The tools and techniques we use to send the strongest possible proposition across the start line
The principle
Our job ends at launch. Yours is just beginning.
The data on Year Two is unambiguous. Winners separate from losers not at launch but in Month 18, as distribution bottlenecks clear, trial converts to repeat, and awareness compounds. The brands that survive the trial-to-repeat valley are the ones whose senior teams keep funding the launch when the early numbers wobble. The brands that fail are the ones where someone in finance pulls support in Month 7.
That commitment isn’t something we can deliver for you. But the strength of the proposition you’re committing to absolutely is.
So our role at this stage is singular: send the strongest possible proposition across the start line. One that has been stress-tested on pack, claim, price, and story — before the launch budget gets spent, not after.
The question we ask, on every project: what would have to be true for this launch to succeed — and what would have to go wrong for it to fail? And have we eliminated, in advance, every weakness we possibly can?
The human work that doesn’t change
There are three pre-launch stress tests that have made the difference between the launches we’ve worked on that succeeded and the ones that didn’t.
Simulated test markets. Concept Sauce or even BASES-style quant validation of the optimised proposition, with proper sample sizes and defensible methodology. The job is to identify any remaining weaknesses in pack design, claim hierarchy, price architecture, and proposition clarity — before the launch is irrevocable.
Story-product match testing. The single biggest predictor of repeat purchase is whether the product delivers the experience the proposition promised. We test the claim against the product, the product against the claim, and refine both until the gap closes. Nielsen BASES’ 15x finding rests on this discipline.
Pre-mortem and pre-retrospective workshops. Two structured stress tests we run before sign-off. The pre-retrospective imagines the launch has succeeded and works backwards to identify what had to be true for that outcome — surfacing the assumptions the proposition depends on. The pre-mortem imagines the launch has failed and works backwards to identify the proximate causes — surfacing the weaknesses the team is currently rationalising away. Used together, they catch the things that quant testing misses: the political risks, the executional fragilities, the commercial assumptions nobody has questioned.
This is the work that builds a launch worth committing to. Skip it, and your Year Two budget is at the mercy of whichever doubt surfaces first.
How we optimise at this stage
Pre-launch optimisation is the final pass on the gap between story and product. We run co-creation sessions specifically focused on whether the experience delivers what the claim promised — and where it doesn’t, we refine until it does. This is the work that drives repeat purchase, which is what determines whether the launch survives the trial-to-repeat valley. Without it, you’re committing a Year Two budget to a proposition whose product hasn’t yet earned it.
Where AI is now genuinely useful
The honest position: AI is less transformative here than at the insight or proposition stage, by some distance. But there are three places where it meaningfully sharpens the pre-launch stress test.
Predictive concept screening. Tools like Zappi, Quantilope and Upsiide now use machine learning trained on historical concept-test outcomes to predict in-market performance from concept text. Zappi’s work with PepsiCo has demonstrated 50% greater accuracy in forecasting short-term sales impact than legacy pre-testing — a finding cited on the record by Stephan Gans, PepsiCo’s Chief Consumer Insights and Analytics Officer. We use this work to narrow the option set before committing to a full BASES round, not to replace BASES itself.
Visual attention prediction. Neurons’ Predict tool, trained on twenty years of eye-tracking data from 300,000 participants, predicts where consumer attention will land on a pack design, an ad, or a shelf set. We use it to pressure-test pack hierarchy, claim visibility and creative impact before the work goes into quant. It catches the obvious mistakes — brand asset invisible, claim buried, hierarchy wrong — at a cost that makes iteration genuinely affordable.
AI-augmented story-product match testing. Using structured Claude-based pipelines, we stress-test the proposition narrative against the actual product experience — looking for the gaps where the claim overpromises, the experience underdelivers, or the language doesn’t match what consumers themselves would say. This is methodology rather than vendor tool, but it materially closes the gap between proposition and product before the launch budget gets committed.
Where AI isn’t the answer
Year Two success is not an AI problem.
The biggest predictor of whether a launch survives the trial-to-repeat valley is whether the senior team has the courage to keep funding it. No predictive model substitutes for a CFO who holds the budget. No attention-prediction tool substitutes for a CMO who can still tell the proposition story compellingly twelve months in.
What AI does, at this stage, is reduce the number of weaknesses your team has to defend. It catches the obvious mistakes earlier and cheaper, so the proposition that reaches launch is structurally stronger. The commitment to keep funding it is yours to find.
What Brand Development is doing today
At the pre-launch stage, our AI use is more selective than at the insight or proposition phase — by design.
Where AI sharpens the stress test (predictive screening, attention prediction, story-product match analysis), we deploy it. Where the pre-launch work is fundamentally human — the pre-mortem, the pre-retrospective, the conversation with the senior team about what they’re prepared to commit to — we focus our energy on that, because that’s what the launch actually needs.
The lever it pulls
Optimised propositions are four times more likely to succeed than unoptimised ones. That 4x multiplier is built before launch, not after — and it is the strongest single lever in the entire NPD process.
Our job at this stage is to make sure that when the proposition crosses the start line, every weakness we can find has been found, every assumption has been tested, and every gap between story and product has been closed.
What happens next is up to you. We just make sure you start from the strongest possible position.
Next
[Read about The Lever That Changes the Odds →]
[Or talk to us about a pre-launch stress test →]