


Purchase intent is one of the most frequently used measures in market research. What matters is how it is measured: as a statement in a survey or as an observed decision in front of a concrete offer.
Whether it is a new product, a higher price or a different product variant: many business cases start with an assumption about how many people would buy. This assumption often comes from a question such as "How likely are you to buy this product?" with a scale from "definitely not" to "definitely".
Research shows that these statements are related to later behaviour, but only to a limited extent. Intentions are acted upon in roughly half of all cases, and for new products the relationship between stated purchase intent and later sales is weaker than for existing ones. So precisely where the most is invested, the stated measure is at its most uncertain.
On top of that, the survey lacks price comparison, alternatives and the moment in which you actually decide for or against an offer. Planning on stated purchase intent means planning on a statement without consequence.
Stated purchase intent: respondents rate a concept on a scale. Often the share of the top two answer categories is reported. This is fast, easy to compare across concepts and, together with open questions, provides valuable insight into motives.
Measured purchase intent: real people see a realistic offer in their familiar online environment, with a price and product description, and decide whether to click the buy button. What counts is the last binding click before the reveal. Horizon measures purchase intent in this way with Painted Door Tests: nothing is sold, and anyone who chooses the offer is then told transparently that it was a test.
A manufacturer is considering offering a cordless kitchen machine in two versions: a basic version and a version with app recipes at a higher price. In a survey, 41% choose one of the top two categories for the app version and 36% for the basic version.
In the behavioural test, each person sees only one of the two offer pages. Of the visitors to the basic version, 3.4% click "Add to basket", for the app version 2.6%. The ranking flips: at the moment of the offer, the price premium weighs more heavily than in the survey.
The two sets of figures are not directly comparable because they rest on different bases. What is meaningful is the comparison of variants within the same method.
Willingness to pay asks how much someone would pay at most; purchase intent asks whether someone chooses an offer at a given price. Sign-up intent is the counterpart for tariffs and contracts, for example in insurance, energy or mobile. Purchase intent is not a purchase: in a Painted Door Test it is measured without a contract being created or anything being paid.
Measured purchase intent is not sales either. It shows how people decide in front of an offer, not how often they later buy again, return the product or tell others about it. It mainly answers the questions "which variant" and "how strongly" in comparison, less the question of exact market volume.
Stated purchase intent remains useful for pre-sorting many concepts and understanding the why. For the few variants that are actually invested in, measured purchase intent complements these statements with observed behaviour.
Sheeran & Webb 2016: Intentions are translated into action in roughly half of all cases. The Intention-Behavior Gap, Social and Personality Psychology Compass 10(9). Source
Morwitz, Steckel & Gupta 2007: Stated purchase intent is related to later sales, but considerably more weakly for new products than for existing ones. Meta-analysis of purchase intent and later sales, International Journal of Forecasting 23(3). Source
Chandon, Morwitz & Reinartz 2005: Among surveyed customers, the relationship between intention and purchase is 58% stronger than it would be without a survey: asking itself inflates the measured validity. Study of self-generated validity effects in survey research, Journal of Marketing 69(2). Source
No. In the behavioural test nothing is sold and nothing is paid. What is measured is whether a person decides to buy in response to a realistic offer.
It is related to behaviour, but only to a limited extent, and more weakly for new products than for existing ones. For investment decisions, an additional behavioural data point is therefore worthwhile.
The share of visitors to an offer page who make the last binding click before the reveal, in other words the measured purchase intent per variant.
Bring your decision question, and we will outline a possible test design.
You will speak with Daniel Putsche
Founder & CEO, 30 minutes
Read more