cash responsibly, as we say. That builds trust with leadership. For example, we have a promo that has worked on .com for four years. I wanted to test new promos, or no promo, on the performance side, and we tested that across platforms for four months. We got interesting insights, but the leading promo still wins. So that's our promo for social, a differentiator against retailers that keeps .com happy, and in other channels we test and scale other things.

Chris: What's a good duration before deciding to cut something?

Danielle: If you ask my merchandising team, it's seven to twelve days, which in media is impossible to learn anything from. Some things you can see in-platform. We've tested OpenAI ads, because everybody's testing AI, but it's been slow. The ads are contextual and hard to serve, because people aren't searching in a way that triggers them. For me, 60 to 90 days. I know it feels slow, but if you can see the outcomes, it works on the brand side. On the agency side it was more like two weeks: you'd better show performance fast. Kate's probably thinking, "If I could get 60 to 90 days for my clients, that would be amazing."

Kate: Absolutely. We're pressured to deliver results fast, which compresses testing time. We push for a good readout, and four to six weeks is a good benchmark for test duration.

01:00:38 – One Takeaway for 2027

Chris: Our last question: one takeaway for 2027. Danielle?

Danielle: There are a lot of shiny pennies out there. Protect the core of what you know works for your business. Every planning cycle, it's tempting to chase whatever new channel or partner is getting attention. Keep testing, absolutely, but testing is different from abandoning the fundamentals that drive your revenue today. My warning from 2026 is what I call measurement anxiety: the urge to prove every dollar is incremental can push you to overcorrect and cut investments just because they're harder to measure cleanly, like brand. Staying true to the brand will matter more than ever in 2027, with more players chasing the same channels and the same AI-driven efficiency plays. The brands that stand out won't lose their identity chasing the next platform. Test the edges, protect your core.

Thaynna: I agree we need to test, and I like to test faster. This year we learned to test faster and keep the learnings. We used to test and scale. Now we test, learn, and then scale. For the first time, we have an incremental 360-degree view across all media channels for all our brands, so we can allocate the next budget, the next dollar, based on the true incremental view. Test, learn fast. If it works, scale. If it doesn't, stop and keep going. That doesn't mean we stop investing in Google. One month is one scenario, the next month is another. We keep testing YouTube and other formats and making new decisions based on what we see.

Kate: I agree on testing. What I'd also like to see more of in 2027 is to stop asking what a channel drove last quarter and start asking what the next dollar is worth. We need to shift MMM usage from backward-looking reporting to forward-looking action. I recently read an eMarketer report showing that fewer than 30% of marketers say their organization is effective at using MMM insights. So 70% of those with access to MMM aren't using it effectively. We have to change how we use MMM insights to drive the next decision.

01:05:20 – Q&A: Building Confidence to Grow Brand Investment

Chris: From reactive to proactive. Classic MMM has a reputation for staying high-level, but you need to get hands-on. Insights need to arrive today to be useful. If you're looking at last quarter, you're missing out big time. We're at time, but let's take one question from the audience: how did you gain confidence to grow investment in brand marketing?

Danielle: When we first decided to spend more in upper and mid funnel, finance asked us what budget we wanted, so they could cut it down, obviously. It differs by industry, so I won't give a percentage, but I researched CPG categories and startups to understand typical brand-to-performance ratios. That changes at every stage of a business, same as marketing cost of sales. Using industry data meant finance couldn't say we'd pulled the number out of thin air.

The biggest question from finance was when they'd see the return. I said it could be anywhere from 12 to 24 months. It could happen in four to six, but I wanted to set a longer range for brand.

Then there was a lot of re-education internally. People said, "We're not seeing new .com customers tick up with this brand investment." But new Pura users overall, buying at retail, receiving it as a gift, or buying on .com, are growing. That's brand. If I spend on CTV, I won't necessarily see a direct return on .com. Someone sees a CTV ad, then a Meta ad, searches on Google, then walks into Target, sees us on an end cap, and buys. It's not as black and white as it used to be, so we educated finance on the shopper journey. We're marketers, but we're consumers at our core.

Then we needed a tool to validate it, and for us that was the MMM. It showed what we'd seen within the first six months, and from last year's investment. Now, as Kate said, we use it for forward planning: our blue sky model and the MMM help finance plan the next fiscal year, so it isn't just "we'll only spend this much because of EBITDA." We can show that we may need to take a little off the top to keep growing.

Kate: Being at a full-funnel agency, we're lucky to have strong leaders who educate clients on why to invest in brand and upper-funnel tactics. Like Danielle said, a lot of it is education: how much to invest, why, and what it will do for the business. Then we bring in MMM to read out the true value of those upper-funnel tactics.

01:11:03 – Wrap-Up

Chris: We're over time. Thank you, Kate, Danielle, and Thaynna. I could keep going for another hour. Thanks to everyone who joined. We'll share the recording on our website, and we'll follow up on the questions we didn't have time to answer.