Porters Five Forces Analysis for VC Market Screening
Summary
A porters five forces analysis tests whether a market lets companies keep their profit. For a VC analyst, the useful version takes about 90 minutes: write a one-sentence market boundary, score buyers, suppliers, entrants, substitutes and rivals from 1 to 5 with a source next to each number, then state a verdict. The page does not approve a deal. It tells you which two questions to put to the founder.
A porters five forces analysis tells you whether a market can pay a fund back, before you spend a week on the team. For a VC analyst, the useful version takes about 90 minutes, produces one page, and ends with a verdict on structural profit, not a tidy diagram. Below is the version we would actually run on a Monday morning shortlist, with the limits stated up front.
The framework answers one question: who keeps the profit?
Michael Porter's original argument, published in Harvard Business Review, is that competition for profit comes from five places. Direct rivals are only one of them. The other four are customers, suppliers, potential entrants and substitutes.
Together they set the structure of an industry. Structure decides how much of the revenue a company can keep once everyone with bargaining power over it has taken a share. That is the part an analyst cares about.
Early-stage investing adds a twist. You are rarely scoring a mature industry. You are scoring the one that might exist in seven years, with a company that is a few months old. So the analysis is a set of hypotheses, each tied to evidence you can check this week.
Skip it if the market is genuinely new and no buyer has ever paid for the product. In that case you have no structure to analyze, only a bet on demand. Use a different tool for that.
Start by drawing the market boundary, because everything else depends on it
Most bad five forces pages fail at step zero. The analyst writes "fintech" or "climate software" and then every force looks moderate. A market that wide has no structure.
Define the boundary the way a buyer would. Who writes the cheque, for what job, and what do they do today instead? "Accounts payable automation for mid-size German manufacturers" has a structure. "Fintech" does not.
Write the boundary in one sentence and put it at the top of the page. If a partner cannot argue with it, it is too vague.
Rate each force from 1 to 5 and write the evidence next to the number
Scores without evidence are opinions. The format that survives an investment committee is a table with four columns: force, score, evidence, source. Keep each cell to a sentence.
Here is what to look for in each force for a typical B2B software company.
Buyer power
Count the buyers and check concentration. If three customers make up 60% of revenue, buyer power is high no matter how good the product is. Then check switching cost: how long would it take a buyer to rip the product out? Weeks means high power. A year of integration work means low.
Supplier power
For software, suppliers are cloud providers, data vendors and model providers. If the product depends on one API that can double its price, supplier power is real. Ask the founder what happens to gross margin if the largest input cost rises 40%.
Threat of new entrants
Look at what actually blocks a second team. Capital is rarely the answer in software. Regulation, proprietary data, distribution and trust are. For a European healthcare tool, a certification that takes 14 months is a better barrier than any feature.
Threat of substitutes
Substitutes are the force analysts underweight most. A spreadsheet is a substitute. So is an in-house hire, or a service firm. Ask what the buyer did last year with the same budget, then price that alternative.
Rivalry among existing players
Count named competitors, check how differentiated they are, and watch pricing pages over time. Three vendors discounting the same feature is a rivalry signal. A single vendor with a waiting list is the opposite.
Use the output to decide the next question, not to approve the deal
A completed page rarely says yes or no. It says where the next hour should go. A market with weak buyers, weak suppliers and high entry barriers is attractive on structure, and the diligence shifts to the team. A market with a strong force on one side tells you what the founder must have answered already.
We find three patterns useful.
One dominant force. The thesis lives or dies on it. Spend the call on that force alone.
Moderate everywhere. Structure will not carry the deal. Execution and timing will, so say so in the memo.
Strong rivalry plus strong substitutes. Usually a pass at seed, unless the company has a wedge no one else can copy.
Write the verdict as the first line of the memo section. Partners read the first line.
Where the evidence comes from, and which tools shorten the hunt
The framework is cheap. The evidence is where the hours go. Three sources do most of the work.
