How do you measure purchase intent for a new product?
TL;DR
- Ask about one product, at one price, within a set time window. Vague questions get vague answers.
- Read the top box ("definitely will buy") and the top two boxes separately, and always against a control.
- Research on 40 studies found intent tracks purchases better for existing products than new ones, and better when people compare options than when they see one.
- Neon Blue returns intent, the reason behind it and the likely source of volume, by segment, labeled as simulated. Run your first study free
Neon Blue builds simulations of human behavior. You measure purchase intent by showing people a specific product at a specific price and asking how likely they are to buy it, usually on a 5-point scale from "definitely will not" to "definitely will". The score is useful for comparing options and spotting weak ideas. It is not a sales forecast on its own, and new products are exactly where stated intent drifts furthest from real buying. Neon Blue measures intent on a simulated audience of your buyers, with the follow-up questions that make the number easier to trust.
What is the standard purchase intent question?
"If this product were available at [store type] for $[price], how likely would you be to buy it within 4 weeks?"
- Definitely would buy
- Probably would buy
- Might or might not buy
- Probably would not buy
- Definitely would not buy
Three parts of that wording matter. The product description must match what will be on shelf or screen. The price must be the real price. The time window keeps people from answering "someday". The same question sits inside a full concept test questionnaire, after liking and believability.
When does stated purchase intent predict real sales?
In 2007, Vicki Morwitz, Joel Steckel and Alok Gupta looked at this in the International Journal of Forecasting. Across a wide range of intent studies, they found intentions were more closely linked to later purchases:
- for existing products than for new ones,
- for durable goods than for everyday consumables,
- over short time horizons rather than long ones,
- when people were asked about a specific brand or model, not a whole category,
- when the outcome was trial rather than total market sales,
- when intent was collected by comparing options rather than one product alone.
For a new consumable product, the very thing most concept tests measure, nearly every factor works against you. That is why teams read intent relatively, against a control or a past launch, rather than as an absolute number.
How do you make a purchase intent score more useful?
- Add a comparison. Show your concept next to what people buy today, or ask a forced choice after the monadic read. Comparative intent was the stronger signal in the 2007 work.
- Ask what it replaces. "If you bought this, what would you buy less of?" separates new volume from cannibalized volume.
- Ask for a reason. A "probably will buy" with "only if it's on promotion" is a different customer from one who names a real need.
- Read by segment. A total of 30% top-two box can be 60% in one segment and 10% in another; set the cuts before you field the study.
- Vary the price. Intent at one price says little about elasticity. A short price ladder, as in measuring willingness to pay, shows where intent breaks.
What does a purchase intent study look like on a simulated audience?
Hypothetical: a frozen food brand has a new high-protein breakfast line and two possible prices.
- Audience: US adults who bought frozen breakfast items in the past month, split by shoppers who track protein and those who don't, and by household size.
- Design: monadic concept read at each price, the brand's current breakfast line as a control, then a forced choice against two products shoppers already buy.
- Readout: top-box and top-two intent by segment and price, the product it would replace, and reasons in shoppers' words.
Example output layout (illustrative, not a real result)
| Segment | Top-box at lower price | Top-box at higher price | Versus control | Would replace (coded) |
|---|---|---|---|---|
| Protein trackers | __% | __% | Above / level / below | Coded product type |
| Other shoppers | __% | __% | Above / level / below | Coded product type |
| Households of 3+ | __% | __% | Above / level / below | Coded product type |
If intent is strong among protein trackers but most of it would come from the brand's own line, the launch adds less than the headline score suggests. That trade-off belongs in the go or no-go call.
What are the limits of purchase intent research?
- People don't account for shelf competition, stockouts or forgetting. In-market trial will be lower than stated intent.
- A simulated read helps rank options and expose weak ones. For a large launch, add live interviews with real shoppers on the finalist and keep the sources separate.
- For products people have never seen a version of, describe the job it does, not just the features. Getting that description right is its own study, covered in positioning and value proposition testing.
If you also need to know how much people would pay, not just whether they'd buy, see measuring willingness to pay.
Questions about measuring purchase intent
What is a good purchase intent score?
There is no universal cut-off. A score means something only against a control: your past launches, your current best-seller, or another concept in the same study.
What is top-box purchase intent?
The share of people who choose the highest answer, usually "definitely would buy". Many teams weight it more heavily than "probably would buy", which tends to overstate trial.
Should purchase intent be asked before or after price?
After. Show the product, ask liking and believability, then ask intent at the stated price so the price is part of the decision.
How does Neon Blue measure purchase intent?
Neon Blue fields your intent questions on a simulated audience of your target buyers, with comparison, replacement and reason follow-ups, and labels every result as simulated. Start a free first study.