Search for SaaS competitor analysis and you get pages written for marketing teams at companies that already sell something. They assume you know your competitors by name, that you have a product to compare feature by feature, and that the output is a battlecard for a sales team.
If you are pre-launch, none of that is true. You have an idea, a guess about who has the problem, and no idea who else is already serving them. The hard part is finding the competitors you have never heard of, including the ones that are not software.
This is the version of the process that works at that stage.
Why the standard framework misfires before launch
The usual advice starts with a competitor grid: list the players across the top, features down the side, mark who has what. It is a good tool at the wrong moment.
The grid answers "how do we compare", which presumes the market already accepted that this category should exist. Before launch your question is different. You need to know whether people are solving this problem today, what they use, what it costs them, and whether any of that leaves an opening.
A founder who runs the feature grid too early gets a confident-looking document built on a competitor list that came from one Google search. The three companies with the biggest marketing budgets show up, and the workaround that eighty percent of the market actually uses never appears, which is usually the row that decides whether the business works.
Step 1: Build the list you do not have
Five sources, in the order that produces the least obvious results.
Review site adjacency. Find one product that is roughly in your space on G2 or Capterra, then read the "compared to" and "alternatives" sections. Those lists are built from what buyers actually evaluated side by side, which is more honest than any category page. Follow the chain two hops out and you will surface companies that never rank for your keywords.
The word "alternative" in communities. Search Reddit, Hacker News and the relevant Slack or Discord for "alternative to" plus the category. People asking for alternatives are describing an unmet need in public, and the replies name products no listicle covers.
What people are paid to do manually. Look at freelance marketplaces and agency service pages for the outcome your product would deliver. If someone sells it as a service, the demand is proven and the current solution is expensive. That is a competitor even though it is a human.
Recently launched and recently dead. Product Hunt and launch communities show who arrived in the last two years. Equally useful is who left: a shutdown post explaining why a company in your space stopped is the cheapest market research available, and founders write them honestly.
The internal build. In B2B especially, your biggest competitor is often the customer's own engineer with a weekend and a database. Ask any potential buyer what they would do if your product did not exist, and listen for "we'd probably just build something".
Stop when new sources stop producing new names. That usually happens faster than founders expect, around the fifteen-name mark, and the list then needs cutting rather than growing.
Step 2: Sort into three buckets, and take the third one seriously
Direct. Roughly your product, roughly your buyer. These are the ones you will be compared to on a sales call.
Indirect. Same problem, different shape of solution. A scheduling tool and a virtual assistant service both fix a calendar. They compete for the same budget line even though a feature grid would say they have nothing in common.
Substitutes. The spreadsheet, the email thread, the checklist in someone's notes app, the intern who does it every Monday. This is where most early-stage revenue actually comes from, because you are not persuading someone to switch products. You are persuading them to stop doing it by hand.
The reason the third bucket matters most is inertia. A user paying a competitor $50 a month has already accepted that this problem is worth money, and switching is a comparison. A user doing it manually has decided the opposite, and your job is to change their mind about the category before you sell them anything. Those are two completely different products, two different pitches, and two different prices.
Get this wrong and you build for the switcher while your actual buyer is the manual one.
Step 3: Record six things, not thirty
For each of the five to eight companies that survive the cut, capture only what changes a decision.
Who they say it is for. Their own words from the homepage, not your paraphrase. The specificity of that sentence tells you how narrow their focus is, and a vague one is an opening.
Entry price and what it unlocks. The cheapest real plan, and what a customer is blocked from doing on it.
What sits behind the highest tier. This reveals where they believe the value is, which is often not where their marketing puts the emphasis.
Their most recent shipped thing. Changelog or release notes. A product that has shipped nothing in eight months is not defending its position.
The complaint that repeats in reviews. One sentence. Method for finding it is the next step.
What they would have to break to serve your customer. This is the strategic question and the only one that requires judgment. If serving your segment would mean cannibalizing their enterprise revenue or rebuilding their data model, they will not follow you there. If it would take them a sprint, your differentiation is a feature and not a position.
Six fields across eight companies fits on one page. If the document runs longer than that, it has stopped being useful.
Step 4: Read the one-star and three-star reviews
Feature pages tell you what a company wants to be true. Reviews tell you what it is like on Tuesday.
Skip the five-star reviews, which are mostly incentivized, and skip the one-star reviews that are really billing complaints. The three-star review is where the honest text lives, written by someone who kept using the product and still had something to say.
