How to research competitors before starting a business

Competitor research is not about looking at a landing page and deciding you can design a better logo. It is a mathematical exercise to determine if a market has money and where that money is currently poorly served. Most founders waste weeks staring at competitor feature matrices. You need to extract four specific data points from the incumbents: their estimated revenue based on review volume, their pricing boundaries, the demand signals hidden in their negative feedback, and the features keeping users locked in.

The Math of Review Volume

Do not guess if a competitor is making money. Use their app store or directory reviews to run a baseline calculation. Only a fraction of paying users leave a review. By understanding this ratio, you can reverse-engineer their active user base and their revenue floor.

For iOS apps, a standard heuristic is that one in fifty active users leaves a rating. If a productivity app has 2,000 ratings, they likely have around 100,000 historical users. To get revenue, multiply that user estimate by their lowest paid tier. If the app costs $5 a month, and we assume a conservative 5 percent conversion rate from free to paid, that yields 5,000 paying users. That is $25,000 a month in recurring revenue. This is a floor. It tells you the market exists and people are willing to pay.

For B2B software, the math changes. Look at G2, Capterra, or Trustpilot. B2B review rates are much lower, often around one in two hundred users, unless the company aggressively incentivizes reviews with gift cards. If a niche CRM has 50 reviews, assume at least 10,000 historical users. If their average contract value is $1,200 a year, and 10 percent of those users are active paying customers, you are looking at $1.2 million in annual recurring revenue.

Do not just look at the total number of reviews. Look at the velocity. Sort the reviews by date. If a competitor has 5,000 reviews but only ten were left in the last year, the product is dead. The revenue you calculated is a trailing indicator. You want to find competitors accumulating dozens of reviews every week. High review velocity indicates active user acquisition. It proves the problem is still relevant today.

You extract this data to prove the pain is acute enough that people pull out a credit card. If you find a competitor with a terrible product but thousands of reviews, you have found a validated market. A bad product with high revenue is the best signal a founder can find. It means the problem is so severe that customers will pay for a subpar solution.

Mapping the Pricing Floor and Ceiling

Pricing is the most visible expression of a competitor strategy. You must map the floor and the ceiling of your chosen market. The floor is the cheapest viable paid option. The ceiling is the enterprise tier.

Pick three direct competitors. Document their monthly subscription price, their annual discount, and the specific feature that forces a user to upgrade from the basic tier to the pro tier. This is the paywall trigger. If three different habit trackers all restrict data export to their $9 per month tier, data export is a high-value feature. Customers are demonstrating they will pay a premium to control their own data.

Pay close attention to arbitrary limits. Many SaaS companies price based on usage metrics like seats, storage, or API calls. If a competitor limits users to 100 API calls on their $20 tier and jumps to $100 for 1,000 calls, they are taxing their most active users. If your infrastructure costs allow it, offering unlimited API calls for $40 instantly steals their most profitable cohort.

Never compete on price. If the incumbent charges $10, charging $5 does not win you the market. It halves your margin and attracts the most demanding, least loyal customers. Instead, use their pricing to find the gap.

Look at the annual discount. If a competitor charges $20 a month but offers an annual plan for $100, they are desperate for upfront cash to fund customer acquisition. A massive annual discount usually indicates high churn. Users are quitting after three months, so the company is trying to lock in a year of revenue upfront. This is a weakness you can exploit by building a stickier product.

If the market has a $10 tool for beginners and a $500 enterprise suite for large corporations, the gap is a $50 tool for power users. These are users who have outgrown the cheap option but cannot justify the enterprise suite. Your research must pinpoint exactly where these gaps exist.

Harvesting Demand Signals

Users tell you exactly what to build, provided you know where to look. Do not look at five-star reviews. Five-star reviews are useless platitudes. Look at the two-star and three-star reviews. One-star reviews are usually angry users experiencing a temporary bug or a billing error. Two-star and three-star reviews contain the actual demand signals.

These users wanted the product to work. They paid for it, they tried it, and it failed them in a specific way. You are looking for recurring patterns. If ten different reviews for a meal planning app mention that the grocery list export formats incorrectly for a specific supermarket, that is a demand signal. It is a specific, actionable flaw you can solve.

You must build a system to extract these signals: read Trustpilot, Reddit threads and niche forums on a schedule, and log every complaint about a competitor in the same format so patterns show up.

Go to Reddit and use specific search operators. Search for the competitor name alongside keywords like alternative, issue, missing, or hate. Do not just read the original post. Read the comments. The comments are where power users explain exactly why a product is failing them. You are looking for the feature the incumbent is too bloated to fix. If a competitor has been ignoring a specific user request for three years, that request is your wedge into the market.

If your competitor has a public roadmap or a community forum, you have a goldmine. Look at the feature requests with the most upvotes that have been sitting in the planned column for over a year. Incumbents move slowly. They have technical debt and enterprise clients to appease. A highly requested feature ignored by the incumbent is a demand signal screaming for a standalone product.

Identifying the Lock-In

You must understand why users do not leave a bad product. Every successful product has a switching cost. If you do not know what keeps a user tied to your competitor, your launch will fail. Users will express frustration about the incumbent, try your product, and then go right back to the tool they hate.

Switching costs come in three forms. The first is data lock-in. This happens when a user has years of history in an app. A journaling app is hard to leave because the user past is trapped inside it. The second is workflow lock-in. This happens when a whole team has learned a specific interface. Retraining a team of twenty people costs more than the software itself. The third is network lock-in. This happens when a user clients or friends are already on the platform.

To extract this information, look at churn discussions. Find the Reddit posts where someone asks for an alternative to the incumbent. Read the replies carefully. You will see users say they dislike the current tool but stay because of one specific integration or data format.

If they refuse to switch away because of a QuickBooks integration, your new product must have a QuickBooks integration on day one. It is not an optional feature. It is the table stakes for entering the market. If users mention they cannot export their historical data, your first marketing campaign should be a free migration tool that pulls their data out of the competitor platform. You must dismantle the switching cost before you can sell the solution.

Sometimes the lock-in is purely psychological. A user might refuse to switch because they spent hours configuring a complex dashboard. To break psychological lock-in, your product must offer immediate time-to-value. If the competitor takes three hours to set up, your onboarding must take three minutes. You must prove the grass is greener before their patience runs out.

Synthesizing the Attack Plan

You now have a concrete dataset. It contains estimated revenue floors, paywall triggers, specific demand signals from three-star reviews, and the exact switching costs you must overcome.

Your business idea is no longer a guess. It is a targeted strike. You know the market can support a $30 per month price point because the incumbent is making at least $50,000 a month at that level. You know to build a robust export feature because the demand signals show users are frustrated by data silos. You know to build a CSV importer first because data lock-in is the primary reason users refuse to switch.

This is how you research competitors before writing a single line of code. You do not copy their landing page. You dissect their business model, find the structural weaknesses, and build a product that exploits those gaps. If you cannot find evidence of revenue, clear demand signals, and a switching cost you can break, you do not have a business idea. You have a hobby. Move on to the next market and run the math again.