All work

Consumer App Usability Evaluation

A hierarchical task analysis and heuristic evaluation of logging an oil change in the Carfax Car Care iOS app, with Figma redesign mockups for the three issues it found.

OVERALL GOALLog a completed oil changeAdd Vehicle to AppOpen app andtap “add car”Enter VIN orlicense plate numberEnter state/provinceTap “search”Verify and confirmvehicle detailsSelect “Yes thisis my car”Verify ZIP codeVerify total mileageVerify yearly mileageNavigate to servicecategoryNavigate to“Oil Change”Tap “AddService Record”Enter servicerecord detailsSelect date ofoil changeSelect oil changelocationEnter odometerreading at timeof serviceSelect any otherwork done on car,if applicableSave service record
Hierarchical task analysis for logging a completed oil change. Highlighted steps are the three risk points.
Project type
Self-directed
My role
Sole evaluator
Methods
HTA, heuristic evaluation
Tools
Figma
Date
Sep 2026
Status
Usability test planned

The task and the approach

Carfax’s Car Care app lets people track their vehicle’s service history. I evaluated one routine task, logging a completed oil change: add a vehicle, confirm its details, choose a service category, enter the service details, and save.

I used two complementary methods, then compared their results:

  • Hierarchical task analysisBreaks the goal into five sub-goals and their steps, showing where in the flow the design creates ambiguity, load, or room for error.
  • Heuristic evaluationChecks the same flow against Nielsen’s 10 usability heuristics, explaining which design principle each problem violates.

An issue that turns up in both methods is likely built into the structure of the flow. One that turns up in only the heuristic evaluation is more likely a matter of wording.

Three findings and redesigns

Each issue is mapped to the task step where it occurs and the heuristic it violates, with the original screen beside a Figma redesign.

1. VIN input guidance

Task step
1.2: Enter VIN or license plate
Heuristic
#6 Recognition rather than recall
Severity
Low

The Add a Car screen offers a VIN or a license plate and state, with no hint about which to use. A VIN is more precise and error-resistant, but the design doesn’t say so, so first-time users have to guess.

Fix: add a “Recommended” label next to the VIN option, keeping the license plate as an alternative.

Original screen: Add a Car screen
Before
Redesign mockup: Add a Car screen
After

2. Oil change location

Task step
4.2: Select oil change location
Heuristic
#2 Match between system and the real world
Severity
Low

The app asks “Was this oil change done at a service shop?” with Yes and No buttons. That framing assumes a shop by default, so people who changed their own oil have to translate “No” into their situation.

Fix: replace the yes/no question with two labeled buttons, “Service Shop” and “Self-Service,” matching how people describe where it was done.

Original screen: oil change location question
Before
Redesign mockup: oil change location question
After

3. Service dropdown overload

Task step
4.4: Select other work done
Heuristic
#8 Aesthetic and minimalist design; #6 Recognition rather than recall
Severity
Low to medium

Each of the nine “Add other service” dropdowns holds up to 100+ options in one flat list, many sharing the same prefix. The search bar doesn’t help people who don’t know the exact term, so anyone unfamiliar with car terminology has to scan the whole list.

Fix: group each dropdown’s options by prefix, so people scan a few groups instead of a hundred items.

Original screen: service dropdown list
Before
Redesign mockup: service dropdown list
After

What the two methods showed together

FindingFound byWhat it means
1. VIN inputHeuristic evaluation onlyA wording issue. Likely a brief pause, not task failure.
2. Oil change locationHeuristic evaluation onlyA framing mismatch. Minor hesitation, not a blocker.
3. Service dropdownsBoth methods, independentlyA structural issue, and the highest-priority fix, especially for non-experts.

What I learned and next steps

Two methods surfaced more than either would alone. The task analysis showed where load builds up in the flow; the heuristic evaluation gave me the vocabulary to explain why those spots fail. Comparing them showed which issue to fix first.

Next step: a moderated usability test to check whether each redesign actually reduces hesitation and error, and whether people with no car knowledge find the grouped dropdowns easy to navigate.

See all projects

Questions about this project? jbrazina2324@gmail.com