The buy-side manifesto
You bought the measurement. Now own it.
Your brand measurement lives in twelve places. None of them is the answer. Every vendor delivers their fragment and calls it the source of truth. You pay for measurement, and what you get back is an assignment.
This is why that happened, why nobody fixed it, and what changes now.
6 minute read · Sonaflo · 2026
You pay for measurement. You get fragments.
You opened your quarter last Monday. Six Brand Lift studies in the last 90 days. Nielsen for the CTV flight. Kantar for the spring campaign. DISQO for always-on. YouTube Brand Lift buried in Google Ads. Meta Brand Lift in an XLSX with columns that changed again. An Upwave PDF landed in your inbox with no context.
You paid four hundred thousand dollars for that.
What you have is six PDFs and two spreadsheets. No single chart for the CMO. No way to answer “is awareness up or down this year” without two days of analyst time.
Six checks. Six fragments. No answer.
This is not a story about bad vendors. Each one delivered exactly what you paid them for. The story is what you paid them for: a fragment. The industry sells fragments. It has never sold integration. It has no economic reason to start.
Two vendors. Same metric. Different numbers.
Two of your vendors measured aided awareness on the same campaign. They both called it aided awareness. They came back with different numbers. Sometimes four points apart. Sometimes twelve.
Different control groups. Different exposure validation. Different audience weighting. Different question wording. Different timing windows.
Every vendor knows this. It is in their methodology disclosures — the ones nobody reads.
What they tell you is: “Don't compare across vendors. Methodologies differ.” Which is true. It is also commercially convenient. It forecloses on the question you actually want to answer: which one is right? Or, fine, what does the consensus look like?
You have two options. Pick one vendor and ignore the rest. Or run all of them and pretend they are interchangeable.
Both are losing trades. Both persist because the only fix — a tool that reconciles vendors honestly — is something no individual vendor will ever build.
Standards failed because vendors set the standards.
The MRC. The ARF. The IAB. Each has a measurement working group. Each working group is staffed by — guess who — the vendors.
The vendors fund it. The vendors vote. The vendors certify themselves compliant with the standards they wrote.
This is not a conspiracy. It is the predictable outcome of asking competitors to standardize a market that rewards them for not being standardized. Methodologies become “proprietary.” Audience cells become “exclusive.” Benchmarks become “branded.” Every layer of the stack gets harder to substitute. By design.
Twenty years of standards work. More working groups. More compliance certifications. Zero usable dashboards.
The fix has to come from somewhere with no methodology to defend, no panel to sell, no certification to charge for.
The fix has to come from the buy side.
You own the measurement. Now act like it.
When you pay Kantar two hundred thousand dollars for a study, who owns the deliverable?
You do. Always have.
So why, for twenty years, has nobody built you a workspace to organize the work you already own?
Two reasons.
First — the workspace commoditizes the vendor. A platform that lets you compare Nielsen and Kantar honestly is a platform that lets you notice when they disagree, and ask why, and use the answer in your next negotiation. Vendors lose the asymmetry that pays the bills. No vendor will build that.
Second — integrating eight vendor formats with quietly changing templates was prohibitively expensive until very recently. Reading a Kantar deck and pulling a structured fact table was a human analyst job. Doing it across your portfolio was a team of analysts. The integration tax killed every previous attempt.
That tax is gone.
AI changes the math.
For most of the last twenty years, the bottleneck on the buyer-owned workspace was not design or sales. It was parsing.
Modern language models read PDFs. They read slides. They read spreadsheets. They read charts embedded as images. They read methodology disclosures and pull canonical audience-cell definitions. With deterministic validation on top — sanity checks, schema enforcement, human review on low confidence — the integration problem is finally tractable.
Tractable. Not solved.
Vendor templates change without notice. A chart baked into an image is hard to read. Co-branded reports are weird edge cases. All real problems. All engineering problems — not staffing problems. That is the threshold change.
The constraint stops being “can we extract the data.” It starts being “can we organize it well enough that you actually use it.”
That is a product question. Product questions have answers.
One study tells you something. Years tell you a story.
We are not adopting a new framework. We are not declaring a new metric. We are not certifying vendors. We have no interest in being the arbiter of measurement truth.
We are building the workspace you should have had for twenty years.
Your measurement. Organized. In one place. So you can finally see your brand across vendors, across time, across audiences — and hand the picture to your CMO without rebuilding the deck.
A single study answers a question about one campaign, at one moment. Years of studies answer a bigger one: how is the brand actually moving? Every report you have ever commissioned is an asset you already paid for. Sonaflo makes them cumulative — so the next question starts from everything you have learned, not from a blank page.
This is your data, in your workspace, on your terms. You bought the measurement. Now you get to own it. That is the whole idea.
We are building this. Now.
If you have been waiting a decade for someone to fix this — if you are a brand or insights lead rebuilding the same QBR deck for the fifth Monday in a row, tired of watching years of expensive research scatter and disappear — talk to us.