A decision is not made
in the moment.
Every decision is the outcome of everything that came before, and of the people and systems around the decision-maker. What people do is the real decision; what people say is the intended one. We set one against the other, and use behavioural science to explain the difference. That difference is the clearest clue to real thinking, and shows where the surrounding systems are working, or not.
What people value and believe shapes everything that follows, and no survey reaches values and beliefs directly. The way in is to listen to what people say, then watch what people do. Across 126 studies, self-descriptions explained about one twentieth of what automatic reactions showed.1
What people tell you depends on who is asking, what answer seems wanted, and what the speaker hopes to gain. Useful, and never the whole picture.2
What people actually start, skip, delay, abandon and repeat. Every number you report comes from here. Across eight studies that checked actions against stated intentions, the median who followed through was 47 in a hundred.3
The processes, technology and incentives between intention and action decide which of the three wins. Change what surrounds the decision, and people act on existing intentions, without being asked to think differently. That is the path of least resistance, and the place we work.4 Where something has to be hard, we say so, set honest expectations and take care over the difficult parts. And every change has to work for the person and for whoever pays, or the change won't last.
Sources
- Nisbett and Wilson, “Telling more than we can know”, Psychological Review, 1977. Greenwald and Banaji, “Implicit social cognition”, Psychological Review, 1995. Charlesworth and Banaji, “Patterns of implicit and explicit attitudes II”, American Psychologist, 2021. Hofmann and colleagues, “A meta-analysis on the correlation between the Implicit Association Test and explicit self-report measures”, Personality and Social Psychology Bulletin, 2005: 126 studies, mean correlation 0.24.
- Krumpal, “Determinants of social desirability bias in sensitive surveys”, Quality & Quantity, 2013. Larson, “Controlling social desirability bias”, International Journal of Market Research, 2019. Tourangeau and Yan, “Sensitive questions in surveys”, Psychological Bulletin, 2007. Murphy and colleagues, “A meta-analysis of hypothetical bias in stated preference valuation”, Environmental and Resource Economics, 2005: median ratio of stated to actual willingness to pay 1.35; List and Gallet, 2001 found factors up to three.
- Our compilation. Eight studies that asked people about a planned action, then checked. The share of stated intentions that were carried out: Sheeran, “Intention-behavior relations”, European Review of Social Psychology, 2002, 422 studies and 82,107 people, mostly health: a median 53 in a hundred; Rhodes and de Bruijn, “How big is the physical activity intention-behaviour gap?”, British Journal of Health Psychology, 2013, ten studies, 3,899 people: 54 in a hundred (42 per cent of the sample were intenders who acted, 36 per cent intenders who did not); DellaVigna and Malmendier, “Paying not to go to the gym”, American Economic Review, 2006, 7,752 members: forecast 9.5 visits a month, actual 4.17, so 44 in a hundred of the promised visits; Harris, Maurer and Lurie, “Do people who intend to get a flu shot actually get one?”, Journal of General Internal Medicine, 2009, 1,527 adults: just over half; Rogers and Aida, “Vote self-prediction hardly predicts who will vote”, American Politics Research, 2014, 29,403 people across three elections: fewer than half of the people who predicted voting went on to vote; White, Hardisty and Habib, “The elusive green consumer”, Harvard Business Review, 2019: 65 per cent state a wish to buy from sustainable brands, 26 per cent do, so 40 in a hundred; Ganczak and colleagues, “Does vaccinating against influenza in a given epidemic season have an impact on vaccination in the next season?”, International Journal of Environmental Research and Public Health, 2022, a Polish follow-up study: of patients willing to have the next season’s flu vaccine, 31.9 per cent had the vaccine; Cowe and Williams, Who are the ethical consumers?, Co-operative Bank, 2000, the “30:3” gap, 30 per cent say and 3 per cent buy, so 10 in a hundred, as reported in Bray, Johns and Kilburn, Journal of Business Ethics, 2011. Range 10 to 54; counting the two “about half” results as 50, the median of the eight is 47. Background: Webb and Sheeran, “Does changing behavioral intentions engender behavior change?”, Psychological Bulletin, 2006: across 47 experiments, a medium-to-large change in intention (0.66) produced a small-to-medium change in behaviour (0.36). Sheeran and Webb, “The intention-behavior gap”, Social and Personality Psychology Compass, 2016. Conner and Norman, “Understanding the intention-behavior gap”, Frontiers in Psychology, 2022. Stirratt and colleagues, “Self-report measures of medication adherence behavior”, Translational Behavioral Medicine, 2015.
