For funds
Trial and repeat for a launch: how simulated consumer surveys estimate it
TL;DR
- Trial is the first purchase in the launch window. Repeat is the second purchase among triers. Intent to "try someday" is not enough for either line.
- A Neon Blue study asks likely buyers whether they would try, when, whether they would buy again if they liked it, how soon, and what they would stop buying.
- The output is trial share, repeat intent among triers, timing and cannibalization by segment. It is a labeled input, not a stock call.
- Results come as a report. An ongoing data stream and an API are in build.
Neon Blue builds simulated audiences of a company's customers and category buyers, AI models that answer the way a defined group of people would, grounded in Neon Blue's own paid studies and AI-moderated interviews with real people. For funds, a trial-and-repeat study answers the two rates that decide whether a launch holds after the first shelf reset: who tries, and who buys again.
What is trial and repeat, and why does the model need both?
Trial is the share of the target audience that makes a first purchase in a defined launch window. Repeat is the share of those triers who make a second purchase within a defined follow-up window (often four to twelve weeks for CPG). Together they drive:
- Units in the launch quarter. Trial times units per trip.
- Units after the first reset. Repeat is what keeps the SKU on shelf after launch spend fades.
- Cannibalization. A repeat that replaces the company's own older line is not net growth.
A purchase intent read for a new product tells you who likes the idea. Unit demand for a launch turns buyers into units. Trial-and-repeat sits between them: it separates first buy from second buy so the model does not treat every intending buyer as an ongoing user.
How does a Neon Blue study estimate trial and repeat?
- Define the likely triers. Current brand buyers, category buyers from rivals, and category non-buyers. Each cell converts differently.
- Show the product as it will be sold. Real description, real price, real channel. Leaving out price inflates trial.
- Ask trial and timing. Would you buy once in the launch window, later, or never.
- Ask repeat conditional on liking it. If you tried it and liked it, would you buy again in the next four to twelve weeks, how many times, and at what pace.
- Ask what it replaces. An older SKU from the same brand, a rival, or new spend.
- Combine by segment. Trial share times segment size for first buys; repeat share among triers for the second-buy line.
Stated repeat runs high when the second buy is far away. The useful part is the shape: which segments say they would repeat soon, and how much of that repeat is cannibalized. How simulated audiences work explains how the audience is grounded.
What questions go into a trial-and-repeat study?
| Question | Model input | Why it matters |
|---|---|---|
| Would you buy this once in the launch window? | Trial share | First-buy line |
| If you bought it and liked it, would you buy again in the next four weeks? | Repeat intent among triers | Post-reset line |
| How many times would you buy again in that window? | Repeat occasions | Units after trial |
| What would you buy less of? | Cannibalization and substitution | Net vs gross units |
| Where would you buy it? | Channel mix | Distribution assumptions |
| What would stop a second buy? | Failure modes | Price, taste, availability, rival |
What does a trial-and-repeat study look like?
Hypothetical: a staples company just launched a protein snack at a stated price in grocery and club. The analyst wants trial and repeat inputs before the next quarter.
- Study question: "At this price, would you buy this protein snack at least once in the next eight weeks? If you tried it and liked it, would you buy it again in the four weeks after that first purchase, and would that replace another snack you buy now?"
- Audience: category snack buyers in the last ninety days, split by brand buyers vs rival buyers vs category-light buyers, household income and club vs grocery channel.
- Follow-ups: occasions per month if they repeat, and which SKU they would cut.
How the analyst reads it: high trial with low repeat among rival buyers is a launch spike that fades. High repeat that mostly replaces the company's own line is units without net growth. Both go into the model labeled as simulated.
Want this run on your own name? start a study.
When should a trial-and-repeat study run against the reporting calendar?
Run it when the SKU is announced or first on shelf, again four to eight weeks later among people who say they tried it (or a matched sample) to refresh repeat intent, and ahead of the first full quarter that includes the launch. The timing logic matches pre-earnings consumer checks. Staples and retail desks often keep consumer staples and retail open while they wire the inputs.
Which sectors is this most useful in?
- Consumer staples and food: snacks, drinks, beauty and household launches where second purchase decides shelf life.
- Consumer electronics and gaming: first device or title buy vs follow-on attach and season pass. See unit demand for a launch.
- Consumer internet and media: first paid month vs renew after the promo window. See subscriber churn.
- Private equity diligence: whether a target's new line has real repeat, not only launch trial. See commercial diligence.
What are the limits of a survey-based trial-and-repeat read?
- People overstate repeat when the second buy is weeks away. Prefer short follow-up windows and read repeat next to what they would cut.
- Distribution and out-of-stocks change realized trial without changing intent.
- A small group of heavy category buyers can dominate units. Weight by category spend, not only by headcount.
- Treat the read as one directional input, not a settled number.
Questions about estimating trial and repeat
What is the difference between trial and purchase intent?
Purchase intent is interest in buying. Trial is a first purchase inside a dated window. You can have high intent and still miss the launch quarter on timing.
Why not fold repeat into a single unit forecast?
Because launch spend and shelf resets hit trial and repeat on different schedules. Separating them shows whether the SKU survives after the first push.
Does high trial with low repeat still help the print?
Sometimes in the launch quarter. It usually fails the quarters after, which is why the repeat line belongs in the model early.
Does this predict a company's stock?
No. It gives trial and repeat inputs by segment for your unit assumptions. Neon Blue does not make stock calls or give price targets.
How do I start a study?
Bring the SKU, the price, the launch window and the trial or repeat assumption you trust least. Start a study.