Last updated: 18 September 2026
Quick Answer: A consumer insight states an observed behavior, the motivation behind it, and what a brand should do. IHOP's 2018 example: to earn business after breakfast, prove that IHOP takes its burgers as seriously as its pancakes. Without motivation and implication, it is a finding, not an insight.
In late 2009 Domino's brand tracker found that speed had stopped being a reason anyone bought pizza, and taste ranked it last among the big three chains. Most of what circulates internally as an insight is a finding with a motivational verb bolted on: whether a colleague could have written the opposite sentence from the same data.
What Three Parts Does Every Consumer Insight Need?
An insight has three parts, none optional:
- Observation is what the evidence shows people doing.
- Motivation is why, in human terms rather than a category label.
- Implication is what the brand should do differently.
The ICC/ESOMAR Code frames research as fact-based evidence; the motivation clause makes it an insight, checkable against one template: because [observed behavior], which happens because [motivation], we should [implication]. A weak statement still clears review: younger shoppers seek authenticity, which names no behavior, no motive, no action. Rewritten: because 18 to 24 year olds open a brand's returns page before its product page, expecting disappointment, we should move returns terms onto the product page.
The three parts hold regardless of what the insight is about. A category insight, a brand insight and a cultural insight, covered further down, all still need an observation, a human motivation and an implication, or they are findings wearing an insight's grammar.
Keep the evidence attached: who, how, how many, when, per AAPOR's disclosure standards. A separate discipline governs turning a finding into a decision.
Consumer Insight Examples From Real Campaigns
Seven worked examples follow, each traceable to a published, datable source: two documented by an independent advertising-effectiveness awards body, one drawn from a brand's own campaign research as engaged with in independent peer-reviewed literature, and four measured directly by researchers, universities or public health and policy teams with no stake in a brand outcome.
IHOP, 2018
Observation: indulgent non-breakfast occasions were 19 percent of restaurant occasions, only 6 percent of IHOP's. Insight, per the ARF award case study: prove that IHOP takes our burgers as seriously as our pancakes. Action: it renamed itself IHOb for a summer; burger sales quadrupled.
Domino's, 2009 to 2010
Observation: four straight years of falling sales, last on taste among the big three. Insight: fifty years of speed advertising had bought a reputation it could no longer trade on. Action: it reformulated crust, sauce and cheese, then built the Pizza Turnaround campaign around the criticism, documented in a 2011 case study.
UK Organ Donation, 2013
Observation: nine in ten supported donation; fewer than one in three were registered. Insight: the gap was the moment and framing, not belief; reciprocity was the lever. Action: a trial across a million people tested eight variants; the winner, asking whether the reader would accept a transplant, was projected to add roughly 96,000 registrations a year.
England Stoptober, 2012
Observation: smokers who wanted to quit were asked to commit forever. Insight: a bounded, dated, collective attempt is something a smoker can agree to this week. Action: Stoptober asked for 28 days in October, credited with 350,000 extra quit attempts in year one, though a six-year evaluation found the effect varied yearly with budget.
SuperAmma Handwashing, 2011 to 2012
Observation: households in rural Andhra Pradesh already knew dirty hands spread disease; handwashing with soap at the moments that mattered stayed rare anyway, per a cluster-randomized process evaluation. Insight: a population that already has the health facts doesn't need more of them; nurture, status, disgust and social norms move behavior that information alone does not.
Action: the SuperAmma campaign, run by St John's Research Institute and the London School of Hygiene and Tropical Medicine, built animated films, street theatre, school-based sessions, public pledging ceremonies and visual reminders on those four levers, with no disease messaging at all. Handwashing at key occasions reached 37 percent of intervention households at six months against 6 percent of control households, up from 19 percent against 4 percent at six weeks, across seven intervention villages.
Dove Self-Esteem Project, 2022
Observation: Dove's own 2022 research, distributed through PR Newswire, reported that seven in ten girls who unfollowed accounts pushing idealized beauty content said they felt better about themselves afterward. Insight: filtering the feed, not another message about self-acceptance, was the one action inside a teenager's own control.