- Fogg, “A behavior model for persuasive design”, Persuasive ’09, 2009: behaviour happens when motivation, ability and a prompt meet. Michie, van Stralen and West, “The behaviour change wheel”, Implementation Science, 2011. DellaVigna and Linos, “RCTs to scale: comprehensive evidence from two nudge units”, Econometrica, 2022. Sheeran and Orbell, “Using implementation intentions to increase attendance for cervical cancer screening”, Health Psychology, 2000: attendance 69 per cent without a when-and-where plan, 92 per cent with one. Gollwitzer and Sheeran, “Implementation intentions and goal achievement: a meta-analysis”, Advances in Experimental Social Psychology, 2006: 94 tests, effect size 0.65.
The decision
stays yours.
We're not experts in your business. You are. We're experts in how people and systems behave, including the system you work inside.
So we don't hand over a recommendation and leave. We bring the plan to the point of decision: the questions that have to be answered, the evidence against each one, and the person whose name goes on the plan.
Then you decide, and you act. The argument has already happened on paper, where argument is cheap.
Most of what we use is established practice, and we say so: behavioural science, user experience, service design. The Uservox method connects all three. We work alongside your teams and the agencies you already have, in the room and between the rooms, until the plan has decisions rather than intentions.
When the same need keeps appearing and no single team can meet that need alone, we build a product once. XMap® maps what people actually do: customers, buyers, patients, clinicians. Tested hardest in pharma, where the evidence bar is highest; nothing about XMap is specific to pharma.
We make visible what is decided invisibly.
Wherever plans
outrun decisions.
Industries differ. People don't, and nor do the places where a plan quietly stops being a decision. Global brands, from the United Kingdom, since 2011.
Worked with teams at
Risk says one thing, product another, and compliance signs off on a version neither quite recognises. The disagreement is real and worth having, but never gets resolved on paper. We make the choice explicit: what has to be true, what evidence supports the choice, who signs. Before the work reaches the backlog, not after.
Every studio has a roadmap and a room where the strongest opinion wins. The argument is usually not about the data, but about which question the data was meant to answer. We get the question agreed first, the evidence second and a named owner third. The next strong opinion then has something to argue with.
A loyalty programme is a set of decisions nobody owns together. Marketing sets the mechanic, finance sets the margin, partners set the terms, and each is defensible alone. We put those decisions in one place with the trade-offs visible, and hand back a set of decisions rather than a set of assumptions.
Launch, lifecycle, access, a new indication: the pattern holds. Medical, market access, marketing and country teams each mean something different by "agreed", and the plan moves forward on the ambiguity. We turn plans into decisions rather than documents: what must be true, what evidence supports each decision, whose name is against the decision.
Trading, digital and store operations each have a plan, and each quietly assumes the other two will absorb the cost. Nothing looks wrong until the three plans meet in the same quarter. We get the decisions onto one page, with the question, the evidence and the owner, before the budget commits to all three.
Adoption is everyone's objective and nobody's decision. Product owns the build, commercial owns the sell, service owns the aftermath, and the thing that actually stops people falls between the three teams. We make that a decision with an owner, rather than a theme in three separate plans.
Human first.
AI forward.
Nobody decides alone. Every person sits inside systems of other people, places and technology, and those systems shape behaviour as surely as any belief. The brain is complex. Artificial intelligence is complex. We understand both, and use that understanding to make the complex simple.
heard
Uservox means the user's voice. The user is whoever the system is failing: the customer you serve, or the person in your team who has to make the system work. Everything else is shaped around that, never the other way round.
understood
Technology runs through how a business operates and how customers connect. We know technology well enough to design with, not around, and to keep technology a tool in service of people.
used with intent
Companies are unsure about AI. We are not. AI learned everything from people, so we read AI the way we read people: by actions, not words. We put AI to work at a scale people alone can't reach, with a person deciding what's true and what matters.
made simple
Simple means something you can use on Monday: one picture your whole team agrees on, a short list of what to change first, and evidence anyone can open. No one should need us in the room to read the result.
Talk to us.
Tell us what you're trying to decide. We'll tell you what would have to be true for the answer to be yes.
hello@uservox.com
UK +44 (0)1935 676980
Thank you.
A person will read your message and reply, usually within one working day.
A copy is on the way to .