AdTech companies preach rigorous measurement to the market and then run their own acquisition on two creatives and a guess. A real testing system – structured variants, clean reads, fast iteration – is the difference between a demand engine that compounds and one that plateaus the moment a channel saturates.
The Problem
You preach measurement but test creative on vibes
AdTech sells attribution, incrementality, and clean measurement, then runs its own paid acquisition on two ad variants chosen by whoever spoke loudest in the standup. The hypocrisy is invisible to prospects but fatal to your own pipeline – you are leaving conversion lift on the table in the exact discipline you charge customers for. When the founder asks why CAC is climbing, nobody can point to a test that explains it, because there is no test framework underneath.
Signal loss makes naive testing read noise as signal
Cookie deprecation, iOS privacy changes, and platform attribution windows have degraded the data most teams use to call a creative winner. Teams declare winners on tiny samples, attribution gets blurry across the buying journey, and the same conversion gets counted by three platforms. Without a testing design that accounts for signal loss – holdouts, geo-tests, or incrementality reads – you are iterating toward whatever the broken measurement happens to flatter. You optimize confidently in the wrong direction.
Iteration is too slow, so winning variants decay before scaling
By the time a generalist team designs a test, ships variants, waits for significance, and produces the next round, the winning concept has already fatigued in-channel. AdTech audiences – performance marketers, agency buyers, technical brand teams – are unusually ad-aware and burn out fast. Slow iteration means you scale yesterday's winner into today's decay, and the testing program never builds the compounding library of learnings that makes the next test smarter.
No learning system means every test starts from zero
Most AdTech teams run tests as one-offs – run it, pick a winner, move on – with no structured record of which hooks, formats, and messages won across audiences. The institutional knowledge lives in one marketer's head and walks out the door when they leave. Without a testing taxonomy and a living creative learnings library, each quarter relearns what the last quarter already knew, and the cost-per-learning never drops. The program stays expensive and slow because nothing accumulates.
How We Help
We start with how you are actually calling winners today. In the first 30 days we audit your current testing setup – sample sizes, attribution sources, significance thresholds, and iteration speed – and we usually find that the measurement underneath the tests is too broken to trust the verdicts. For an AdTech company, that is the most important and most embarrassing finding, so we fix the read before we touch the creative.
Strategy designs a testing system fit for a post-cookie signal environment. We define a creative testing taxonomy – hook, format, message, audience – so every test isolates a real variable instead of changing five things at once. We design clean reads using the tools that survive signal loss: holdout groups, geo-based incrementality, and platform-agnostic conversion tracking, the same measurement rigor your product sells. We set significance and sample thresholds appropriate to your spend so winners are real, not noise.
Execution runs the iteration loop fast. We ship structured variant batches, read them against clean measurement, kill losers early, and turn winners into the next round of hypotheses within days, not weeks. Because AdTech audiences fatigue quickly, speed is the whole game – we build the production-to-test-to-iterate cadence that scales winners while they are still working. This connects directly to your creative production and demand-gen motions so the test pipeline never starves.
The fractional model means you get a senior growth operator running the testing discipline plus the analytical and creative capacity to feed it, without hiring a full-time growth lead, an analyst, and a creative strategist. We embed in your existing marketing org and run the testing system as part of your weekly demand cadence rather than as an outside experiment.
Measurement is the point of the whole engagement, so we hold ourselves to it. We track cost-per-learning, win rate of tested versus untested creative, conversion lift from winners, and how fast the iteration loop turns. The deliverable that matters most is a living creative learnings library – what wins, for whom, in what format – so cost-per-learning drops every quarter and the program compounds instead of resetting.
What we deliver
- Creative testing taxonomy isolating hook, format, message, and audience as discrete variables
- Clean-read measurement design using holdouts, geo-incrementality, and platform-agnostic tracking
- Fast iteration loop with structured variant batches and early-kill criteria
- Living creative learnings library that compounds across quarters
- Significance and sample framework calibrated to your spend level
- Testing performance dashboard tracking cost-per-learning and conversion lift
AdTech companies sell measurement rigor to the market and run their own creative testing on gut feel. The fix is not more tests – it is fixing the read first, because a fast iteration loop on broken measurement just optimizes confidently in the wrong direction.
