Market Research Methods

TAM/SAM/SOM Calculation

TAM — Total Addressable Market — represents the total revenue opportunity if you captured 100% of every potential customer globally. SAM — Serviceable Addressable Market — is the portion of the TAM you can realistically reach given your distribution channels, geography, and product capabilities today. SOM — Serviceable Obtainable Market — is the realistic revenue you can capture within the next three to five years given your team size, current growth rate, and competitive environment. Investors use all three numbers; the SOM is the one that drives their return calculations.

The bottom-up calculation is the only credible approach for early-stage companies. For a B2B SaaS product targeting mid-size HR teams at US companies: there are approximately 45,000 US companies with 100–1,000 employees, roughly 30,000 of which have dedicated HR departments. If your average contract value is $12,000 per year, the SAM is 30,000 × $12,000 = $360 million. The SOM at year three, capturing 1.5% of the SAM with a small sales team, is $5.4 million ARR — a concrete, defensible projection. Top-down calculations that cite Gartner reports and apply a percentage ("1% of a $50B market") are immediately recognisable as lazy and don't survive investor questioning.

Competitive Analysis

Competitive analysis serves two purposes: it prevents you from building in ignorance of adjacent solutions, and it forces you to articulate why your specific approach wins against alternatives that already exist. The worst competitive analysis slide in a pitch deck lists five competitors in a grid with checkmarks. The best competitive analysis identifies the two axes on which the market actually differentiates — price versus functionality, self-serve versus enterprise sales, single-feature versus platform — and maps every competitor and your own product clearly against those axes.

Use Crunchbase to understand the funding context of competitors: a competitor that has raised $50 million has a two to five year execution advantage but also a sales motion built around that burn rate, which may leave the lower end of the market underserved. SimilarWeb shows competitor traffic volumes and traffic sources, revealing whether they're growing organically or through paid acquisition. Ahrefs reveals which keywords competitors rank for organically — a company ranking for 10,000 product-specific keywords has built a distribution moat that takes years to match. Map this competitive intelligence against your own strengths before building your positioning; the goal is to find the segment where you win on the dimensions that matter most to that specific buyer.

Customer Segmentation

Segmentation is the process of dividing your potential customers into groups that share enough characteristics to be addressed with the same product and message. The most useful segmentation dimensions for B2B startups are industry vertical (retail, healthcare, fintech), company size (1–10 employees, 11–50, 51–250, 250+), geography (US, Europe, global), and buying process (self-serve credit card sign-up versus sales-led demo and procurement). Each combination of these dimensions produces a segment with different acquisition costs, different feature requirements, and different willingness to pay.

The mistake early-stage founders make is refusing to segment. "Our product is for everyone" means your product is positioned for no one — the messaging doesn't resonate, the acquisition channels don't focus, and the sales conversation starts from scratch with every prospect. Pick the segment where your product is already strongest, where you can acquire customers most cheaply, and where the lifetime value justifies the acquisition cost. The five-customer segmentation exercise: take your five best customers, identify three characteristics they share (industry, size, workflow), and write those down as your primary segment. Build everything else around serving that segment brilliantly before expanding.

Primary Research Techniques

Twenty customer interviews provide more useful product insight than two hundred survey responses. Interviews generate hypotheses — unexpected answers that reframe what you thought you knew. Surveys confirm hypotheses at scale — they tell you whether a pattern you discovered in interviews is universal or specific to the people you interviewed. Doing surveys before interviews produces questions that miss the most important dimensions because you didn't know what to ask. The sequence is always: interviews first, surveys second.

The three primary research techniques that generate the highest signal per hour are as follows. First, customer discovery calls — 30-minute conversations with potential customers focused entirely on their current workflow and pain, not on your solution. Second, observation sessions — watching a customer perform the task your product addresses in real time, noting where they struggle, hesitate, or open a new tab. Third, competitor teardowns — signing up for every relevant competitor, completing their onboarding, and documenting exactly where they're strong and where they're weak. Google Trends adds a quantitative layer to qualitative research: a search term trending upward over three years indicates a problem gaining urgency, which is a favourable signal for market timing regardless of what your interview subjects say.

Frequently Asked Questions

How accurate does a TAM/SAM/SOM calculation need to be for a pitch? It needs to be defensible, not precise. Investors expect uncertainty in market size estimates; they don't expect them to be wrong by an order of magnitude. Build the calculation from first principles using named data sources — US Census Bureau company counts, industry association membership figures, government employment statistics — so you can explain every assumption. If an investor challenges a number, walk them through the sources rather than defending the total.

What is the best free tool for competitive intelligence? SimilarWeb's free tier provides competitor traffic estimates that are directionally accurate for sites with more than 100,000 monthly visits. Google's free Keyword Planner shows search volume for competitive keywords. Crunchbase's free tier shows funding history for public companies. Twitter and LinkedIn advanced search reveal what competitors' customers are saying publicly. These four tools together provide a substantial competitive intelligence picture for $0.

How do I find customers to interview before I have a product? Post in communities where your target customers already spend time: industry-specific subreddits, LinkedIn groups for specific job functions, Slack communities for specific tools or workflows. Be honest that you're researching a problem, not selling a product. Offer a $25 Amazon gift card for a 30-minute conversation — most people will take it, and it signals that you respect their time. Cold LinkedIn outreach to potential customers with a specific, brief message about what you're researching converts at 15–25% for interview requests.

What should I do if my customer interviews reveal that I've targeted the wrong segment? This is useful information, not a failure. Document what segment your interviews are actually describing — who has the problem most acutely, who is most motivated to solve it — and reframe your target segment accordingly. It's better to discover you were targeting the wrong segment after 10 interviews than after 10 months of product development. The most valuable customer discovery interviews are the ones that change your mind.

How do I track competitive changes after my initial research? Set up Google Alerts for each competitor's name and key product terms. Follow competitors' engineering blogs and changelog pages. Join the communities their customers use and notice when people praise or complain about specific features. SimilarWeb and Ahrefs have change-tracking features that alert you when a competitor's traffic or keyword rankings shift significantly. Schedule a quarterly competitive review to update your analysis — the competitive landscape in most startup markets changes faster than annual reviews can track.