Competitive Research: Building a Repeatable Method
Last updated on · written from the review set and vendor pages; the oldest reading dates from .
Before you start
3 things to have ready. Each one is assumed by a step below; get it first.
A decision the research will inform
Research without a decision to serve drifts into collecting facts; knowing the choice up front tells you what to keep and what to skip.
Your own offer, written down
Gaps only show against something, so you need your own product, price and buyer described in plain words before you compare.
Time to read public sources
The method relies on public pages and published material, so what you need is time to read and note them, not special access.
The steps
7 steps, in order. Tick each one once its check passes; “How you know it worked” is what you should see before moving on.
- Define the decision the research serves
- List direct and indirect competitors
- Choose sources you can check
- Build a profile of each competitor
- Measure each competitor's online reach
- Compare the profiles with your offer
- Schedule the next pass and an owner
Write the decision this research must support, such as entering a segment, changing a price or sharpening your pitch. Then list the questions about competitors that the decision depends on, and drop anything that does not bear on it.
How you know it workedYou have a written decision and a short list of questions tied to it.
If it failsEvery question seems important: keep only those whose answer would change the decision.
Write down every company a buyer could pick for the same job. Direct competitors sell a similar product to the same buyer; indirect ones solve the same problem another way, including a service, a manual process or doing nothing.
How you know it workedYour list names each competitor and marks it as direct or indirect.
If it failsYou cannot tell who competes: ask recent customers what else they considered before they chose.
For each question, pick sources a reader could verify: the competitor's own pricing and product pages, published announcements, customer reviews and job postings. Prefer sources that show a date, and write down which source answers which question.
How you know it workedEach question on your list has a named source and a note of when you read it.
If it failsA claim rests on a lone page: confirm it on another public page or mark it unconfirmed.
Fill the same template for every competitor: what it sells, to whom, at what price and how it sells, plus its size, funding, hiring and the technology it uses. Identical fields let the profiles compare side by side.
How you know it workedEvery competitor has a profile with the same fields, each fact tied to a dated source.
If it failsA field stays empty: write not found and leave it that way, never estimate.
Add what shows from outside: how much traffic a competitor draws, which search terms and ads it works on, and what people say about it in public conversations. Whether you use competitive research tools or read pages by hand, note the date of each figure.
How you know it workedEach profile now holds reach figures and public talk, each with the date you read it.
If it failsFigures from two sources disagree: keep both with their dates and treat the fact as unconfirmed.
Put your own product in the same template and mark where each competitor is stronger, weaker or equal. State every gap as a fact a buyer could check, leave opinions out, then answer the decision you wrote at the start.
How you know it workedEach gap is a checkable fact, and the decision has a written answer.
If it failsYou find no gap: report that, and look at how buyers choose rather than only at what products do.
Treat competitive intelligence research as a loop, not a project. Choose a rhythm that fits how fast your market moves, name who redoes it, and log the date and what changed at each pass.
How you know it workedA named owner, a date for the next pass and a log of changes exist.
If it failsNobody has time: shrink each pass to the competitors that came up in recent decisions.
Tick each step once its check passes.
Keep going: a step whose check fails is redone, not skipped.
Every step ticked. Run the final checks below.
Did it work?
4 checks before you call it done. If one fails, the line says which step to redo.
The decision and its questions are written down
Pass whenThe decision and its questions sit together in a short note that a colleague could read and understand.
If notredo step 1 · Define the decision the research serves
Competitors are marked direct or indirect
Pass whenEach company on the list is marked direct or indirect, with the buyer need that links it to you.
If notredo step 2 · List direct and indirect competitors
Every profile fact has a dated source
Pass whenEach fact in a profile names the public page it came from and the date you read it.
If notredo step 4 · Build a profile of each competitor
Gaps are stated as checkable facts
Pass whenEach gap in your comparison is something a buyer could check on a public page, not an opinion.
If notredo step 6 · Compare the profiles with your offer
A check that fails sends you back to one step, never to the start.
If something breaks
3 failures people hit once the steps are done, and what to do about each.
The research keeps growing and never ends
Do thisNarrow it back to the decision you wrote: drop every competitor and field that would not change that decision.
Profiles are too different to compare
Do thisFields were chosen competitor by competitor: fix a shared template, refill each profile with the same fields and mark missing data as not found.
The comparison flatters your own product
Do thisYour side was filled from memory while competitors came from sources: give your product the same dated sources and fields before judging any gap.
The tools named in these steps
Grouped under the step they serve, not ranked. A name opens our review; Try goes to the vendor.
Also reviewed: Bright DataClaudeGeminin8nOpenAISurfer SEOZapier
Grouped by the step they serve, alphabetical within a step. No ranking implied. 12 of the 54 tools we score in AI Automation.
Related questions
Competitive research studies the companies a buyer could choose instead of you: what they sell, to whom, at what price and how. Market research studies the buyers and the market itself, such as segments and needs.
Start from the buyer's need, not the company's look. A close substitute for your product is direct; a different route to the same result is indirect. The same company can be direct for one buyer and indirect for another.
A profile should let someone who has not seen the sources judge the competitor: its offer and buyer, how it earns money, signs of scale, and for every fact a source and a read date. Keep the fields identical across competitors.
No. A SWOT lists your own strengths, weaknesses, opportunities and threats. Competitor analysis profiles the others and compares them with you. A SWOT can come out of competitive analysis research, but it does not replace it.
As often as your market moves, and after any event that changes your decisions, such as a new entrant, a price change or a lost deal. Log the date of every pass so you can see what changed.
Take a company's own price and offer from its dated pages, and confirm anything it says about itself with a source outside that company. Check an older page again before relying on it.
This page walks through one procedure; it does not score tools.
Affiliate linksThis page contains affiliate links: some vendors pay us a commission when you subscribe through their Try button. Scores come from public user reviews, on the same grid for every scored tool. Affiliate disclosure