Frictions Map

What is the Frictions Map?

The Frictions Map connects recurring UX problems to the exact moments they affect a user's journey, helping teams choose and verify the right response.

Friction is a demand placed on the user. The map does not treat every demand as a defect. For each one, it asks whether the demand is necessary, understandable and timed appropriately.

Use the map when you need to locate and evaluate recurring problems within a specific journey: a checkout, a sign-up, an onboarding path or any flow with a clear goal. It works inside the broader evaluation context of the UX Clarity Framework. It shares TYPENORM's pattern registry with the other frameworks in the family: the 5L UX Onboarding Flow, the Research-to-Design Loop and the UX Clarity Canvas.

The unit of analysis: an expectation gap

Every finding on the map is an expectation gap:

At this point, the user expects X, but encounters Y, which affects their goal by Z.

Record the basis for the expectation. A researcher assumption must not be presented as a participant statement.

One finding can span several screens. A specific case supplies the evidence and context; a pattern describes the recurring problem. Matching a pattern does not by itself establish severity, reach, cause or user behavior.

How to apply the method

1. Define the goal. Specify the user, context, intended outcome, journey boundary and conditions. For onboarding, explicitly define first value. Bounding a journey and choosing one persona is covered in Customer journey mapping 101.

2. Observe the journey. Record what the interface does and link the relevant captures. Separate facts from interpretation. The free journey map template gives you a stage-by-stage frame to record them in.

3. Match patterns. Choose a primary pattern with a short fit rationale. Add related patterns only when they explain distinct aspects. Patterns come from the registry's eight categories:

   • CTX — Context & Continuity
   • CLR — Clarity
   • EFF — Effort
   • DEC — Decision Support
   • ERR — Errors & Recovery
   • TRU — Trust
   • FBK — Feedback & System Status
   • CNV — Conversion Friction

   Seeded patterns start as candidates, and an approved definition is not evidence that a pattern is validated. Keep unmatched findings rather than forcing a fit; they may inform future registry work. Research synthesis helps turn raw observations into findings before you match them.

4. Assess consequences. Use the existing case severity system, with its severity levels (critical, major, minor and cosmetic). Record impact on this task, reach if known, and evidence confidence separately. Unknown reach stays unknown. The heuristic evaluation walkthrough rates severity by cost in the same spirit.

5. Choose a response. Remove, explain, move or preserve the demand, and describe the actual intervention. Draw on the registry's solution suggestions where appropriate.

6. Verify the result. Record an owner, expected effect, verification method and relevant guardrails. The next action may be further investigation when evidence is insufficient. A usability test is one way to check the result.

The map adds no composite score, and several pattern tags on one finding still count as one problem.

The map artifact

Read the map in journey order, then use its findings to prioritize work. A workshop board or a table can show the same information.

Field	Required content
Journey moment	Step or transition, user goal, and relevant conditions
Expectation gap	Expected versus encountered behavior, with the basis for the expectation
Evidence	Specific observation and capture references
Pattern match	Code, name, fit rationale, and current registry maturity
Consequence	Task impact and existing case severity; reach and uncertainty separately
Response	Remove / explain / move / preserve, specific intervention, and rationale
Verification	Owner, expected effect, check, guardrails, and eventual outcome

Response choices

• Remove: eliminate an unnecessary demand.
• Explain: make the purpose, state, consequence or recovery path understandable.
• Move: place a necessary demand at the moment it becomes relevant.
• Preserve: retain useful friction with a documented reason, such as preventing a consequential mistake.

These are decision prompts, not a complete catalog of design solutions. A response may combine them. A protective confirmation is not automatically a defect.

Worked example: Best Buy price discrepancy

This entry comes from the section "The price that moved" in our Best Buy UX teardown. It is a historical observation from the teardown's capture, not a fresh test or a completed case.

Field	Entry
Journey moment	Moving from the Sony WH-CH720N search result to its product page
Expected behavior	The advertised price persists, or a change is explained. This expectation is an analyst hypothesis.
Observed fact	The results card displayed $129.99 with a $48.01 saving and a $178.00 comparison value. Less than a minute later, the product page displayed $178.00 without that saving or an explanation.
Pattern	CTX-003 — Same fact shown differently across screens
Fit	The same product's price differs between two screens in one journey without explaining the difference.
Interpretation	The discrepancy may make the actual offer harder to determine.
Unknowns	Cause, frequency, affected users and behavioral consequences. The capture does not establish abandonment or lost trust.
Severity	Not yet assigned; apply the existing case rubric during review.
Response	Investigate consistency; synchronize the displayed offer or explain a legitimate change before commitment.
Verification	Repeat the transition under matched conditions, then assess whether users can correctly identify the applicable price and conditions.
Owner / result	Unassigned / not tested

Template

Use this as a working template for each finding:

At [step/transition], [user in context] expects [outcome], but [observed behavior].
Evidence: [observation + capture references].
Pattern: [code — name], because [fit].
Impact: [task consequence]; severity: [existing case rating]; reach: [known/unknown].
Uncertainty: [what remains unverified].
Response: [decision + intervention].
Owner and verification: [person, predicted effect, method, guardrails].
Result: [not tested / supported / contradicted / inconclusive, with evidence].

How to start

Run a first pass over one flow with UX Snapshot, then map the gaps it surfaces. To see where the map sits inside a broader review, read about the full UX audit. If you want our team to build the map with you, start with a Full UX audit.