Action: Dove launched #DetoxYourFeed that year, continuing the Dove Self-Esteem Project's push against idealized imagery. A peer-reviewed 2025 study analyzing more than 500 public comments on the Self-Esteem Project's own video posts found the audience read them as authentic rather than promotional, and generally positive in tone.
UK Tax Reminder Letters, 2014
Observation: most people who owed overdue tax already knew paying was required; standard reminder letters still treated non-payment as a private matter between one taxpayer and the tax office. Insight: a struggling payer breaks a rule; a payer told that the vast majority of people like them have already paid breaks a norm, and norms move people that a penalty warning alone does not.
Action: HM Revenue and Customs and the Cabinet Office's Behavioural Insights Team tested five reminder messages across 100,000 taxpayers; the version stating that the recipient was in the small minority who had not yet paid produced the largest effect, a 5.1 percentage point rise in payment worth an estimated £2.37 million in a 23-day window. A second trial across nearly 120,000 taxpayers replicated the finding.
How Do Category, Brand and Shopper Insights Differ?
Insights are also grouped by what they're about, not just how they were gathered. A category insight explains why people buy the kind of thing at all; a brand insight explains how they place one brand against its rivals; a shopper insight explains what happens at the moment of choice, which can contradict what the same person says in an interview.
A fourth, cultural insight, explains a shared norm the other three sit inside, and it is the one most often mistaken for common sense because everyone inside the culture already assumes it.
- Category insight: true for every brand in the category, and the reason a market exists before any single brand competes inside it.
- Brand insight: can be true of one competitor and false of its closest lookalike, which is what makes it specific enough to act on.
- Shopper insight: about the pick at the shelf or the checkout, where stated preference and actual behavior diverge; a survey answer is not a purchase.
- Cultural insight: about a shared norm, identity or belief, and the hardest to see from inside it.
This is a grouping by scope, not the descriptive, diagnostic, predictive and prescriptive ladder analytics teams use for data maturity, which does not fit consumer insights at all (see the FAQ below).
Which Type Explains Which Campaign
Several of the campaigns above are cultural insights rather than product insights. Stoptober worked because a bounded, collective attempt is a social contract a smoker can accept, not because of anything specific to cigarettes. The tax reminder letters worked by making non-payment a visible departure from what almost everyone else was already doing, not by explaining the penalty structure more clearly.
IHOP's insight sits closer to brand: burgers were available everywhere already, so the problem was never the category, only where IHOP fit inside it. Domino's sits there too, since a taste ranking is a claim about one chain against its named rivals, not about pizza as a category. SuperAmma is a hybrid of the two: the category insight, that soap already gets used somewhere in every home, mattered less than the cultural one, that status and nurture rather than hygiene facts decide when it gets used at the sink. A category insight and a cultural insight call for different evidence, which is why the sourcing question below has no single right answer.
Where Do Consumer Insights Actually Come From?
Every example rests on evidence that limits its claim; the table runs alphabetically by the first column, not as a recommendation.
| Evidence source | What it observes directly | What it can support | What it cannot tell you |
|---|---|---|---|
| Behavioral and transaction data | What people did, at scale | Where behavior shifts, and for whom | Why, or what was considered and rejected |
| Ethnography and in-context observation | Behavior in its real setting, workarounds included | Motives people cannot articulate on request | Whether the pattern holds beyond those homes |
| Field experiments and A/B tests | Whether one change moves one outcome | Causal claims about the version tested | Why the losing version lost |
| In-depth interviews | Reasoning, contradictions, decision order | Motivation, and the language people use for it | Size, share, or price sensitivity |
| Panel and syndicated datasets | Category trends across many brands | Context and benchmarks for your own numbers | Anything specific to your customers |
| Social listening and reviews | Unprompted language from people who posted | Vocabulary, complaint clusters, timing | What the silent majority thinks |
| Surveys at scale | Stated attitudes and claimed behavior | Incidence, sizing, segment differences | Motives respondents cannot or will not report |
What Each Method Actually Catches
The table above is a starting point, not a checklist to complete. Ethnography and in-context observation earn their place because of the say-do gap: someone can describe their decision process accurately in an interview and still be watched doing something else entirely at home, and the two are not a contradiction so much as two different questions answered honestly.