Our Methodology
Our creative testing build for AdTech runs as a 90-day system installation. Phase one audits how you currently call winners – sample sizes, attribution sources, iteration speed – and fixes the measurement underneath the tests before touching creative. For an AdTech company that sells incrementality, an honest read is non-negotiable, so this phase usually exposes that the data was too broken to trust prior verdicts.
Phase two designs the testing system: a taxonomy that isolates hook, format, message, and audience as discrete variables, a clean-read design using holdouts and geo-incrementality that survives cookie deprecation, and significance thresholds calibrated to your actual spend. This is what separates a testing program from a pile of A/Bs.
Phase three installs the iteration cadence and the learning system. Structured variant batches ship on a fast loop, losers get killed early, winners become the next round of hypotheses, and every result feeds a living creative learnings library. Unlike agencies that run tests as billable one-offs, we build a system where cost-per-learning drops every quarter because the knowledge compounds.
How We Work
Initial engagements run 3 to 5 months because building a testing system requires fixing measurement, designing the taxonomy, and running enough iteration cycles to prove the loop compounds. The first 30 days are the testing audit and measurement fix with marketing and analytics leadership. Days 31 to 60 design the testing taxonomy, clean-read framework, and significance thresholds. Days 61 to 90 run the iteration loop at full speed and stand up the learnings library.
Our team includes a growth operator who owns the testing system, an analyst who designs and reads the experiments, and a creative strategist who turns learnings into the next variant batch. From your side we need access to ad accounts and conversion data, alignment with creative production so the variant pipeline never starves, and a single owner for spend decisions. We handle test design, measurement, iteration cadence, and the learnings library.
Weekly test reviews read live experiments and set the next batch of hypotheses. Monthly business reviews tie testing to CAC, conversion lift, and cost-per-learning trends. Most AdTech companies see iteration speed improve within 30 days and measurable conversion lift from winning variants within 60 to 90 days, with cost-per-learning dropping each quarter as the library grows.
- Conversion lift from systematically validated winning creative
- Lower cost-per-learning as the creative learnings library compounds
- Faster iteration that scales winners before they fatigue
- CAC stability through disciplined, clean-read testing
If your adtech company needs creative testing & iteration 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 creative testing cost for AdTech companies?
Most AdTech creative testing engagements run between $20K and $50K per month depending on spend level, test volume, and how much variant production sits inside the engagement. That is less than hiring a full-time growth lead, an analyst, and a creative strategist, which is the team a real testing system actually requires. Cost scales with how many channels you are testing across and the iteration speed you want to hit.
How long before we see results from a creative testing engagement?
Iteration speed typically improves within the first 30 days once the testing taxonomy and clean-read framework are live. Measurable conversion lift from validated winners shows up within 60 to 90 days as the loop surfaces and scales winning variants. Cost-per-learning drops each quarter after that as the creative learnings library accumulates and tests get smarter.
How does the creative testing team integrate with our existing staff?
We embed in your weekly demand cadence and work directly inside your ad accounts and conversion data rather than operating as a detached experiment. We need alignment with creative production so the variant pipeline keeps feeding the test loop and a single owner for spend decisions. Your team keeps channel ownership while we run test design, measurement, and the iteration system.
What makes Winston Francois different from a traditional testing agency?
Most agencies run tests as billable one-offs with no learning system, and many use the same broken attribution your team already mistrusts. We fix the measurement read first, build a taxonomy that isolates real variables, and stand up a learnings library so cost-per-learning drops every quarter. We are operators applying the same measurement rigor your product sells to the market.
How do you measure ROI from a creative testing engagement?
We measure conversion lift from winning variants, win rate of tested versus untested creative, cost-per-learning over time, and CAC stability as testing scales. The headline metric is conversion lift per dollar of test spend, trending up as the learnings library compounds. Most AdTech companies see clear testing ROI within 60 to 90 days and a declining cost-per-learning curve thereafter.
What type of AdTech company is the right fit for this service?
Growth-stage AdTech companies running real paid acquisition spend who are calling creative winners on gut feel or broken attribution. Companies between $5M and $100M ARR with enough spend to read tests cleanly and a creative pipeline that can feed iteration see the strongest fit. The first step is a testing audit that exposes how reliable your current winner-calling actually is.