Buyer and entrant evidence sits in your CRM and your data provider. Past conversations with comparable companies are a record of who buys, how fast, and who tried to enter. A relationship CRM like Affinity keeps those notes searchable, which turns your own history into a source for the buyer power row.
Market size and funding activity sit in private-market datasets. PitchBook is the standard for counting recent rounds in a segment, which is a direct read on entrant pressure. Count rounds, not headlines, and note the date range.
Demand-side signals sit in web and traffic data. Similarweb estimates traffic and audience overlap, useful for checking whether buyers are splitting across vendors. Treat the numbers as estimates, because they are.
Then there is where the page lives. Keep it in a shared workspace such as Notion so every deal gets the same template and you can compare forces across deals later.
AI agents help with the collection step, not the judgment step. An agent can read 40 sources and return a list of named competitors with pricing in minutes. It cannot decide that a substitute is a real threat. That call is the analyst's job.
A worked example: a Zurich seed deal in invoice automation
Take a hypothetical seed company selling invoice automation to Swiss and German mid-size manufacturers. This is an illustration with estimated figures, not a real deal.
Boundary: accounts payable automation for manufacturers with 200 to 2,000 staff in DACH.
Buyers score 3. Customers are numerous, but each integration into an ERP takes about six months, which lowers their power after signing. Suppliers score 2. The product runs on commodity cloud and a document model that has several providers.
Entrants score 4. Two large ERP vendors could ship the same feature as an add-on, and a funded competitor raised a Series A last quarter. Substitutes score 4, because the main alternative is an offshore outsourcing team already on the payroll. Rivalry scores 3.
The verdict: structure is moderate to weak, driven by ERP vendors and outsourcing. The diligence question for the founder is clear. Why does the buyer choose a startup over the ERP add-on, and what does retention look like after month 12?
That is a useful page. It did not say pass or invest. It told the partner which two questions decide the deal.
Why the 1 to 5 scale beats a free-text summary
Free text hides disagreement. If one analyst writes "buyers have some leverage" and another writes "buyers are fairly weak", the partner cannot tell whether they differ or just phrase things differently. A number forces the argument into the open.
Keep the scale honest. A 1 means the force is almost absent. A 5 means it is the main reason the industry earns thin margins. Most forces in most markets sit at 2 or 3, and a page with five 4s is usually a sign the boundary is too wide.
Score twice if you can. Once on today's market, once on the market in five years if the company wins. The gap between the two columns is often the most interesting line in the memo, because it shows what the founder is betting on.
What founders get wrong when you ask about it
Founders tend to answer the rivalry question first and the substitutes question never. Ask them directly: what does your best customer do if your product disappears tomorrow? A concrete answer, with a cost attached, is worth more than a competitor matrix.
Also ask who the customer called before they called the company. That person or vendor is the real incumbent, and it is often not on the slide.
Five mistakes that make the page worthless
Scoring the company instead of the market. Strong product, weak market is a common trap. Keep the two apart.
Using today's structure for a seven-year bet. Write what changes if the product succeeds. Success attracts entrants.
Treating all five forces as equal. In most markets one or two dominate. Say which.
Skipping the date. A competitive structure from last year can be wrong. Date every row.
Ignoring complementors. Porter's model has no slot for platforms and partners, and in software they often decide the outcome. Add a sixth note if they matter.
That last point is the honest limit of the method. The framework is from 1979 and built for industries with clear edges. Platform markets blur those edges, so treat it as a first filter and not a full model.
What to do on Tuesday
Pick one deal on this week's shortlist. Write the one-sentence boundary, fill the table with sources, and stop at 90 minutes. Then put the verdict at the top of the memo and send it to one partner.
If the page changes the questions you ask the founder, the exercise paid for itself. If it only confirmed what you already thought, check that your evidence column is not empty.
Want the memo template we use for this? Download the memo sample and adapt the force table to your own fund.