Tech Investing Is Not a Gut Call: How to Build Your Own 100-Point Evaluation Framework

Before you read: This is general educational information and a practical orientation, not investment, legal, accounting, or tax advice; it promises no returns, no fundraising success, and no exit. The weights, scores, and thresholds below are illustrative — they exist to show a method, not to set a universal standard. For any real investment decision, deal term, or governance question, consult a qualified professional.

Two investors look at the same deal — one says invest, the other says pass. Most of the time it is not that one is smarter; it is that the weights in their heads are different, and neither wrote them down. Information in early-stage tech investing is never enough. A mature company has financials, an industry position, comparable valuations, and a long track record. An early-stage company usually has none of that — what you see is a founder, a product prototype, a few early-customer signals, and a market story that is still moving. In that setting, "I think this will work" and "I don't think so" are not judgments; they are impressions. And the biggest problem with an impression is that it cannot be checked, challenged, or audited afterward.

Breaking your instinct into a hundred-point evaluation sheet is not about pretending early-stage investing can be calculated — it can't, and any framework that claims otherwise is lying to you. The real use of a scoring sheet is that it forces you to say out loud how much weight you put on the market, how much on the team, how much on the terms — so you know what you are actually believing, what risk you are taking, and what you tend to miss. The score is never the answer; it is the start of a discussion. Its greatest value is that two people who reach different conclusions can lay out and compare their weightings, and six months later, when the deal has an outcome, you can look back and see which cell you got wrong, and why. Without that sheet, you cannot even review your own mistakes.

Six Dimensions, and Why These Weights

The framework works simply: first assign each of six dimensions a weight that sums to exactly one hundred; then, for each dimension, score it within its own weight according to the evidence in hand; add the six cells and you have the deal's total (out of a hundred). The weights below are only a reasonable starting point, not a correct answer — the real work is adjusting them to your sector, stage, and circle of competence.

Dimension	The question to ask	Example weight
Market / Timing	Is this a big enough shift? Is now too early, just right, or too late?	20
Founder / Team	Can this team learn, correct, and navigate uncertainty fast?	25
Product / Moat	Is the solution clearly better than the alternatives? Does the barrier come from technology, data, process, or channel?	20
Early Validation / Unit Economics	Is there customer behavior, willingness to pay, retention, or a unit-economics signal?	15
Deal / Risk Control	Are the valuation, terms, information rights, and headroom for the next round reasonable?	10
Portfolio / Post-Investment Value	Does this fit your portfolio logic? Can you actually help?	10

The team usually carries the highest weight, for a very practical reason: at the seed stage the market shifts, the product gets torn down and rebuilt, and the first business model is almost certainly wrong — the only variable present from start to finish is this group of people. So what you look at in the team is not how polished the resume is, but how they decide under incomplete information, how fast they correct after the market proves them wrong, and whether they will admit what they do not know. The trap in this cell is that founders easily learn to "perform candor" — volunteering a harmless small weakness to look honest — so what you chase is how they handle the problem that genuinely hurts, not whether they will own up to a minor flaw (for a deeper view, see Founder assessment: judgment, execution, and resilience).

Market and timing are often flattened into "is the market big enough," but for early-stage investing timing is frequently more fatal than size. A technology moving from hype toward usable, with customers starting to actually pay or pilot, is a completely different risk from one still stuck in media buzz and conference talk. Too early, and you burn your cash before the market is ready; too late, and the price already reflects everything everyone can see. What is really worth asking is "why now" — what cost has fallen, what regulation has loosened, what behavior has changed, that makes this doable today and not before.

In the product cell, what you look at is real differentiation, not how slick the demo is. Ask a sharp question: if a customer stitched it together themselves with general-purpose tools, could they reach seventy percent of your effect? If they could, the differentiation is thin. The barrier might come from technology, accumulated data, depth embedded in the customer's workflow, or understanding of a specific domain. One point is often misjudged here: relying on heavy customization is not necessarily bad — in hardtech and semiconductors, living off NRE (commissioned design) or custom engineering work early while accumulating platform and IP is a healthy and common opening move. The real red flag is when customization is forever the endgame — every new customer adds another pile of bespoke work that never finishes — while the team packages itself as a scalable SaaS.

