What the expectation-disconfirmation model says
Customers do not judge an experience in isolation. They arrive with an expectation of what it will be like, built from adverts, product pages, listings, reviews, what a salesperson or support agent said, past experience and what friends told them. The experience is then compared with that expectation. "Disconfirmation" is the name for the gap: a shortfall is negative disconfirmation, an experience better than expected is positive, and an experience that matches is confirmation.
The consequence for marketing is uncomfortable. A business controls a large share of the expectation through its own claims, and the pressure on every claim is to raise the expectation, because a higher expectation usually wins more attention and more sales. Past a certain point, the same claim starts creating disappointment that the product cannot repair. A software advert that calls installation instant turns a ten-minute install into a let-down; a photograph of a room in its best light makes the room look smaller when the guest walks in.
The guide to checking marketing promises against the experience turns the idea into a recurring audit: collect every promise, find the moment in the journey where the customer tests it, and fix either the step or the claim. The message match guide applies it to the shortest gap of all, between the advert a person clicked and the page they land on. This page describes the model that sits behind both.
Where marketing sets the expectation
Every touchpoint before the purchase adds to the expectation, and each one is tested at a particular moment afterwards. The table pairs the two.
| Touchpoint | The expectation it sets | How disconfirmation shows | What to read |
|---|---|---|---|
| Advert or social post | What the product does and who it is for | Clicks with little engagement on the page | Landing page behaviour after the click, not CTRShare of impressions that result in a click. alone |
| Landing or product page | The specific offer, price and conditions | Abandoned checkouts; questions the page should have answered | Checkout drop-off; pre-purchase questions by topic |
| Marketplace or app store listing | What the first use will look like | Low ratings and reviews describing a different product | Rating trend; review text; early uninstalls |
| Photographs and video | Size, quality, condition and setting | "Not as pictured" complaints and returns | Return RateShare of orders or units sold that customers sent back. and return reasons |
| Sales or support answer | Fit, delivery, compatibility, terms | Complaints that quote what an agent said | Complaint themes matched to the answers given |
| Onboarding emails and first session | How quickly value arrives | Early cancellations; support contacts in the first weeks | Early Churn RateShare of customers at the start of a period who left during it.; first-week contact reasons |
| Price and guarantee wording | What is included and what happens if it does not suit | Disputes over extras, fees or refund terms | Refund and dispute reasons |
Two rows deserve special attention. The app store guide for app installs asks that the screenshots describe what the person will see when they open the app, because a listing that promises more wins the install and loses the user. The guide to answering pre-purchase questions treats an agent's answer as a claim: an answer that promises more than the product delivers creates the complaint or return later, so agents use the approved wording and add no promise the page does not make.
Setting expectations that the experience can meet
Make each claim specific and checkable. "Fast delivery" sets whatever expectation the reader brings; "delivered within the window shown at checkout" sets one the business can keep and the customer can verify. Specific claims are also easier to audit against the experience.
Show what the customer will actually see. Product photographs taken in ordinary light, screenshots of the screen the user meets first, a description of the room as it is: each moves the expectation towards the experience. The value proposition canvas helps here by separating the gains a product really creates from the ones the team would like it to create.
Keep the promise the same along the path. Message MatchHow closely a landing page repeats the promise and wording of the ad or link that sent the visitor. is usually discussed as a conversion technique, but it also keeps the expectation steady: an advert, a page and a confirmation email that say the same thing leave no room for the customer to build a larger expectation than the business intended.
Read the test past the click. The Meta creative testing guide warns that a hook that overpromises wins on clicks and loses on everything after, and reads results on the metric that matches the objective. The same applies outside paid social: a headline, subject line or listing that wins on attention should also be checked against what happened after the purchase.
Close the loop with the after-purchase signals. Complaints, returns, ratings, reviews and early cancellations are where disconfirmation becomes visible. Reading them by the claim or touchpoint they mention, rather than as a single satisfaction figure, shows which promise is responsible.
Worked example (illustrative)
A holiday cottage agency rewrites its listings
Dazzevrin Holiday Cottages, an invented agency letting 38 cottages, reads the 410 guest reviews from its last season against each cottage's listing. Most reviews are positive, but 52 contain a complaint, and the complaints cluster: 31 of them concern just 9 cottages.
In those 9 listings the team finds the same pattern. Wide-angle photographs make small bedrooms look spacious, "sea views" means a glimpse from one upstairs window, and "a short walk to the beach" is a 25-minute walk on a road without a pavement.
The agency rewrites the 9 listings. Photographs are retaken at normal focal length, each bedroom is described with its bed size and floor area, the view is described from the room where it can be seen, and the walk is given in minutes with the route described. Enquiries for 2 of the cottages fall at first.
Over the next season the agency tracks complaints and rebookings by cottage. Complaints about the 9 cottages fall to 8, and rebooking enquiries from previous guests at those cottages rise from 6 to 14.
Limits and criticisms
Expectations are hard to observe. Customers rarely state them in advance, and they shift during the experience as the customer learns more, so the expectation a business reconstructs afterwards is partly a guess.
Raising expectations also raises interest. A bolder claim usually attracts more people, and the model cannot say how far a business should promise before the extra sales are outweighed by the disappointment. That trade-off has to be judged from the business's own after-purchase evidence.
A promise set too low can lose the sale before the experience starts. Underpromising is safe for satisfaction and unsafe for revenue, and a customer who never buys is never disappointed or delighted.
Satisfaction is not the same as buying again. A customer whose expectation was met can still leave for a lower price, a change of circumstance or a competitor that is easier to reach, so satisfaction measures are not a substitute for retention measures.
And the model says nothing about what customers want in the first place. It explains how an experience is judged against an expectation; it does not say which product, offer or experience is worth building.
Related frameworks
Customer journey mapping lays out the touchpoints in the table above in the order the customer meets them, with the thoughts and feelings at each stage, which shows where each expectation is formed and where it is tested. AIDA is a checklist for writing persuasive messages; this model is the counterweight to its desire stage, a reminder that the desire a message builds becomes the expectation the experience must meet. The marketing frameworks library lists the other funnel and journey frameworks by the decision they serve.
Common mistakes
- Judging creative on the click alone. A claim that wins attention can lose the customer after the purchase.
- Vague promises. A claim the customer cannot check lets them supply their own, larger expectation.
- Best-light photographs. Pictures that flatter the product set an expectation the product cannot meet.
- Agents adding promises. An answer in chat or on the phone is a claim, and it is tested like one.
- Reading satisfaction as one number. Complaints and returns read by claim show which promise failed.
- Promising too little to be safe. An expectation set too low loses the sale before it can be met.
- Treating satisfaction as loyalty. A met expectation removes one reason to leave, not all of them.
