CTV doesn't click, doesn't carry a cookie, and doesn't report the same number twice across Roku, Hulu, YouTube TV, and your DSP. Winston Francois builds the unified reporting, incrementality testing, and revenue attribution that tells you what your CTV spend is actually doing – not just what each platform says it did.
The Problem
Your CTV impressions don't click through
CTV ads run on a ten-foot screen with a remote, not a mouse, so there's no click for standard digital attribution to hang on. Most teams fall back on view-through windows borrowed from display advertising, which credit CTV with conversions it may have had nothing to do with. Depending on which window you pick, the same campaign can look either highly effective or completely dead. That's not a measurement result, it's a coin flip dressed up as a number.
Every platform grades its own homework
Roku, Hulu, YouTube TV, Amazon, and your DSP each generate their own dashboard, with their own definitions of reach, frequency, and completed view. None of them talk to each other, and none of them are incentivized to show you a number that looks bad. Stitching five walled gardens into one picture is a manual exercise most teams run once a month, usually too late to change a media plan already in flight.
No cookies means no individual-level truth
CTV devices sit mostly outside the cookie and mobile-ID ecosystem, so you can't deterministically tie one household's ad exposure to their later purchase. Probabilistic matching and IP-based identity graphs fill part of the gap, but they carry real error rates most vendors don't publish. Teams that treat CTV like a performance channel with clean, person-level attribution end up making budget calls on numbers that were never that precise.
Impressions never make it to the revenue conversation
Marketing reports reach, completion rate, and frequency; finance wants pipeline, revenue, and payback period. Without a pipeline connecting CTV exposure data to your CRM or revenue system, those two conversations happen in different rooms and never reconcile. That's usually where CTV budget gets cut first in a downturn – not because it isn't working, but because nobody in the room can prove it is.
How We Help
We start with an audit of what you already have: which CTV platforms you're buying on, what each one's native dashboard actually measures, what your DSP reports versus what the publisher reports, and where those numbers already disagree. Most audits turn up gaps before we add anything – duplicate reach counted across platforms, completed-view definitions that don't match, or an attribution window nobody remembers choosing.
From there we build a measurement plan sized to your spend and your decision cadence, not a generic template. A brand spending modestly on CTV needs a lighter approach – directional geo tests and simple trend reporting. A brand spending heavily can support a running incrementality testing program with its own testing calendar. We size the plan to the budget, not the other way around.
Incrementality sits at the center of the plan, not bolted on afterward. Matched-market or geo-holdout tests – turning CTV on in some markets and holding it off in comparable control markets – are the only reliable way to isolate what CTV is actually adding versus what would have happened anyway. We design the test cells, select comparable markets on demographic and historical-performance fit, and run the read once the flight closes.
Alongside testing, we build a single reporting layer that pulls impression, reach, frequency, and completion data out of every platform – Roku, Hulu, YouTube TV, Amazon, your DSP – into one dashboard with shared definitions. This is unglamorous work: reconciling how each platform counts a completed view, normalizing date ranges and time zones, deduplicating households counted twice across platforms.
For brands running CTV alongside paid social, search, and other channels, this measurement work extends into media mix modeling to estimate each channel's contribution to revenue over time, including how they interact with each other.
The last piece connects CTV exposure data to whatever system finance actually trusts – CRM, revenue reporting, sometimes downstream product usage signals when the business is in that shape – so a media plan review and a board conversation about pipeline are working from the same numbers instead of two different ones. That reporting layer gets folded into your broader marketing strategy work, not run as a side project nobody owns.
If your CTV reporting currently lives in five browser tabs and a gut feeling, that's the starting point for this engagement, not a disqualifier – book a strategy call and we'll walk through what a unified view would look like for your specific platform mix.
What we deliver
- Unified CTV reporting dashboard across all platforms and DSPs, with shared metric definitions
- Incrementality testing program – matched-market and geo-holdout test design plus post-flight analysis
- Media mix model incorporating CTV alongside your other paid channels
- CTV-to-revenue attribution pipeline connected to your CRM or revenue system
A completed-view rate is not a business outcome – if you can't say what would have happened without the CTV spend, you don't have a measurement program, you have a vendor's dashboard.
