Technology analysis blueprint
Tech analysis fails when investors lead with multiples and lag on the business. The right order is: understand the unit economics, stress-test the moat, then frame the valuation.
On this page
Overview
Most weak technology research starts with a P/S or EV/NTM revenue multiple and works backward. The strongest research starts with a clear-eyed view of growth durability, competitive position, and capital efficiency — then arrives at a valuation that can survive the next bad quarter.
Read this first
Write these prompts down
Full framework
The revenue architecture determines what growth is worth. A recurring dollar earned from an existing customer is worth more than a new dollar that required significant incentives to close.
Recurring subscription, usage-based consumption, transaction fees, hardware, and professional services all have different margin profiles, growth dynamics, and terminal value implications. A blended revenue mix needs a blended valuation framework, not a shortcut multiple.
Why it matters
The wrong revenue classification leads to the wrong discount rate, the wrong margin normalization, and the wrong comparable set.
When it matters
Before building any model, opening a comp table, or calling the stock cheap.
Investor take
Write one sentence on what the primary revenue engine is and what makes its economics durable or fragile before you assign any multiple.
Blended ARR growth mixes new logo additions with expansion and contraction from the existing base. If NDR is 90% and new logo growth is 40%, reported ARR grows but the business is leaking at the foundation. Separate the two streams and judge them independently.
Why it matters
NDR deterioration typically shows up in cohort behavior one to two quarters before it appears in reported ARR growth.
When it matters
Every quarter before the earnings call, and every time a bull thesis relies heavily on new logo momentum without addressing the existing base.
Investor take
If NDR is below 105% and the bull case depends on new logo growth, the math is harder than it looks — model the implied logo count required.
Mission-critical infrastructure with deep workflow integration survives budget pressure. Discretionary productivity tools with shallow integration get cut or replaced first. Customer switching costs and budget priority are the real moat, not the product roadmap.
Why it matters
Pricing power, renewal rates, and the durability of expansion all track closely to how central the product is to the customer's operating process.
When it matters
When evaluating whether NDR is structurally high or cyclically elevated, and when assessing what happens to renewal economics in a soft macro environment.
Investor take
Ask whether the product is in the customer's 'cannot turn off' category or the 'nice to have' category — then test the answer against actual renewal behavior.
A product with genuine network effects — where each additional user makes the product more valuable to all other users — compounds differently from a product that just has good features. Features can be matched by a well-funded competitor. Network effects compound over years.
Why it matters
Investors who mistake strong product differentiation for network effects pay for moats that will erode under competitive pressure.
When it matters
When the competitive moat analysis is the central part of the valuation thesis and justifies a premium multiple.
Investor take
Name the mechanism by which the product becomes more valuable as the installed base grows. If that mechanism is not concrete, treat the moat as product differentiation, not network effect.
Growth driven by partner incentives, reseller push, aggressive free-trial-to-paid conversion, or M&A may look like organic adoption in the income statement. Cohort analysis and customer acquisition source data — when disclosed — will surface whether organic motion is accelerating or if the business is buying revenue.
Why it matters
Inorganic growth inflates current-period ARR and leads to overconfident long-run extrapolation, then consensus cuts after the one-time driver laps.
When it matters
Before forecasting growth beyond the next four quarters, and when evaluating whether management is growing through discipline or spend.
Investor take
If the company discloses channel vs. direct mix, track the trend — a shift toward resellers often signals that direct sales efficiency is weakening.
The market often assigns a moat premium based on brand, market share, or management narrative. Real moat analysis asks whether customers actually cannot or will not leave — and why.
Switching costs are real when renewal rates are high even when price increases, when customers expand over time despite competing offers, and when the sales cycle to displace is consistently measured in years. Those are behavioral indicators. The product specification sheet is not.
Why it matters
Overstating switching costs is the most common error in software analysis. The market eventually discovers whether the moat is structural or assumed.
When it matters
When the valuation depends on a sustained revenue multiple that requires durable retention to justify.
