Startup & GTM Terms

A working glossary of the strategy, unit-economics, go-to-market, and funnel vocabulary that recurs across the teardowns. Built for engineers: each entry is a tight definition and, where it helps, a one-line example or test — not an essay.

Strategy & positioning

Wedge — the one sharp capability you use to enter a market and earn a foothold before expanding. Narrow enough to actually win, expandable enough not to dead-end. Stripe: developer-friendly payments. Figma: browser-based real-time collaboration.

Beachhead — the first narrow segment you fully dominate before expanding outward. The wedge is the capability; the beachhead is the market slice it wins you.

Moat — a durable competitive advantage that makes you hard to copy or displace. Common types: network effects, switching costs, economies of scale, brand/trust, proprietary tech or data. Test: if a funded competitor tried to copy you, what stops them? If the answer is “just time,” it’s not a moat yet. The wedge gets you in; the moat keeps you in.

Defensibility — whether that advantage actually holds over time. A feature is a head start; defensibility is what stops it from being competed away.

Arcane — obscure and esoteric enough that mastering the domain takes specialized, hard-won knowledge most builders won’t acquire. In opportunity-hunting it’s a feature: high-dollar workflows that run on paper, fax, portals, and tribal knowledge are unglamorous enough that no AI-native company has shown up — the arcaneness is the moat, because the barrier that kept incumbents manual is the same one that deters new entrants.

Nascent — just beginning to exist or take shape — an idea, market, or category in its earliest, still-forming stage, but on a trajectory to grow larger and more established. In opportunity-hunting it cuts both ways: a nascent market means no entrenched incumbent and rules still unwritten (room to define the category), but also no proven demand, thin tooling, and buyers who don’t yet know they have the problem — you’re betting on the trajectory, not the current size.

Bespoke — custom-built for one customer or case rather than standardized for everyone. In software it’s usually a warning: bespoke per-customer work doesn’t scale and erodes margin (every deal needs new code), so the goal is to turn bespoke demand into a configurable product. The flip side — bespoke integrations or workflows competitors won’t bother replicating — can be a moat.

Incumbent — the established, dominant player already holding the market you’re entering — the one with the distribution, brand, and customer base a startup has to displace or route around. Their weakness is usually structural (legacy tech, channel conflict, slow release cycles), which is what a wedge exploits; the incumbent’s advantage is the moat you have to out-build.

PMF (product-market fit) — the product solves a real problem well enough that the market pulls it from you (usage, retention, word-of-mouth) rather than you pushing it.

ICP (ideal customer profile) — the precise customer type you’re built for: industry, size, role, pain. Sharp ICP makes targeting, messaging, and CAC efficient; a fuzzy one leaks money.

TAM / SAM / SOM — market sizing, narrowing down: Total addressable (everyone who could use it) → Serviceable (those you can actually reach/serve) → Obtainable (the slice you can realistically win near-term).

Unit economics & metrics

CAC (Customer Acquisition Cost) — what it costs to land one customer. Total sales + marketing spend ÷ new customers acquired in the period.

LTV (Lifetime Value) — total profit expected from a customer over the whole relationship. ≈ avg revenue per customer × gross margin × avg lifetime (or revenue ÷ churn rate).

LTV:CAC ratio — the headline health check. <1:1 loses money per customer (dead); ~3:1 is healthy for SaaS; >5:1 often means you’re underspending on growth and could acquire harder. Watch the trap: cheap early CAC from founder network/organic doesn’t scale — the ratio at scale is the real story.

CAC payback — months of revenue needed to recover CAC. Under ~12 months is good for SaaS; under 18 is acceptable.

ARR / MRR — annual / monthly recurring revenue: the predictable subscription run-rate (ARR ≈ MRR × 12). The default scoreboard for SaaS.

Gross margin — revenue left after the direct cost of delivering the service (hosting, inference, support). High margin is what makes LTV work — and why per-request inference cost matters so much for AI products.

Contingency fee — a pricing model where you’re paid only on a successful outcome, typically a cut of the money recovered or saved, with no fee on failure. Common in legal, insurance subrogation, and tax appeals. Upside: it bills from money found/saved rather than a budget line, so zero budget objection and instant ROI proof. Downside: revenue is outcome-dependent and lumpy, and you carry the cost of the losses. The recurring monetization alternative to subscription or per-transaction for “recovery” workflows.