Our Methodology
We treat CTV measurement as three separate problems that usually get conflated into one: reporting (what happened), incrementality (what CTV caused), and attribution (how it connects to revenue). Most teams only have the first, mistake it for the second, and never get to the third. We build all three in sequence, because each depends on the last one being solid – a beautiful dashboard built on unvalidated attribution logic just makes bad numbers easier to read.
That's why we prioritize test design over dashboard polish in the first phase. Geo-holdout and matched-market tests get set up before we spend real effort on visualization. Once we know CTV is actually moving outcomes, the reporting layer has something real to report, instead of just relaying whatever the platforms want you to see.
How We Work
The first 30 days are audit and design. We pull every platform's native reporting, document the discrepancies, and design the first incrementality test – market selection, holdout structure, and the metric we're testing against. You leave this phase with a written measurement plan and a first-pass unified dashboard built from what already exists.
Days 30 to 60 are execution. The incrementality test runs live, the unified dashboard gets automated instead of hand-pulled every week, and we start media mix model data collection if that's in scope. You're already seeing one reporting view instead of five, well before the test results land.
By day 90 the first incrementality read is complete, the dashboard runs on its own, and we deliver a report on what CTV is actually contributing versus what the platform-reported numbers implied. That gap – between reported impressions and measured impact – is usually the single most useful output of the whole engagement.
You work with a small, consistent team: a lead who owns the measurement strategy and a data and analytics operator who builds and maintains the pipeline, not a rotating account team you have to re-explain your business to. Cadence is a standing weekly check-in during the build phase, moving to biweekly once reporting is automated and the testing calendar is set.
- Unified visibility across fragmented CTV platform dashboards
- A causal read on CTV's actual contribution to revenue, not just platform-reported impressions
- Defensible numbers for the media plan conversation with finance
- A repeatable testing calendar instead of one-off measurement projects
If your ctv / connected tv company needs data, reporting & analytics leadership, we should talk.
Expand your marketing team output with our experts
Let us take a custom approach to your growth goals by assembling and leading the best-in-class marketing team to support your next stage.
Frequently asked questions
How much does CTV measurement and reporting work cost?
Cost scales with how many platforms you're reconciling and whether an ongoing incrementality testing program is in scope, so we don't quote a flat number before we've seen your platform mix. A brand running CTV on two or three platforms with one testing cycle a year is a smaller engagement than one running five platforms with a quarterly testing calendar.
How long until we see results?
You get a written measurement plan and a first-pass unified dashboard within the first 30 days, built from platform data that already exists. The first incrementality test read – the part that tells you what CTV is actually causing, not just reporting – typically lands by day 90, timed to your test's flight length.
What does the team structure look like?
You work with a lead who owns the overall measurement strategy and a data and analytics operator who builds and maintains the reporting pipeline day to day. It's a small, consistent team rather than a rotating group of account staff, so you're not re-explaining your platform mix and business context every quarter.
What makes Winston Francois different from relying on the CTV platforms' own reporting?
The platforms selling you CTV inventory have an incentive to report numbers that make their own inventory look good, and none of them are set up to tell you what would have happened without the spend. We don't sell media, so we have no reason to protect any single platform's number – our job is to build the independent, unified view and run the incrementality tests that show what's actually working, even when that means telling you a platform isn't performing.
How do you measure ROI without cookies or click-through data?
We rely on matched-market and geo-holdout incrementality tests, which turn CTV on in some markets and hold it off in comparable control markets to isolate the causal lift, and on media mix modeling to estimate CTV's contribution alongside your other channels over time. Neither approach depends on a cookie or a last click, which is the point – they're built for a channel that was never trackable at that resolution to begin with.
What kind of company is the right fit for this?
This fits Series A/B and growth-stage companies, typically $5M to $100M in ARR, that are already spending meaningfully on CTV across more than one platform and can't currently get a straight answer on what it's contributing to revenue. It's a poor fit for a company just starting to test CTV with a single platform and no budget to run a real incrementality test yet – that team should start with a smaller pilot before building out full measurement infrastructure.