Early validation means separating "someone is trying it" from "someone bought it, and stayed." Traction here (real, verifiable early progress) does not have to be large revenue — a paid PoC, repeat usage, retention, referrals, or a unit-economics signal all count. Unit economics is about "do you make or lose money serving each additional customer," most often read as the ratio of customer acquisition cost to customer lifetime value; at the seed stage you usually cannot compute a credible number yet, so what you look at is the signal and the direction, not a verdict. Be careful: a one-off trial contract is often a PR deal bought with an innovation budget, and does not mean a formal procurement budget will pick up behind it.

The last two cells are easily drowned out by enthusiasm. A good company is not the same as a good investment, and here many people need to break a counterintuitive point first: a higher valuation is not better. The valuation sets your entry cost; too high, and even if the company succeeds you will not return a meaningful multiple, and it raises the performance bar the next round has to clear — miss it, and you may face a down round. So what you look at is whether the valuation leaves room for the next round, whether the SAFE (an early-stage agreement that defers valuation to the next round, exchanging cash now for future equity) or priced-round terms are clear, and whether your information rights and pro-rata (the right to invest proportionally in the next round) are looked after (for the vocabulary of terms, see Which term-sheet clauses an angel should watch). And a single deal's score must finally be read back inside the portfolio — whether you have enough deal flow, can avoid the obviously unfit, and can double down on the few that turn out well; those are the things early-stage investing actually lets you control.

The Weights Are Not Fixed: Adjust by Stage, Sector, and Your Circle of Competence

That set of weights is not for you to copy — it is a starting point that moves with the situation. Three things should move it most: stage, sector, and your own circle of competence.

The earlier the stage, the more you should load the score onto team and market — at the seed stage almost every concrete number will change, and what you are really betting on is whether this group can steer the direction right. By the Series A you start to have revenue, retention, and unit economics in hand, and it makes sense to shift weight toward early validation and economics, because there is now data to verify; falling back on "bet on the people" logic then becomes an evasion. Sector decides which cell is life-or-death: in biotech and hardtech revenue arrives very late, and forcing a high weight onto traction makes you miss every good topic that needs time — those deals should put the score on technical risk, regulatory path, and IP. B2B software is the opposite: whether it sells in is the biggest uncertainty, so sales momentum and renewals deserve high weight. Consumer or platform plays return to market, timing, and retention. Last is your circle of competence — if you happen to know a sector, can help with introductions, or can plug the gap the team is missing, the post-investment value cell is worth more to you than to others and deserves added weight, because that is a variable you can genuinely control.

In other words, there is no "correct" weighting table — only one that is more honest about this stage, this sector, and this you. If you find yourself using the same numbers on every deal, that is usually not discipline; it is not yet having thought through where this particular deal is stuck.

Let's Score One

All the abstract talk about weights matters less than scoring one for real. Here is a de-identified example.

A B2B AI operations-tooling startup is raising a seed round: three enterprises are running trials, the demo works well, but only one has moved into small paid usage; deployment needs the customer's internal process to cooperate, and data quality directly affects the result; the valuation is on the high side for the stage, and several terms are still unsettled.

First set the weights. An investor who knows enterprise IT procurement and believes they can genuinely help with introductions might allocate (summing to one hundred): market 20, team 25, product 20, early validation 15, deal 10, portfolio 10. Then score cell by cell — each score is "how many points this dimension earns within its own allocation":

• Market 15 / 20: enterprise AI adoption really is at an inflection point, but "the market is turning" is not the same as "this company can capture it," so dock a slice.
• Team 16 / 25: the founder is technically strong and can personally sell the first few customers — a good signal; but no one on the team can yet scale sales, and the hardest gate in B2B is exactly selling into large organizations. This is a hireable gap, not a death sentence, so dock points without cutting to the bone.
• Product 12 / 20: the demo is slick, but if a customer could stitch together seventy percent of the effect with a general-purpose model, the differentiation does not hold — mid-to-low.
• Early Validation 7 / 15: one paying customer is a signal, but the key question is whose budget — a business lead funding it from their own budget is very different in value from a one-off innovation-experiment budget; stay conservative until it is clarified.
• Deal Terms 5 / 10: the valuation is high, which raises the performance bar the next round has to clear, and the terms are unsettled.
• Portfolio / Post-Investment Value 8 / 10: you happen to know this sector and can help with the introduction window, so this cell is worth more to you than to others.

