Skip to content
UX Atlas
Plate RES·019 · Research methodGrid H1 · Rev. 2026-09-12

Competitive Analysis

Study how others solved the same problem to learn the conventions your users already hold, not to assemble a feature list.

Research methodCraft conventionFoundational

Definition

Competitive analysis examines comparable products against defined criteria. Done well it establishes what your audience already expects and where a genuine difference is available. Done badly it produces a feature matrix that justifies copying.

Craft convention. True because the industry converged on it. Breaking it costs familiarity, not correctness.

On this plate
  1. 1Definition
  2. 2Running one that is worth reading
  3. 3Using the result
  4. 4What to record per competitor
  5. 5In practice
  6. 6Local chart
  7. 7Continue from here

Reach for this when

  • Entering an established category.
  • Deciding whether to follow or break a convention.
  • A stakeholder is convinced a competitor solved something you have not looked at.
  • An argument is running on what everyone else does, and nobody has actually looked.

Common mistakes

  • Feature matrices with scores.
  • Only direct competitors included.
  • Marketing pages examined instead of the product.
  • Competitor decisions treated as validated.
  • Reviewing only direct competitors, when the convention your users hold comes from a different category entirely.

Running one that is worth reading

  1. 1

    Define the comparison set

    Direct competitors, adjacent products solving the same job, and the products your audience uses most in general. The last group sets their conventions, and it is usually the group nobody includes.

  2. 2

    Write the criteria first

    Derived from your framed problem. Criteria chosen after looking are criteria chosen to flatter.

  3. 3

    Complete real tasks

    Sign up, buy, cancel, recover a password. Screenshots of marketing pages tell you nothing about the product.

  4. 4

    Record the decision, not the surface

    'Two-step checkout with address validation on step one' is useful. 'Clean modern design' is not.

  5. 5

    Separate convention from choice

    Where every product does the same thing, that is a convention your users hold and deviating from it costs. Where they differ, there is a decision available.

Using the result

Do

  • Note what everybody does, and treat it as the baseline expectation.
  • Note what nobody does, and ask whether that is an opportunity or a well-known dead end.
  • Capture the failure modes you hit, since competitors' bad states are the cheapest lessons available.
  • Date the analysis. It is out of date within a quarter.

Do not

  • Do not treat a competitor's design as validated. You cannot see their data, and they may be shipping a mistake.
  • Do not build a feature matrix and score it. The scoring implies a precision the method does not have.
  • Do not substitute this for talking to your own users.

What to record per competitor

A comparison that lists features produces a feature list. A comparison that records decisions produces an argument you can use.

ColumnWhy it earns its place
The flow, screen by screenWhere the steps are is the decision. A feature name is not.
What they ask for, and whenAccount, payment details, personal data. The order tells you what they think they can charge the user for.
Defaults and preselectionsThe cheapest read on what the business actually optimises for.
What they refuse to doOmissions are decisions too, and usually the considered ones.
Public complaintsReviews and support forums show where the design fails after the demo.

In practice

  1. i.

    The adjacent product that set the expectation

    A B2B tool's users judged its search against a consumer search engine, not against other B2B tools. Including that comparison explained complaints the direct competitor analysis could not.

  2. ii.

    Copying the outcome of a constraint

    The usual failure

    A team matched a competitor's three-step checkout without knowing the third step existed to satisfy a payment regulation in one market. They imported the cost and none of the reason.

Local chart

RES·019 · Region IV · Grid H1

This entry and the entries it is stated to connect with. Each line is labelled with what the connection means.

Authority BiasBIAS·006 · PRINCIPLES BEHIND THISJakob's LawLAW·003 · PRINCIPLES BEHIND THISMental ModelsMOD·001 · PRINCIPLES BEHIND THISHeuristic EvaluationRES·002 · RELATED CONCEPTDiscoveryFLOW·004 · RELATED CONCEPTProblem FramingFLOW·007 · RELATED CONCEPT

Continue from here

Each link says what the connection is, so you can tell a principle from an alternative from a thing people mix this up with.

Principles behind this

The reasoning this solution is an application of.

Related concept

Connected closely enough to change how you apply this.