Method 16 of 100 · Stage 2: Check demand before you build

Competitor Analysis Grid

Put 5 to 10 competitors, workarounds and the option of doing nothing in one spreadsheet, then find the gap nobody serves well.

What is a competitor analysis grid?

A competitor analysis grid is one spreadsheet where each row is something your customer could use instead of you, and each column is a fact you can check: price, who it’s for, what it promises, what reviewers love and hate. Filled in properly, the grid shows you where the market is crowded, where everyone makes the same promise, and which customers nobody takes seriously.

The rows matter more than founders expect. There are three kinds of competitor. Direct competitors sell a similar product to the same customer. Indirect competitors solve the same problem another way, like a spreadsheet, a freelancer or a generic app. And doing nothing, living with the problem, is the competitor you lose to most.

When should you use it?

Use it when:

  • You’ve found a problem and need to know whether people already pay to solve it. Competitors with paying customers are good news: they prove demand.
  • You are about to set a price or write the headline for your first page.
  • Someone tells you “there’s no competition”, which almost always means the founder hasn’t looked at indirect rivals or doing nothing.

Keep it short when:

  • You are still choosing between five problems. A 10-row grid for each is overkill. Do 3 rows per problem and come back later.

How do you do it, step by step?

  1. Write down the job to be done. Put one line at the top of the sheet: the job your customer hires a solution for. Say, “share household chores fairly between two adults”. Every row must be a way to get that job done.

  2. Find 5 to 10 rows. Search Google for the job, the problem and “[top competitor] alternative”. Check app stores, Product Hunt and Reddit threads where people recommend tools. Add at least 2 indirect rows (a spreadsheet, a general to-do app, a person you hire) and 1 row called Doing nothing.

  3. Set the columns. Use these 10: name, type (direct, indirect, nothing), who it’s for, main promise (copy their homepage headline word for word), price and model, how they get customers (ads, SEO, partners, word of mouth), review score and count, top 3 complaints, what they do brilliantly, and the gap you see.

  4. Fill the facts from their own pages. Price from the pricing page, promise from the homepage, audience from their testimonials and case studies. Paste the Wayback Machine snapshot of their pricing page from a year ago next to today’s: price rises and removed plans tell you what they learned.

  5. Read the 1, 2 and 3 star reviews. Use the app stores, G2, Capterra or Trustpilot, depending on the product. Read at least 20 low reviews per direct competitor and write the three complaints that repeat. For a deeper pass, follow the competitor review analysis method.

  6. Fill the “Doing nothing” row honestly. Price: what living with the problem costs in time, money or arguments. Promise: “free and familiar”. Complaints: what people say in interviews about their current workaround. If doing nothing is cheap and painless, that’s the competitor to beat.

  7. Colour the gaps. Turn the promise column green where promises repeat. Turn audiences red where nobody serves a group. Look for one column where every row is weak. That column is your angle.

What does a worked example look like?

Say Zainab wants to build an app that helps couples split household chores. The rows below are invented types of competitor, so no real product is described.

Her grid has 8 rows: 3 chore apps for couples, 1 family organiser app, 1 general shared to-do app, 1 shared Google Sheet template, 1 whiteboard on the fridge, and Doing nothing.

Patterns jump out fast. All three direct apps promise “never argue about chores again” in nearly the same words. Two charge a subscription and one is free with ads. Their low reviews repeat the same complaint: my partner stopped opening the app after a week. The person who sets up the app stays engaged, and the other partner drifts off.

The Doing nothing row is brutal: zero cost, no new habit, and a weekly argument that most couples have accepted.

Her gap: every product is built for the organised partner. Nobody makes it effortless for the one who doesn’t want another app. She notes a test idea: chores assigned through a weekly text message, no app for the second partner at all. That becomes the promise on her one-page offer.

Which free tools help?

Tool What you use it for Free plan
Google Sheets The grid itself, with colours and filters Free with a Google account
Wayback Machine Old versions of competitor pricing and homepages Free
Trustpilot, G2, Capterra Reviews of consumer and software products Free to read
App Store and Google Play App reviews and ratings Free to read

What mistakes should you avoid?

  • Listing only direct competitors. You lose most deals to a spreadsheet, a cousin who helps, or nothing at all. Give those rows the same attention.
  • Copying features into the grid. Feature lists make every product look the same. Promise, audience and complaints show the real differences.
  • Skipping the reviews. A competitor’s 1-star reviews are free interviews with frustrated customers who have already paid.
  • Picking a fight with the biggest name. Find the customer group the leader ignores and serve them better.
  • Filling it once and forgetting it. Prices and promises move. Put a reminder in your calendar to refresh the grid every quarter, and set Google Alerts monitoring for each name.

How do you know it worked?

You can finish this sentence from the grid: Unlike [rows], we are for [group] who [complaint], and we promise [gap]. Our rule of thumb for a go: at least 2 direct or indirect competitors have paying customers (proof of demand), and you can point to one complaint that shows up in at least 3 rows. If no one pays for anything in the space, test willingness to pay before building. If every row is strong and nobody complains, pick a narrower customer group.

What should you do next?

Turn the price column into a full competitor pricing map, then choose the group nobody serves with beachhead segment selection. Use bottom-up market sizing to check that the gap is big enough to pay you. The Fe/male Switch startup game gives you an AI co-founder that walks through competitor research and positioning with you.

FAQ

How many competitors should be in a competitor analysis?

5 to 10 rows is enough for an early-stage founder. Include at least 2 indirect options and the Doing nothing row, because together they decide most purchases.

What is the difference between direct and indirect competitors?

Direct competitors sell a similar product to the same customer. Indirect competitors solve the same problem in another way, like a spreadsheet, an agency, a freelancer or a general tool.

What if I can’t find any competitors?

Search for the problem instead of your solution, and look at what people use today, even if it’s clumsy. If people truly do nothing and complain little, demand is weak. If they do nothing and complain loudly, you have an opening worth testing.

Updated 26 October 2026 · Written by the Fe/male Switch team

Methods that go well with this one

Method 10

Competitor Review Analysis

Read low-star reviews of market leaders on G2, Capterra and Trustpilot and map why customers switch away from them.

Find the problemIntermediate4 to 6 hours
Method 49

Competitor Pricing Map

Collect competitor prices into one table, convert them to price per unit of value and plot them to find an open spot for your price.

Price and pre-sellBeginner4 to 6 hours
Method 15

Bottom-Up Market Sizing

Count the customers you can actually reach, multiply by what they would pay, and get a market size you can defend in one spreadsheet.

Validate demandIntermediate3 to 4 hours
Method 50

One-Page Offer Sheet

Put your whole offer on one page: who it's for, the result, what's included, price, guarantee and deadline, then test it on sales calls.

Price and pre-sellBeginner3 hours to write, 2 weeks of calls to test
Method 28

Beachhead Segment Selection

Score your possible customer groups on reach, urgency, budget and word of mouth, then commit to the one narrow group you can win first.

Customer discoveryIntermediate2 to 3 hours, plus a week of quick tests

Go deeper