Competitive analysis is one of those tasks that sounds straightforward and always expands to fill a week. Left unstructured, it produces a thirty page slide deck of screenshots and no decision. This method keeps it focused: map the field, compare on the dimensions that matter, turn the findings into a decision, and leave the reader with something they can act on. It is the same sequence we run for clients before a product launch, a pricing change, or a go-to-market plan.
Step 1: Define the competitive set
The first decision is who is in scope. Start with direct competitors: the players a customer would genuinely choose instead of you for the same job. Then add two adjacent players: businesses that serve a nearby need today and could move into your space if it becomes attractive. Leave out everyone else. A set of five to eight well chosen players beats thirty loosely related names, because every extra row in the analysis costs time to research and dilutes the comparison.
A common mistake is defining the set by what people in the office fear, rather than what customers actually consider. Ask a real buyer who else they looked at before choosing, and that list is usually better than the one the product team produces. Write each competitor in one line: who they are, what they sell, and who they sell to. If you cannot describe a competitor in one line, you do not understand it well enough to analyze it.
Step 2: Choose comparison dimensions that matter
Pick three to five dimensions tied to what customers actually decide on: pricing, core functionality, target customer segment, platform, and support model are the usual starting points. Resist the urge to compare everything. Comparison tables with fifteen rows are hard to read and harder to act on, and the extra rows are usually dimensions where nobody differs anyway.
The test for a dimension is simple: if it would not change a customer's choice or your strategy, cut it. For a SaaS product the deciding dimensions are often price point, depth of functionality, and ease of onboarding. For a consulting firm they might be sector expertise, senior team composition, and pricing model. For a physical product they are likely distribution reach, unit economics, and brand. Choose the list that matches the purchase decision in your market, and write a one line definition for each dimension so the comparison stays consistent.
Step 3: Gather evidence, not vibes
For each competitor, collect what you can point at. Pricing pages give you the price points and the packaging. Feature lists and documentation show what the product actually does. Case studies and review sites show what customers say after buying. Hiring signals show where the company is investing. Funding announcements show what they plan to do next, and job postings are often the earliest evidence of a new market move.
Keep a source next to every claim from the start. The reason this matters is that the final output will be read by people who may act on it, and every claim in it should trace back to something public and verifiable. This is also where the AI hallucination risk lives, so keep sources attached and treat any unverified number as missing rather than guessed. If you cannot find evidence for a claim, either find it or leave the cell empty.
Timebox the research as well. Set a limit per competitor, for example half a day for the first pass, and treat anything found after that as a refinement rather than a requirement. The point of the exercise is a decision at a point in time, not perfect knowledge. If two competitors are nearly identical, you do not need two parallel research efforts at double the cost; you need the same depth of source for each and a note on where they genuinely differ.
Step 4: Build the comparison
Put the competitors on the rows and the dimensions on the columns, then fill in facts rather than adjectives. "Cheap" is an opinion. "$49 per user per month on the starter plan" is a fact. For each cell, add a short note on what the fact implies: that the price puts them below the mid-market point, that the feature is missing entirely, that onboarding takes three days. Then add a final column titled "what this means" for the single most important takeaway per competitor.
The gap analysis comes out of this table. Read across each row and ask where no one is strong. That empty space is either the opening you can take or the space no one wants, and the research determines which. Read down the columns and look for crowding: if every player scores identically on a dimension, that dimension is table stakes and not where you should compete.
A worked example: you are comparing three project management tools for an internal recommendation. Pricing reads $10, $12, and $15 per user per month, functionality is broadly equal, but one has no offline mode and another caps file storage. The row writes itself: the decision comes down to which constraint the team can live with, and the table has already surfaced it. That is the point of building the comparison in facts: the implications are visible without interpretation.
Step 5: Turn it into strategy
The output is not the table. It is the answer to two questions:
- Where is the wedge? The dimension where you can win and competitors are weak. A competitor may be strong on price but weak on service, or strong on features but weak on onboarding. The wedge is the combination a buyer would choose you for.
- Where are the red flags? A competitor about to move into your segment, a feature you are missing that is becoming table stakes, or a pricing move that undercuts your position.
Write three moves you would actually make as a result: one offensive move that leans into the wedge, one defensive move that covers the red flag, and one test you would run to confirm the gap is real. A competitive analysis that ends in actions is worth the week. One that ends in a table is not.
Common mistakes to avoid
- Defining the set by keyword, not by buyer. Every company that ranks for the same search term goes in, whether or not they compete for the same customer. The table grows, the decision does not.
- Comparing different levels of evidence. One cell is verified against a pricing page, the next is a guess from a sales call. Decide a consistent evidence standard for the whole table.
- Skipping the refresh. A scan done in March is stale by June if the market is moving. Put a rerun date on the file before you close it.
- Presenting the table instead of the moves. The deliverable that gets acted on is the wedge, the red flags, and the three actions, not the matrix itself.
Speed it up
Competitive scans are exactly the kind of research task where structure saves the most time. Start the comparison with the free competitive analysis tool, then spend your effort on the sources and the strategy that come out of it.