Pre-launch Testing

Why this matters for your decision

After the launch, most of the costs have already been incurred: development, production, listing, campaign, training. What can still be optimised then are details. The big levers, that is whether, for whom and at what price, lie before the launch.

A pre-launch test brings evidence to this point. The closer the evidence is to actual behaviour, the less the business case has to rest on assumptions. Surveyed purchase intent is a weaker indicator for this than it seems: the act of surveying itself can make the link between intent and purchase appear larger than it is.

Forms at a glance

Survey-based pre-launch tests measure acceptance, willingness to buy on a scale or preferences between attributes. They are fast and good for selecting from many ideas. Market simulations extrapolate expected sales from survey data and assumptions. Regional trial launches introduce the product for real in a limited area; this is meaningful, but expensive, slow and visible to competitors.

Behavioural tests before launch, such as Painted Door Tests, sit in between: real people respond to a realistic offer without it having to exist. Horizon runs them with Google and Meta ads, with up to six variants and around four weeks from the question to the data analysis.

What makes a robust pre-launch test

First, a clear decision question: which decision should the test support, and what happens with which result? Second, variants that differ only in the attribute being decided on. Third, a measurement point as close as possible to the later decision. Fourth, a benchmark, such as an existing offer, so that a value can be put into context. And fifth, a sample size and stopping rule defined in advance, so that the result does not depend on when you look.

These criteria apply to every form, whether survey or behavioural test. They determine whether a result holds up in the steering committee.

Example

A consumer goods manufacturer has rated three concepts for a refillable shower gel in a concept test, and two are tied. Before the production decision, both concepts run in a Painted Door Test at an identical price.

Result: concept A achieves 4.2 % measured purchase intent, concept B 2.9 %. In the concept test, both were tied. The decision for A now rests on observed behaviour.

How it differs

Pre-launch test is the umbrella term, and the Painted Door Test is one form of it. Post-launch optimisation, for example via A/B tests, requires an available product and answers the question 'how', not 'whether'. A soft launch is already a real introduction with sales, just on a smaller scale.

Limitations

No pre-launch test reflects all factors of the later market: distribution, competitive response, word of mouth and repeat purchase only arise after the launch. Survey-based forms are additionally subject to hypothetical bias. Behavioural tests measure the first decision in the online environment and do not fit when the purchase decision is made at the shelf in retail, in a B2B buying committee or on comparison portals.