Investor take
Look at gross revenue churn and net revenue churn trend over three or more years. A stable NDR at 120%+ with pricing power is evidence. A declining NDR is not a temporary aberration — treat it as a thesis signal.
A company that controls the workflow layer, the data layer, or the identity and authentication layer has a structurally different competitive position than one that adds value on top of infrastructure it does not own. The closer the business sits to a commodity layer, the more margin pressure is a structural risk rather than a cyclical one.
Why it matters
The competitive threat for businesses sitting above a commodity layer often comes from infrastructure providers extending upward, not just from point-solution startups.
When it matters
When analyzing whether a gross margin trajectory is structurally defensible or at risk from an infrastructure provider who decides to compete.
Investor take
Map the tech stack and identify which layers above and below the company's product are owned by potential competitors — then track whether those competitors are investing in the adjacent capability.
Management will always say pricing power is strong. What gross margin trajectory over three to five years says is harder to spin. If gross margins have held or expanded through product mix changes, cost inflation, and competitive pressure, the pricing conversation is earned. Flat or declining gross margins in a scaling software business deserve an explanation.
Why it matters
Gross margin compression in a high-growth software company often signals competitive intensity that is not yet visible in customer churn data.
When it matters
When a new competitor enters at a lower price point, when cloud infrastructure costs are rising, or when the business is shifting toward enterprise deals with more services attached.
Investor take
Do not accept blended gross margin improvement as pricing power evidence if the mix is shifting toward higher-margin segments while underlying segment margins are flat.
When competitive pressure intensifies, well-capitalized incumbents and challengers both accelerate R&D. If a business is holding R&D flat as a percentage of revenue while its largest competitors are increasing theirs, that asymmetry will show up in product gaps within two to four years.
Why it matters
R&D intensity is a leading indicator of where competitors are placing their bets, and therefore where the product roadmap of the business you own may face structural pressure.
When it matters
When evaluating whether a market leadership position is defensible over a three-to-five-year horizon and when a new competitor is receiving significant venture or public-market capital.
Investor take
Build a peer R&D intensity comparison — R&D as a percentage of revenue — and track whether the gap is narrowing or widening before you underwrite a long-duration thesis.
Technology competitive threats come from two directions simultaneously. Incumbents — Microsoft, Salesforce, ServiceNow, AWS, Google — extend existing platform relationships into new categories at low marginal cost. Startups move up from a wedge product toward enterprise. Most tech businesses that lose market share lose it from one of these two directions, not from a direct horizontal competitor.
Why it matters
Single-product businesses are more vulnerable than platform businesses when an incumbent decides the adjacent market is worth entering.
When it matters
When a hyperscaler or large platform company announces an investment in a product category adjacent to the business you are evaluating.
Investor take
Define the two most credible competitive threats — one incumbent extending down, one disruptor moving up — and track whether the evidence favors or disfavors those scenarios each quarter.
Unit economics tell you whether the business is actually worth a premium multiple. Most investors know the vocabulary but stop before doing the normalization work that makes the numbers honest.
Rule of 40 — growth rate plus FCF margin — is useful for filtering out businesses that are spending growth dollars inefficiently. A business at 20% growth and 5% FCF margin scores 25: that is not a software premium. A business at 20% growth and 20% FCF margin scores 40: the spending is working. Use it to force the conversation, not to auto-justify a multiple.
Why it matters
A high Rule of 40 score means the investment is productive. It does not mean the business deserves an unlimited premium — it means you should look harder at what happens when growth decelerates.
When it matters
When comparing capital efficiency across tech peers and when evaluating whether the current investment rate will produce FCF leverage at normalized growth.
Investor take
Build the Rule of 40 view at normalized growth — when the business is growing at 15-20% instead of 40% — and see whether the FCF margin structure holds.
Gross margin is the ceiling on the long-run FCF margin. A software business with 80% gross margins can eventually generate 30-35% FCF margins. A mixed hardware-software business with 50% gross margins cannot. The multiple should sit on the gross margin ceiling, not on current-period revenue growth alone.
