market research & sizing

Every investor deck has a TAM slide. Most of them are fiction. A PM pulls a number from a Statista report, slaps it into a circle diagram, and declares a $50 billion opportunity. The board nods. The number means nothing because nobody did the work to connect it to reality.

Market research is not about finding large numbers to justify your existence. It is about answering a specific question: how many people will pay how much for this thing, and what stands in the way? If your research does not change a product decision, you wasted your time.

The two approaches to sizing

There are two ways to size a market. You need both, and the gap between them is where the insight lives.

Top-down starts with the total market and narrows. “The Indian SaaS market is $12 billion. Project management tools are 8% of that. We target SMBs, which are 30% of the segment. So our addressable market is $288 million.” This is what most people do. It is easy, impressive-sounding, and nearly useless for product decisions.

Bottom-up starts with your unit economics and scales. “There are 1.4 million registered MSMEs in Maharashtra using digital tools. 22% use some project management software. Our conversion rate from free trial is 4.5%. At Rs 499/month, that gives us a reachable revenue of Rs 33 crore in year one from one state.” This is harder, less glamorous, and actually useful.

The rule: if your top-down and bottom-up estimates are more than 5x apart, something is wrong with your assumptions. The gap forces you to interrogate which inputs are real and which are wishful.

TAM, SAM, SOM — without the self-deception

These three concentric circles show up in every business school slide. Here is what they actually mean and how to compute them without lying to yourself.

TAM (Total Addressable Market): The total revenue opportunity if you had 100% market share. This is a theoretical ceiling. For most products, your TAM is someone else’s published research plus your segmentation logic. Use Statista, IBISWorld, or Gartner reports as starting points — but never use their number directly. Always adjust for your geography, segment, and pricing.

SAM (Serviceable Addressable Market): The portion of TAM you can actually serve given your product’s current capabilities, geography, language, and go-to-market. If your product only works in English and you sell to Indian SMBs, your SAM is dramatically smaller than your TAM. This is where most PMs get lazy. They hand-wave SAM as “we’ll expand to that eventually.” Do not. Define SAM based on what you can serve today.

SOM (Serviceable Obtainable Market): The portion of SAM you can realistically capture in the next 12-18 months given your team size, sales capacity, and competitive position. This is the number your quarterly targets should connect to. If your SOM does not match your revenue plan, one of them is wrong.

The discipline is in the filtering. Each step from TAM to SOM should cut the number by 60-90%. If your SOM is more than 10% of your TAM, you are not filtering hard enough.

Competitive analysis that changes decisions

Most competitive analyses are graveyards of screenshots. A PM spends two weeks building a spreadsheet comparing 15 features across 8 competitors, presents it in a meeting, and nothing changes. Nobody makes a different decision because of it.

Useful competitive analysis answers exactly three questions:

1. Where are customers leaving competitors, and why?

Do not start with feature matrices. Start with churn. Read competitor reviews on G2, Capterra, and app stores — not the 5-star reviews, but the 2-star and 3-star ones. These are people who tried the product, found partial value, and left because something specific was missing or broken. That “something” is your opportunity.

2. What is the competitor’s business model constraint?

A competitor’s pricing model, funding stage, and target customer determine what they cannot do. A VC-funded competitor burning cash on enterprise sales cannot profitably serve SMBs. A freemium competitor with 10 million free users cannot introduce a feature that disrupts their free tier. These structural constraints are more durable than feature gaps — a feature can be copied in a quarter, but a business model constraint takes years to unwind.

3. What is the switching cost, and who bears it?

If switching from a competitor to your product requires data migration, team retraining, or integration rebuilding, the feature comparison is irrelevant. The switching cost is the moat — theirs or yours. Understand it before you build anything.

The India-specific research problem

Market research in India has a data gap that PMs in the US or Europe do not face. Reliable industry reports for Indian market segments are either expensive (RedSeer, Redseer), outdated, or simply do not exist for niche categories. Government data (MCA filings, MSME registries, DPIIT reports) exists but requires significant cleanup.

Three practical approaches that work:

Proxy metrics. If you cannot find a direct market size report for “cloud kitchen management software in India,” find adjacent data. How many FSSAI licenses were issued in the last year? How many Swiggy/Zomato restaurant partners are there? What percentage of restaurants use any digital tool for operations? Build your estimate from public proxies rather than waiting for the perfect report.

Channel-back sizing. Start from your distribution channel and work backwards. If you sell through Google Ads, your reachable market is bounded by search volume for your category keywords. Use Google Keyword Planner to find monthly search volume for “project management tool India” or “invoice software GST.” Multiply by your expected click-through rate and conversion rate. This gives you a ground-truth SOM that no Statista report can match.

Competitor revenue triangulation. For Indian SaaS companies, you can often estimate revenue from three public signals: employee count (from LinkedIn), funding history (from Crunchbase or Tracxn), and pricing (from their website). A 200-person SaaS company with Series B funding and $15/user/month pricing is doing roughly $3-6 million ARR. If there are four such competitors, the served market is at least $12-24 million. This is imprecise but directionally useful.

When to research and when to stop

Market research has a failure mode that is just as dangerous as doing none: doing too much. A PM can spend three months building the perfect competitive landscape and sizing model while the engineering team sits idle and the market moves.

The rule: research until you can make the next decision, then stop. If you are deciding whether to enter a market, you need TAM/SAM/SOM. If you are deciding which feature to build next, you need competitive differentiation analysis and user churn data. If you are deciding on pricing, you need willingness-to-pay data and competitor pricing. Match the depth of research to the decision it informs.

Research that does not lead to a decision within two weeks is academic. Put it down and ship something you can learn from.

Test yourself

Where to go next

• Turn research into strategy: Product Vision & Strategy
• Validate with real users: User Research Methods
• Apply sizing to pricing decisions: Pricing Strategy
• Build the business case: Business Model Thinking