Field experiments and A/B tests are the only method on the list that supports a causal claim rather than a correlation, which is why the tax reminder letters above settle which message worked without needing anyone to explain why in their own words. That strength is also the limit: a field experiment shows that the minority-norm letter beat the basic one, not what a taxpayer was thinking when it worked. In-depth interviews fill that exact gap, at the cost of a sample too small to size anything from.
Behavioral and transaction data show what happened at a scale no interview sample reaches, but a purchase record cannot distinguish a shopper who chose a brand from one who could not find the alternative. None of these substitutes for another; they answer different questions, and an insight resting on only one method is only as strong as that method's blind spot.
Choosing a Platform for a Mid Sized Team
A motive without a size is a hypothesis; a size without a motive, a dashboard. AAPOR's best practices cover both. For a mid sized insights team choosing a platform, order beats the shortlist:
- A named decision with an owner: what it informs, by when, who answers.
- A standing route to the right people, not whoever a panel returns this month.
- One place the evidence lives, via a searchable evidence base.
Personas Built From Transcripts, Not Clusters
A transcript-based persona carries the order a person decided; one inferred from clusters cannot: personas built from real interviews check back against a source.
How Do You Test an Insight Before You Build On It?
Test it the way the organ donation team did:
- The opposite test. Write the contradicting insight from the same data; if it's also plausible, the evidence hasn't chosen.
- The small-stakes test. Try the cheapest artifact available: the organ donation trial above ran eight variants past a million people, and one reduced sign-ups.
- The scale test. Effects shrink at scale: two hygiene campaigns across 181 rural wards in Tanzania moved intermediate behaviors but produced no detectable child health effect; Stoptober's evaluation, above, varied with budget.
- The competing-message test. Don't test your best guess against nothing; test it against your other plausible guesses. The tax reminder letters above tried five messages at once and found a nearly four-fold gap between the weakest and the strongest, a difference a single before-and-after test would never have surfaced.
Who Is Missing From the Evidence Behind Your Insight?
An insight is only as wide as the evidence behind it. ITU's Facts and Figures 2024 finds 93 percent online in high-income countries, 27 percent low-income; an online panel there samples the connected quarter. Pew's mobile fact sheet, surveying 5,022 US adults in 2025, finds 91 percent ownership via an address-based frame reaching what convenience samples skip.
Three questions expose the gap:
- Which channel produced the evidence, and who doesn't use it.
- Which language it was collected in, and who can't explain a motive.
- Who was screened out for convenience, not relevance.
Sample validity turns on the same three. Channel breadth is the lever: Alchemic publishes 57+ languages including Spanish, Arabic and Mandarin, and runs managed fieldwork or bring your own across fourteen markets including the USA and the UK, so end-to-end consumer research at scale is not restricted to a browser link.
Where Insight Writing Goes Wrong
The most common failure is restatement: adding a verb like crave or value to an observation sounds like an insight while saying nothing new. Shoppers abandon carts at the shipping step becomes shoppers crave transparency in shipping, instructing nobody.
Three more failures compound it:
- Hindsight bias. Every example above is told backward from a result, making each decision look inevitable; equally plausible insights that failed are never written up, so case studies teach only survivors.
- Category nouns standing in for motives. Authenticity, convenience, community label motives without being motives; unfalsifiable, they survive review.
- Treating a cultural insight as universal. SuperAmma's nurture-and-status framing worked in rural Andhra Pradesh; the tax letters' minority-norm framing worked on UK taxpayers already primed to see paying as the default. Neither finding transfers to a market where the starting norm runs the other way, and an insight that does not name the population it was found in invites exactly that transfer.
Some decisions need no insight: where the choice is already made, research that follows produces a document, not a decision.