Why it matters
The biggest valuation errors in technology happen when investors apply a pure software multiple to a business where gross margins are 50% or lower because of hardware, services, or infrastructure costs.
When it matters
Before assigning any multiple, and whenever the product mix is shifting or M&A has changed the margin profile.
Investor take
If two businesses have similar revenue growth but one has 75% gross margins and the other has 50%, the 75% gross margin business deserves a materially higher revenue multiple — quantify how much before comparing them.
Stock-based compensation is a real economic cost. It dilutes existing shareholders. Treating it as a non-cash add-back to produce an adjusted FCF figure that exceeds GAAP earnings is standard practice — and it systematically overstates owner cash. Subtract SBC from reported FCF before you declare the business profitable or compare it to a peer that expenses SBC more conservatively.
Why it matters
Companies with SBC above 15% of revenue are paying their talent with equity dilution that belongs in the cost structure, not in a footnote.
When it matters
When evaluating any technology business that reports 'non-GAAP' earnings or free cash flow prominently, particularly high-growth software companies that lean on equity compensation to retain engineers.
Investor take
Build your own SBC-adjusted FCF line and compare it to management's reported figure. The gap between them is the part of the story being hidden in non-GAAP presentation.
Annual billing in advance, one-time cash benefits from deferred revenue build-up during rapid growth, and lumpy contract prepayments can inflate reported free cash flow in ways that do not repeat at normalized growth. Investors who underwrite peak FCF margins as steady-state often find the math does not work when growth slows and billing cycles normalize.
Why it matters
Deferred revenue tailwinds are the most commonly misread FCF inflator in high-growth SaaS. The cash comes in today; the revenue is earned over the contract life.
When it matters
When FCF margins look unusually strong relative to a peer set with similar economics, and when a business is transitioning from rapid growth to a more normalized growth rate.
Investor take
Track the change in deferred revenue as a separate line. If reported FCF is materially higher than operating income plus D&A minus capex after adjusting for deferred revenue changes, investigate before accepting the margin.
TAM is management's argument for why the stock deserves your capital. Cohort retention is the business's demonstrated ability to hold and grow the value it has already created. Terminal growth assumptions should be tied to what the installed base can compound at, not to an optimistic share of a large and vaguely defined market.
Why it matters
DCF models anchored on TAM penetration produce valuations that are always bullish and never falsifiable. Models anchored on cohort retention produce valuations that can be updated as evidence arrives.
When it matters
When building a long-run DCF for a high-growth software business and when evaluating whether management's guidance implies durable growth or requires a step-function improvement in market penetration.
Investor take
Ask whether the terminal growth assumption requires NDR to hold, new logo growth to accelerate, or a new product to succeed. Each of those is a different risk, and at least one of them should show up as a downside scenario.
A technology thesis is complete when it names the variables that would force a revision. If the only update path is upward, the work is not done.
Every technology thesis has a multiple that the market is paying for a set of assumptions. Those assumptions have a most-important metric and a second-most-important metric. Name them both before sizing the position. If NDR drops from 125% to 110%, does the multiple compress from 12x to 8x? Know the answer before the print.
Why it matters
Most investors discover their multiple reset condition the quarter after it happens. Writing it down in advance changes how you size and how you react.
When it matters
Before initiating and whenever adding to an existing position in a technology stock with a premium multiple.
Investor take
Write the specific metric threshold and the multiple that attaches to a miss. Then decide whether current position sizing reflects the implied downside if that scenario plays out.
Technology spending accelerated dramatically during COVID and pulled several years of cloud, collaboration, and infrastructure adoption forward. Businesses that benefited from that pull-forward are now growing through a period where the installed base is digesting rather than expanding. That digestion is not a competitive loss — but the recovery timeline is measured in years, not quarters.
Why it matters
Misreading digestion as structural decline leads to premature selling. Misreading structural adoption as temporary pull-forward leads to holding through a multi-year reset.
When it matters
When a business is growing below its pre-2020 historical rate and management is attributing slowdown to macro rather than to a specific product or competitive dynamic.
