How to conduct competitor analysis that drives growth (2026 guide)

Competitor analysis works when it ends in a decision. The method is straightforward: choose a question you actually need answered, pick a short list of direct and indirect rivals, gather evidence against a fixed set of headings so the results are comparable, then write down what you will build, price or say differently as a result. If the exercise finishes with a spreadsheet rather than a changed roadmap, it has failed, however thorough the spreadsheet looks.

We have spent over twenty years building software for UK businesses, and we see the same pattern repeatedly. A founder arrives with a beautifully organised feature matrix covering nine competitors, and no clear view of what to do next. The matrix was not wrong. It simply was not pointed at anything.

Start with the decision, not the competitors

Before you open a single rival’s website, write the question at the top of the page. Most competitor analysis serves one of three purposes.

• Roadmap decisions. Which gaps in the market are worth building for, and which “missing” features are missing on purpose?
• Positioning and messaging. How do rivals describe themselves, and what claim is still unoccupied?
• Pricing. What value are competitors delivering at each price point, and does your own packaging make sense against that?

These need different evidence. A pricing question is answered by trial signups and pricing pages. A positioning question is answered by homepage copy, ad copy and review sites. Trying to answer all three at once is how the spreadsheet grows to nine competitors and stops being useful.

Write the decision down in a single sentence: “We are deciding whether to add a free tier before the January launch.” Everything you collect should either support or undermine that sentence.

A competitor analysis model you can actually repeat

A model is only useful if a different person can run it next quarter and produce comparable output. Ours has four stages, and it deliberately puts identification before data collection.

1. Identify the field properly

Most teams list only their direct competitors and stop there. That is the single biggest blind spot in the whole exercise.

Type	Definition	What it tells you
Direct	Same solution, same buyer	Feature parity, pricing pressure, win/loss reasons
Indirect	Same problem solved a different way	Whether your category is the one buyers actually want
Substitute	Spreadsheets, manual process, an existing internal tool	The real reason deals stall
Emerging	New entrants with a different approach	Where the market is heading in eighteen months

For most of the projects we work on, the substitute is the one that decides the deal. A recruitment business is not choosing between two back office platforms; it is choosing between a platform and the Excel workbook that has run the office for nine years. If your analysis does not include the incumbent spreadsheet, it is missing the competitor you are most likely to lose to.

Keep the list to five or six names. Three direct, one or two indirect, one emerging is usually enough. A longer list produces thinner evidence on each.

2. Fix your headings before you gather anything

Comparability is the whole point. Decide the columns first, then fill them for every competitor, even where the answer is “unknown”. A blank cell is itself a finding.

Area	Questions worth answering	Where to look
Product	What is the core job it does well? Where does it get awkward?	Free trial, demo recordings, help documentation
Onboarding	How long from signup to first useful outcome?	Sign up yourself and time it
Pricing and packaging	Tiered, per seat, or usage based? Which feature forces the upgrade?	Pricing page, trial limits
Positioning	Who do they say this is for, in their own words?	Homepage headline, paid ad copy
Customer voice	What do reviews complain about repeatedly?	G2, Capterra, Trustpilot, LinkedIn comments
Direction	What are they hiring for? What has shipped recently?	Careers page, changelog, release notes

The careers page is underrated. A company advertising for three enterprise sales roles and a compliance lead is telling you exactly where it is heading, and it costs nothing to read.

3. Become the customer

Screenshots of a features page tell you what a competitor claims. Using the product tells you what is true. Sign up for every free trial on your list and record the journey: how many steps to first value, where you got stuck, what you had to ask support. Note the friction honestly, including where they beat you. Evidence that only flatters your own product produces a document nobody in the business trusts.

4. Convert evidence into decisions

This is where most analyses stop and where the value actually sits. For each finding, write the implication rather than the observation.

• Observation: “Competitor B has no mobile app.”
• Implication: “Field-based users are underserved across the category, and that is a segment we could own if we prioritise mobile in Q1.”

A short, sharpened SWOT works well here, provided each entry names a competitor and a consequence. “Weakness: our reporting is thinner than Competitor A’s, which is why we lost two enterprise deals last quarter” is useful. “Weakness: reporting” is not.

Direct competitor analysis: what to look at closely

When you are comparing yourself with a genuine direct competitor, three things matter more than the feature list.

