Outdoor Voices spent 30-40% of all raised capital on Instagram and Facebook customer acquisition. That was the growth model. The dashboard said it was working. ROAS looked strong. The audience was growing. The brand was scaling.

Then Apple changed its privacy settings. Tracking broke. Overnight, the dashboard went dark on the data that justified the spend. The brand pivoted to wholesale, but by then it had burned through the capital. The founder was ousted by the board.

"We ran into the Zuckerberg paywall. We ran into the Apple privacy changes. We ran into Google third-party cookies. We're like, okay, pivoting from DTC now. We're all about wholesale."

Eric Liedtke, founder, Unless Collective

Eric Liedtke, the former EVP at Adidas who created the Yeezy line, launched Unless Collective as a DTC-first sustainable fashion brand. Same plan. Same dashboard. Same result.

Two brands. Two experienced operators. Same dashboard. Same blind spot. Both built their entire capital allocation strategy on what the marketing dashboard could see. Neither dashboard could see the retail revenue their marketing was generating.

What the dashboard measures and what it misses

The tools your marketing team uses were built for a specific world. A person sees an ad, clicks, visits a website, adds to cart, purchases. The tools measure every step of that path with precision. ROAS. CAC. LTV. CPA. Conversion rate. Every metric tracks what happens on a screen.

A consumer brand selling 40-60% of revenue through physical retail does not live in that world.

The customer who sees your Instagram ad, doesn't click, walks into Target three weeks later, recognizes your packaging, and puts it in the cart. That purchase is invisible. The brand awareness that drove the recognition compounds over months. The dashboard measures in 7-day or 28-day attribution windows. A campaign building recognition that drives shelf purchases six months later shows zero ROAS in week one. The measurement collapses long-term effects into short-term metrics. What can't be measured in 28 days is treated as if it didn't work.

This boundary is not a technical limitation that better software will fix. The tools were designed for DTC. They measure the digital funnel because the digital funnel is what they were built to see. Everything off-screen, offline, in-store, or on a timeline longer than a month is outside the frame. Not because it doesn't matter. Because it was never in scope.

What NIVEA discovered about its own data

NIVEA added omnichannel measurement to its marketing analysis. ROAS jumped 80%.

The campaigns had not changed. Nothing about the marketing had improved.

The retail revenue had been there the entire time. The dashboard was structurally unable to see it. When the measurement frame expanded, the performance that was always happening became visible.

Every budget meeting, every campaign review, every "this channel isn't performing" conversation at NIVEA before that measurement change was happening inside a frame that could not see where the money was actually landing. Prescient AI found the same pattern across its client base. A channel showing 2x ROAS on ecommerce was delivering 4x when retail was factored in.

The campaigns were not underperforming. The measurement was underseeing.

Two scorecards, zero overlap

Your retail buyer evaluates you on margin percentage, turns per SKU, sell-through rate, and same-store sales. Your marketing team evaluates campaigns on ROAS, CAC, CPA, and conversion rate.

These two scorecards do not share a single metric.

A brand could be winning on both scorecards and losing money. Or losing on both and building long-term value. Neither scorecard can tell because neither can see what the other measures.

The CFO hears "our ROAS is 3.5x" and cannot translate that into total profit across all channels. The CMO hears "sell-through is down in Q3" and cannot connect it to the brand campaign they cut in Q1. The data lives in two systems that were never designed to talk to each other. Not because nobody thought to connect them. Because the institutions that produce the data have no incentive to make it interoperable.

"Your CMO and CFO are literally looking at different pictures of performance. Marketing talks about digital metrics. Finance talks about retail revenue. The conversation goes nowhere because they're not looking at the same data."

Prescient AI

P&G solved this by building dedicated teams on each side with a coordination layer between them. That coordination layer is a significant organizational cost. An 8-person brand cannot build it. The brand making the most consequential capital allocation decisions of its life is making those decisions with two partial pictures and no mechanism to combine them.

The loop that validates itself

When the dashboard says a campaign is working, the recommendation is to scale it. When the dashboard says a campaign isn't working, the recommendation is to change the creative, adjust the targeting, increase the budget, or test a new platform.

The recommendation is never that the dashboard might be measuring the wrong thing.

The agency doesn't report on offline performance because the agency can't see offline. The platform doesn't surface data that contradicts its own value proposition. The tool doesn't flag that its attribution window misses long-term effects. Every step in the chain references the step before it. Nobody in the chain provides signals from outside the chain.

You can fire your agency. The next agency runs the same playbook on the same platforms with the same tools. You can leave Meta for TikTok. Same measurement framework. Same attribution models. Same optimization logic. You can switch from Klaviyo to Braze, from Google Analytics to Amplitude. New dashboard. Same boundary.

You have freedom of choice at the actor level and no freedom at the system level.

What this costs when it meets retail

A brand spending $30,000 per month on Meta ads is building awareness that partially drives retail sell-through. But the dashboard shows zero of the retail impact. The CMO, under pressure to justify spend, shifts budget from brand campaigns to performance campaigns. The dashboard validates the decision. Sell-through at Target starts declining two quarters later. Nobody connects the two events because the data lives in different systems.

Co-op advertising, promotional allowances, and markdown money are marketing costs that live inside the retail institution's margin architecture, not inside the marketing budget. The CMO's total marketing spend doesn't include them. The brand is paying for marketing through two institutions and accounting for it in two separate categories. True total marketing spend is always higher than either dashboard reports.

A brand runs a DTC promotion while the retailer runs an in-store promotion on the same product. Marketing says DTC sales spiked. Retail says sell-through held steady. The brand may have cannibalized one channel with the other. Same units sold. Higher total cost. Lower total margin. Each dashboard shows a win. The combined picture is a loss that neither dashboard can see.

Lunya, the DTC sleepwear brand, expanded into physical retail stores. The stores lost $135,000 per month while contributing 8% of revenue. The brand filed for Chapter 11 in 2023. The measurement frame that justified the expansion could not see the cash it was consuming.

The question underneath the metrics

The marketing institution was designed around a digital purchase funnel. The retail institution was designed around in-store economics. Each built its own metrics for its own participants. Neither was designed with the other in view.

The brand sits between them, drawing from one margin pool to pay both. The platforms profit from the current frame. The agencies are paid within it. The retailers measure their own side.

The brand is the only party that needs to see across both institutions. And the brand is the only party whose tools were never built to do it.

We have sat across from founders looking at dashboards that say everything is working while the cash position says everything is breaking. The numbers all look right. The ROAS is healthy. The sell-through is steady. The revenue is growing. And somehow the cash is tighter than the dashboard says it should be. The model says one thing. The bank account says another.

Most assume they did something wrong. They didn't. They built a business inside two measurement systems, each showing them the half it was designed to see.

Your business lives in the space between them. Your tools do not.

MMK Retail helps consumer brands see the design of the retail institution before the institution shows itself.