Investor take
Build a baseline growth rate from the pre-pull-forward period. If current growth is above that baseline, assume some digestion is still ahead. If it is at or below baseline, the digestion is likely through.
CAC payback and NDR are the most paired unit economics metrics in software analysis. Payback tells you how efficiently the business acquires customers. NDR tells you how much value it extracts after acquisition. A business with 24-month payback and 130% NDR is a different machine than one with 12-month payback and 105% NDR. Both can score well on individual metrics but represent fundamentally different growth architectures.
Why it matters
When CAC payback lengthens and NDR holds, the business is finding expansion value even as acquisition becomes more expensive. When both move against you simultaneously, the thesis is under real pressure.
When it matters
Every quarter, and particularly during periods of sales team reorganization, new product pushes, or when the competitive environment is visibly intensifying.
Investor take
Track both metrics in a single view and build a combined efficiency score. If the trend on either metric is worse than it was 12 months ago, make that deterioration explicit in the thesis rather than footnoting it.
Remaining performance obligations — RPO — and backlog disclose contracted revenue that has not yet been recognized. A business with a large and growing RPO has revenue visibility. One with a flat or declining RPO may be pulling current-period revenue forward from future periods while presenting it as demand acceleration. Treat RPO trend as a leading revenue indicator and a sanity check on guidance.
Why it matters
Companies with high-growth guidance and flat RPO trends are either booking shorter-duration contracts — which reduces revenue visibility — or the forward pipeline is thinner than the number implies.
When it matters
When management provides multi-year growth guidance and when the quarter-over-quarter growth rate is meaningfully above the RPO growth trend.
Investor take
Build a simple bridge: if RPO grew 15% year-over-year and revenue grew 30%, ask where the 15-point gap is coming from — duration shortening, pull-forward recognition, or a genuine acceleration in new bookings.
Macro scenarios — recession, rate hikes, enterprise budget freezes — are real risks but they are shared across the sector. The specific bear case for a technology stock is competitive displacement or cohort deterioration that a competitor is causing. Write that bear case as if you work at the competing company and are trying to take the market.
Why it matters
Macro-anchored bear cases are the easiest to dismiss and the least useful for decision-making. Competitor-anchored bear cases are harder to build and more likely to surface the actual fragility in the thesis.
When it matters
Before sizing any position with a premium valuation multiple, and after any quarter where the business loses a named account to a named competitor.
Investor take
If you cannot write a credible 200-word bear case from a competitor's perspective, the moat analysis is not complete enough to justify a premium multiple.
Evidence
Unit economics scorecard
Every metric below is a lens, not a target. Read them together — a single strong number can hide three weak ones.
Valuation by revenue model
The most common tech valuation error is applying a SaaS multiple to a company with hardware-dependent or services-heavy revenue. Match the lens to the economics.
| Revenue model | Primary valuation lens | Key metrics to anchor | Common mistake |
|---|---|---|---|
| Pure SaaS / ARR | EV/NTM Revenue or EV/NTM ARR | NDR, gross margin, Rule of 40, CAC payback | Paying a 15x multiple on ARR that is 60% professional services — services revenue deserves no software multiple. |
| Usage-based / consumption | EV/NTM Revenue with NDR as the quality gate | Net expansion rate, cohort cohesion, gross margin by product tier | Extrapolating linear growth on a nonlinear consumption model without understanding whether the expansion driver is sticky or event-driven. |
| Marketplace / platform | EV/Gross Profit or EV/Revenue with take-rate disclosed | Take-rate trajectory, GMV quality, buyer-seller NPS as a proxy | Valuing gross merchandise value instead of net revenue — the multiple should sit on what the platform actually earns, not what flows through it. |
| Hardware + software (mixed) | Blended: EV/EBITDA for hardware, EV/Revenue for software attach | Software attach rate, recurring revenue mix, gross margin by segment | Applying a pure software multiple to a business where 60% of revenue is hardware at 20% gross margin — the blended economics do not support it. |
Common mistake
Apply and continue