You are looking for repetition across different reviewers. A single frustrated user is noise. The same specific frustration from six people is a structural gap, and the fact that it persists in a funded product usually means fixing it conflicts with something the company cannot change.
Do the same pass on the community threads. When someone asks for a recommendation, the replies that begin "we used X but" are the most valuable sentences in competitor research, because they contain both the choice and the regret.
Write down the exact phrasing people use. It is your positioning copy, and it is better than anything you would have written from scratch.
Step 5: Take the pricing apart
Pricing is the clearest statement of who a company has decided to serve, and it is public.
Look at where the tier jumps are. A product that goes $29, $99, $499 has decided the middle of the market is small. Look at what triggers the jump: seats, usage, or a feature. Seat-based pricing punishes teams and leaves an opening for anything priced per outcome. Usage-based pricing punishes the successful customer, which is why their heaviest users are the most likely to leave.
Then look for the gap. If every competitor starts at $99 a month and the segment you care about would pay $20, that is either your opening or a warning that the segment does not have money. The difference is whether they currently spend anything at all on solving the problem, including their own time.
Free tiers are worth reading closely too. A generous free tier usually means the company needs volume for a network effect or is buying data. A stingy one means the product costs real money to run per user, which tells you something about your own future margins.
Step 6: Convert it into one decision
The analysis is only finished when it produces a sentence you can put in front of a customer.
The shape is: for [specific buyer], who currently [what they actually do today], we do [the one thing that differs], unlike [the closest alternative].
Fill it in. If you cannot name the buyer specifically, the research was too shallow. If the "unlike" clause is a feature that any competitor could ship in a sprint, you have found a feature and not a position. If "what they do today" is nothing at all, you are entering a market that does not exist yet, which is possible but should be a deliberate choice rather than an accident.
Then go test that sentence on five people who are not friends. Everything before this step is desk research, and desk research can be wrong in a way that no amount of additional desk research will reveal. The Mom Test covers how to ask so people tell you the truth rather than the polite version.
What "there are no competitors" actually means
Almost always one of three things, and only one of them is good news.
The most common is that you have searched for the product rather than the problem. Nobody sells software for it because everyone does it in a spreadsheet. Search for the workaround and the market appears.
The second is that someone tried and it did not work. Look for shutdown announcements, acquired-and-buried products, and Reddit threads where a product is discussed in the past tense. Understanding why the previous attempt failed is worth more than any amount of competitor research on live companies, because you are about to make the same attempt.
The third, rarely, is a genuine opening created by a recent change: a new regulation, a new platform, a price shift that made something viable that was not viable two years ago. This one is real, and the test is whether you can name the change and the date. If you cannot, you are probably in case one or two.
An empty market usually means no market. Demand signals before building covers what evidence of a real market looks like when there is no competitor to point at.
What a crowded market actually means
The opposite mistake is abandoning an idea because twelve companies already exist.
Competition proves the problem is worth paying for, which is the single hardest thing to prove. A crowded market has already done your demand validation. What you need to establish is different: whether any meaningful segment is underserved by all twelve.
The answer is usually yes, and it usually lives at the edges. The customers too small for enterprise tools. An industry with a vocabulary and a compliance rule that generic products ignore. A workflow that sits between two products and belongs to neither.
The question to answer is not "is this market crowded". It is "who in this market is being served badly by everyone, and can I reach them". If you can name that group and where they gather, crowding is an advantage, because your competitors have already spent the money teaching the market that this category exists.
How often to redo this
Once properly before you commit to building, and then lightly every quarter.
The quarterly version is fifteen minutes: check the changelogs of your three closest competitors, check whether anyone new has appeared in the community threads, and check whether pricing moved. Competitor analysis that turns into a monthly ritual is a form of procrastination, and the founders who do it most are rarely the ones shipping.
The full version is worth repeating when something structural changes, which usually means a competitor raises a large round, a large player enters, or your own positioning stops working in sales conversations.
Where to start
If you have the idea and not the list, the finding is the hard part, and it is the part that generic tools do not help with because they start from a competitor URL you do not have.
Scoutr's competitor analysis works from a description of the idea instead. It maps direct and indirect players, surfaces the workarounds people currently use, pulls the complaints that repeat, and gives you the six fields above without the two days of tab management. From there the judgment is yours, which is the part that should be.
If you are earlier still and testing whether the problem is real at all, how to validate a business idea covers the full sequence from first conversation to first payment, and the product discovery framework covers the six questions this analysis feeds into.