The upgrade trigger. Which single feature sits behind the paywall on their entry tier? That is what they believe customers value most, tested against real buying behaviour. It is the most honest signal on the entire pricing page.

The onboarding gap. Time from signup to first useful outcome predicts retention more reliably than feature count. If their trial takes forty minutes of configuration before anything happens, that is a gap worth attacking with better interface design rather than more functionality.

The repeated complaint. Read fifty reviews and ignore the one-off gripes. The complaint that appears fifteen times is a structural limitation, usually caused by an architectural decision they cannot easily reverse. That is defensible ground for you, because they cannot fix it quickly.

What matters far less: the total number of features, their funding round, and their follower count. These make for satisfying slides and change nothing about what you should build.

How often to run it

A full analysis once a year suits most businesses we work with, timed to land before budget and roadmap planning. Between those, a quarterly check of pricing pages, changelogs and careers pages takes an afternoon and catches what matters. Continuous monitoring sounds diligent but produces a stream of updates nobody reads.

Mistakes we see most often

Copying the roadmap. If your conclusion is “they have it, so we need it”, you have handed your product strategy to a company whose constraints and customers differ from yours. Ask why the feature exists before you decide whether it should exist for you.

Ignoring substitutes. Covered above, and worth repeating. Manual process is the most common competitor in bespoke software, and it never appears on a comparison site.

Confusing volume with rigour. Nine competitors researched shallowly is worse than four researched properly. Depth beats coverage, particularly when you have to present findings to a board.

Skipping validation. Competitor evidence tells you what already exists, not whether anyone wants what you propose to build instead. It complements customer conversations; it does not replace them. Our guide to idea validation covers how to test a gap before committing budget to it, and the common mistakes in early-stage app development piece covers what happens when teams skip that step.

Frequently Asked Questions

What is a competitor analysis model?

A competitor analysis model is a repeatable structure for comparing rivals so that findings from different people and different quarters remain comparable. It fixes three things in advance: which competitors are in scope, which headings you gather evidence against, and how findings convert into decisions. Without a fixed model, each analysis uses different criteria, which makes it impossible to see what has changed over time or to act on the results with confidence.

How do I make a competitor analysis step by step?

Write down the decision you need to make. Select five or six competitors across direct, indirect, substitute and emerging categories. Fix your evidence headings before gathering anything, typically product, onboarding, pricing, positioning, customer voice and direction. Sign up for each competitor’s trial and record the experience. Then convert every observation into a written implication, and finish with a short list of prioritised actions. The final step is the one teams skip most often.

What is the difference between direct and indirect competitors?

A direct competitor sells a similar solution to the same buyer, so you meet them in deals regularly. An indirect competitor solves the same underlying problem by a different route, and may not appear on comparison sites at all. Substitutes go further still: a spreadsheet, a manual process or an existing internal system. In bespoke software projects the substitute is frequently the strongest competitor, because it is already installed and nobody has to sign anything.

Is competitor analysis legal and ethical?

Yes, provided you rely on publicly available information. Reading websites, pricing pages, published accounts, job adverts, review sites and using a publicly offered free trial is ordinary business research. Anyone could do the same. The line is crossed when you obtain non-public information through deception, misrepresent who you are to gain access, or approach staff under false pretences. Stay with what is published and you are on firm ground.

Which competitor analysis tools are worth paying for?

For search and content visibility, Ahrefs and Semrush are the established options. For traffic estimates and technology detection, Similarweb and BuiltWith are useful, though treat traffic figures as directional rather than exact. Review platforms such as G2, Capterra and Trustpilot are free and often more revealing than any paid tool. None of them replace signing up for the product yourself, which remains the highest-value hour in the whole process.

How does competitor analysis affect what we build?

It should influence prioritisation, not direction. Use it to decide the order of work and to identify segments nobody serves well, then check those findings against your own customer conversations before committing. Once you know what you are building, technical decisions follow, and our guide to choosing a tech stack covers that next step.

Where to take this next

If your analysis has produced a gap worth building for, the useful next question is what it would cost and how long it would take. Our cost estimator gives a realistic range in a few minutes, and if you would rather talk it through with someone who has built this kind of product before, our SaaS development team is happy to have that conversation. You can also get in touch directly.