The six cells add to 63 (out of a hundred). Swap in a different investor who treats "no one can scale sales" as near-fatal and gives the team only 9, then pushes the deal down to 3 over the high valuation and the product to a more conservative 9 — the total drops to 51, below their action threshold, and they pass outright. Neither investor is wrong — the only difference is how much weight each put on the "can't sell into large enterprises" risk, and how deep they docked that cell. The point was never 63 versus 51; it is whether you can point at each cell and say: why I gave this score, relative to what, and what new evidence would make me change it. When the founder next tells you "a second customer moved into paid usage, and the business lead funded it from their own budget," you immediately know to revise the early-validation cell upward — that is the entire point of having a scoring sheet.

The Action Threshold and the Deal-Breakers

Weights alone are not enough; you also need two lines. The action threshold is "how high the total has to be before I will invest more time and move into deeper due diligence or co-investment discussion" — some use a lower bar to start observing, others demand more certainty before moving. The score itself is not the point; the point is that you know why, past this score, you are willing to take on more time and risk.

A deal-breaker is a different thing: it is not part of the sum but a single-issue veto — some risks cannot be touched even when the total looks beautiful:

• Core IP or equity ownership is unclear.
• The founder avoids talking about key risks and does not disclose honestly.
• The market story is big, but the first customers and the path to payment cannot be explained.
• The terms make you carry the risk with no matching upside.
• The team props up revenue with heavy customization while packaging itself as a scalable SaaS.

Keeping deal-breakers outside the sum matters because a weighted total has a dangerous property: it lets one bright enough highlight quietly compensate for a fatal flaw that should have made you walk away. Deal-breakers exist to stop exactly that.

Screen Fast First, Then Evaluate Deep

The hundred-point framework is for the "deep evaluation" — you only run the full version on deals worth the time. But an angel sees far more deals in a year than they can dig into, so you need a "fast screen" in front: five to seven weighted questions answered only with "yes / no / a single number," deciding in a dozen minutes whether a deal is worth a full evaluation. A fast-screen card might look like this:

• Can the founder say in one sentence "why now"? (high weight)
• Has anyone ever paid for this solution, or crossed a key milestone in a regulated domain? (high weight)
• Is the product's edge over existing options "clearly noticeable" or "a twenty-percent tweak"?
• Does the market ceiling match your exit script? (acquisition, a small-to-mid IPO, or going global each set completely different bars.)
• Is the valuation so off that the next round almost certainly cannot connect?

Any "no" on a high-weight question eliminates the deal outright — no need to run all six dimensions. But these are only examples; fast-screen questions must follow your sector and exit script. For deals where revenue arrives very late, like biotech and hardtech, the fast screen should swap in technical milestones, IP positioning, and regulatory-path feasibility, rather than "has anyone paid" and "several times better" — otherwise you will, in ten minutes, eliminate exactly the good deals that need the most time. The fast screen and the deep evaluation share the same dimensions; the only difference is that the fast screen decides "whether to look deeper" and the deep evaluation decides "whether to act." A high score only means "worth looking deeper" — the real decision still returns to the full framework, term-sheet review, and professional advice.

Make the Framework Your Own

Before your next pitch, write down on paper how you split your hundred points; afterward, go back and ask yourself three things: was the dimension I weighted most heavily actually supported by the data? Did I let one highlight inflate the overall score? Did I miss any deal-breaker? If you want to turn the framework directly into live questions, these work well: what is the single biggest upside, in one sentence? If it fails, where is it most likely to die? What is the strongest evidence and the weakest assumption right now? If you could only verify one more thing, which would most change your judgment? The point of these questions is not to make the founder perform a complete answer, but to help you decide the next step — pass, observe, gather more data, or move into deeper due diligence.

Be honest about this method's limits too. It is not a mechanical scoring sheet — a high total does not mean automatic invest, a low one does not mean automatic pass; the score only makes the discussion clearer. It also should not run on one shared weighting for everyone: someone fluent in B2B sales, someone fluent in hardware supply chains, someone who can offer an enterprise channel — different circles of competence should weight differently, and forcing someone else's weights only makes you fake confidence where you can't actually read the deal. The framework does not form in one pass; it grows out of every deal seen, every follow-up question, every failure and correction — and its real product is not a score, but finally being able to say clearly why you invested, and why you didn't.

Sources

• The Almanack of Naval Ravikant — Investing
• Gartner Hype Cycle
• Y Combinator SAFE Financing Documents
• LAUNCH Fund 4

This article cites external material for general educational reference; different markets, deal terms, and individual situations may warrant different judgments, and formal investment decisions should still be verified independently.

Further Reading

• The angel investor learning path
• Due diligence basics
• Founder assessment: judgment, execution, and resilience