Underwriting — assessing the risk of a deal and setting the price and terms to accept it: lending (credit, income, collateral → rate/limit), insurance (claim likelihood → premium), and securities (IPO/bond issuance → offering price). In fintech AI products it’s the core decision being automated — the model is the underwriter, so its accuracy maps straight to loss rates and margin.

Escheatment — the legal process by which property whose owner can’t be located — an uncashed check, a credit balance, a stale dividend, an unused gift-card balance — must be turned over to the state after a multi-year dormancy period. Any business holding someone else’s money is a “holder” obligated to track each item’s dormancy clock, attempt owner outreach (due diligence), and remit on a per-state deadline in a prescribed file format. Why it’s an opportunity domain: compliance is near-universal as an obligation but rare in practice (in California, ~2% of holders comply), and the penalty for missing it is a contingent-fee auditor with a 15-year lookback and authority to estimate liability — an arcane, high-dollar workflow that’s still mostly manual.

Churn — the rate customers leave. Logo churn = accounts lost; revenue churn = dollars lost. Low churn is the foundation of high LTV.

NRR / NDR (net revenue retention) — revenue from your existing customers a year later, including expansion and net of churn. >100% means you’d grow even with zero new logos — the strongest single SaaS health signal.

Burn rate / runway — cash spent per month / months of cash left at that rate. Runway = cash ÷ burn; it sets the clock on the next milestone or raise.

Magic number — sales efficiency: net new ARR ÷ prior-period sales & marketing spend. >~0.75 suggests spending more on growth pays off; well below means fix the funnel before pouring in money.

Go-to-market motion

GTM (Go-to-market) — the plan for how you actually reach and sell to customers: who you target, how you find them, how you price, and how the sale happens (self-serve vs. sales-led).

PLG vs. sales-led — product-led growth: the product itself acquires/converts users (self-serve signup, free tier, bottom-up adoption). Sales-led: reps drive deals top-down. Many AI startups run a hybrid.

Land and expand — win a small initial footprint (one team, one use case), then grow seats and usage inside the account. Cheap to land, expansion does the heavy lifting.

Expansion revenue / upsell — additional revenue from existing customers (more seats, higher tier, usage). The engine behind NRR >100%.

Design partner — an early customer who co-develops the product with you in exchange for influence, access, and pricing. Source of real requirements before you have a market.

North star metric — the single number that best proxies the value customers get (e.g. weekly active documents signed). Aligns the team on outcomes, not vanity counts.

Funnel & growth

Conversion funnel — the staged path visit → signup → activated → paid → retained. A leaky funnel brings traffic in but drops people before they convert; the fix starts with finding which stage leaks worst relative to benchmark.

Top-of-funnel / bottom-of-funnel — TOFU = awareness and traffic (lots of people, low intent); BOFU = the revenue end (fewer people, high intent). A TOFU/BOFU mismatch is heavy traffic that never reaches paid.

Activation — reaching the “aha” moment that makes the product click (e.g. sent first signature request). Signups that never activate are an activation problem, not an acquisition one.

AARRR (“pirate metrics”) — the funnel as five measurable stages: Acquisition, Activation, Retention, Referral, Revenue. A checklist for where to instrument and where to look for leaks.

Cohort analysis — group users by signup week/source and watch their behavior over time. Separates an acquisition problem (bad cohorts coming in) from a retention problem (good cohorts leaking out later).

Retention curve — the % of a cohort still active plotted over time. A curve that flattens above zero = genuine stickiness; one that decays to zero = no durable value yet.

Funnel diagnostics

When traffic rises but users/paid users don’t, the name of the problem depends on where it leaks — and the fix follows the location.

Traffic-quality problem — the wrong visitors arrive (bad targeting, misleading messaging) and convert near zero. High traffic, low conversion by source is the tell — a marketing problem, not a product one.

Monetization / paywall friction — people use the product but won’t pay (the active→paid step leaks). A pricing, packaging, or value-capture problem rather than acquisition or activation.