Skip to content
Get in touch

Where in an app or a checkout do people decide to stay or leave?

Every session reads as a four-phase story, and the alignment phase, when people decide, is where clarity, lower risk, and signs of fit matter most, because a person who leaves there rarely comes back.

A.A., Dartmouth

ActiveInteraction
HCIChoice architectureTrust

Hypotheses.

What the paper argues, written so it can be tested.

  1. H1

    On a decision page, putting the total cost, the key terms, and a specific way out in the same view as the main button lowers abandonment more than adding persuasion earlier in the session.

  2. H2

    People who leave during the alignment phase come back to finish less often than people who leave in any other phase, so changes there move completion rates the most.

The model.

A session told as a four-phase story. The hook gets attention, exploration narrows the options, alignment weighs fit against risk, and resolution carries out the choice. Leaving at alignment is the hardest exit to undo: people who walk away there rarely come back.

A SESSION AS A FOUR-PHASE STORYHookGets attentionSome leaveExplorationNarrows optionsSome leaveAlignmentWeighs fit and riskPeople who leave hererarely come backResolutionFollows throughSome leaveWHAT DECIDES ALIGNMENTKEEPS PEOPLEPUSHES PEOPLE OUTA clear offer in the person’s own termsSurprise fees or hidden termsTotal cost and terms in viewToo many similar optionsA specific guarantee or way outDecision fatigue late in a sessionProof from people like themReviews that look too perfectA few curated choicesLate changes that break expectations
A SESSION IN FOUR PHASESHookGets attentionSome leaveExplorationNarrows optionsSome leaveAlignmentWeighs fit and riskFew returnResolutionFollows throughSome leaveKEEPS PEOPLEA clear offer in the person’s own termsTotal cost and terms in viewA specific guarantee or way outProof from people like themA few curated choicesPUSHES PEOPLE OUTSurprise fees or hidden termsToo many similar optionsDecision fatigue late in a sessionReviews that look too perfectLate changes that break expectations
Drawn from the paper. Dashed lines show what the model predicts.

Partial abstract.

Sessions in e-commerce, onboarding, and conversational agents are modeled as a four-phase narrative arc. The paper proposes that the alignment phase, where a person weighs fit and risk, matters most, since drop-off there is the least reversible.

The Four-Phase User Experience Narrative and the Decisive Role of Alignment in HCI Sessions

Key sources

  • Fogg, B. J. (2003). Persuasive technology: Using computers to change what we think and do.
  • Deck, C. A., and Jahedi, S. (2015). The effect of cognitive load on economic decision making: A survey and new experiments.
  • Long, X., et al. (2024). The choice overload effect in online recommender systems.

How we’ll test it.

No results yet. This is the plan: the data, the measure, and what would count as support.

Awaiting data

The data

Session logs from online checkouts and software sign-ups, tagged by phase, with versions of the decision page that change what sits next to the main button.

The measure

Drop-off in each phase, how often people who left come back to finish, and completion rates for each version of the decision page.

What would count as support

People who leave at alignment come back least often, and the page with costs, terms, and guarantees in view cuts drop-off there more than any change to another phase.

How it connects to Pulp.

Loosethreads cues meet people at the moment they decide.

Also in Interaction

Doing work like this?

Students from any school or field, and mentors from academia or industry: get in touch.

Get involved

Pulp

Color mode

Roll into a random page on Pulp. A few hidden pieces can only be accessed by taking a chance.

Get in touch

Color mode

Kept on this device only.

What are you trying to understand?

So we know who to write to.
An address we can reply to, like name@example.com.
A question, a dataset, an argument that keeps going in circles.

Get in touch
Roll into a random page on Pulp. A few hidden pieces can only be accessed by taking a chance.