포스트

Startup & Product Glossary: Terms Every Solo Founder Should Know

A practical reference of business metrics, strategy concepts, product stages, and tech terms used in day-to-day startup work — organized for quick lookup.

Startup & Product Glossary: Terms Every Solo Founder Should Know

Running a startup solo means switching context constantly: metrics review in the morning, product decisions at noon, launch prep in the afternoon. This glossary collects the terms that come up most in daily standups, retrospectives, and strategy sessions — so you can read and write them without stopping to look things up.

Organized by domain. Jump to what you need.


1. Business Metrics

AbbreviationFull NameOne-line meaning
NSMNorth Star MetricThe single number that best captures your product’s core value delivery
OKRObjective + Key ResultsQuarterly goal (O) + 3 or fewer measurable outcomes (KR)
KPIKey Performance IndicatorOperational metrics tracked more frequently than OKRs
MRRMonthly Recurring RevenueMonthly subscription revenue — excludes one-time payments
ARRAnnual Recurring RevenueMRR × 12
DAUDaily Active UsersUnique users active in a given day
MAUMonthly Active UsersUnique users active in a given month
LTVLifetime ValueTotal revenue a single user generates over their lifetime
CACCustomer Acquisition CostTotal cost to acquire one paying customer
TAMTotal Addressable MarketMaximum size of the market you could serve
WTPWillingness to PayThe price threshold validated before building (₩9,900? ₩30,000?)
D7 / D30 / D60 / D90Day 7 / 30 / 60 / 90Measurement checkpoints after launch for retention and KPI tracking
retentionPercentage of users who return after N days (e.g., D7 retention 30% = 30% still active 7 days post-launch)
funnelStep-by-step user conversion flow (e.g., install → first action → paid upgrade)
activationUser completes the core action at least once
conversionRate of moving from one funnel stage to the next (e.g., free → paid)
close ratePercentage of sales conversations that result in a purchase
ad-LTVAd-based LTVLTV driven by ad revenue rather than subscriptions — typically a hypercasual game model

2. Decision-Making & Strategy

TermMeaning
kill criteriaPre-defined conditions that trigger shutting down a project — defined before you start, so the decision is data-driven not emotional
kill signalA data point that satisfies kill criteria (e.g., D7: 0 of 5 users retained → wrong message or wrong ICP)
green signalA data point that clears a go-ahead threshold (e.g., ≥2 activations → proceed to next stage)
polish trapThe pattern of refining a product past the point of necessity before validating with real users — revenue-free polish is value-zero
pivotChanging direction after a hypothesis fails (e.g., B2C → B2B)
wedgeThe narrow, sharp entry point into a larger market
moatA sustainable competitive advantage competitors can’t easily replicate (data accumulation, network effects, brand)
flywheelA self-reinforcing feedback loop that accelerates over time (users↑ → data↑ → accuracy↑ → users↑)
bootstrapGrowing entirely on self-generated cash flow without external funding
venture-scaleRevenue potential large enough to justify VC funding (typically hundreds of millions ARR ceiling)
anti-goalAn explicit “we will NOT do this” commitment — keeps scope from drifting
JTBDJobs To Be Done — the underlying goal a user “hires” your product to accomplish
MECEMutually Exclusive, Collectively Exhaustive — a framework for classification with no overlaps and no gaps
Go / Hold / DropProject status: continue / maintain without new features / retire
hold modeNo new feature development; monitoring and notifications only

3. Users & Market

TermMeaning
ICPIdeal Customer Profile — the archetype of your best first customer
gatekeeperSomeone who influences adoption without being the direct end user (e.g., a therapist recommending an app to a family)
endorsementA credible recommendation from an authority — more effective than broad cold outreach
cold outreachContacting strangers directly via DM or email with no prior relationship
warm introAn introduction through a mutual contact
pre-screenQualifying a potential user for ICP fit and WTP before investing time
D3 returnedUser returned within 3 days of first use — a strong early signal of genuine engagement

4. Product Stages & Launch

TermMeaning
P0Launch-blocking bug — must be fixed before shipping
P1Pre-launch fix if possible, not strictly required
P2Post-launch fix — deferred deliberately
ship-blockerAny bug or unresolved decision that prevents shipping (equivalent to P0)
pre-launchThe stage immediately before going live to real users
stagingA production-equivalent environment for final verification before deploy
prod / productionThe live environment real users access
listingThe product’s public page on an app store or marketplace
smoke testA quick end-to-end check that the critical path works
gauntletA full scenario-by-scenario QA pass before launch
TestFlightApple’s iOS beta distribution channel
betaLimited-access pre-release testing with real users
rollbackReverting to a previous version after a bad deploy
hot patchAn emergency fix deployed to production immediately after a critical issue

5. Technical Terms

TermMeaning
PWAProgressive Web App — web app that can be added to the home screen and works offline
Service WorkerBackground script responsible for caching and offline functionality in a PWA
OAuthStandard for delegating authentication to external services (Google, Shopify, etc.)
webhookA callback where an external service POSTs to your server when an event occurs
idempotencyProperty where sending the same request multiple times produces the same result as sending it once
dedupDe-duplication — removing repeated or redundant records
atomic updateA write operation that completes as a single unit with no race condition exposure
race conditionA bug caused by concurrent requests interfering with each other (e.g., credit double-spend)
funnel eventA tracked analytics event corresponding to a step in the conversion funnel
on-deviceProcessing that happens entirely on the user’s device — no data sent to a server
inferenceMsTime (in milliseconds) for a single ML model inference pass
p50 / p9550th / 95th percentile latency — more informative than averages for tail-latency issues
bundleThe combined output file produced by a bundler (esbuild, webpack, Rollup)
externalizeExcluding a library from the bundle and loading it at runtime instead — reduces bundle size
localStorageBrowser key-value store that persists across sessions
standalone modePWA launched from the home screen, without the browser’s address bar or tabs
flaky testA test that produces inconsistent results without any code change
dependencyA task ordering constraint: task A must complete before task B can start

6. Common Tools by Category

CategoryTools
Hosting / DeployVercel (web apps, Next.js, static sites)
Backend / DatabaseSupabase (PostgreSQL + Auth + Storage), Prisma (TypeScript ORM)
PaymentsShopify Billing API (app subscriptions), Stripe (direct payments), Gumroad (digital products)
AnalyticsPlausible (privacy-first web), Firebase Analytics (mobile)
Mobile DeploymentTestFlight (iOS beta), App Store Connect, Google Play Console
UI ComponentsShopify Polaris
Mobile FrameworkFlutter (cross-platform Dart)
Bundlersesbuild, Rollup
AI / ML RuntimesTFLite (mobile inference), ONNX

7. Quick Abbreviation Reference

AbbreviationMeaning
B2B / B2CBusiness-to-Business / Business-to-Consumer
GTMGo-to-Market — strategy for reaching customers
SaaSSoftware as a Service — subscription-based software
ASDAutism Spectrum Disorder
ABAApplied Behavior Analysis (therapy method for ASD)
EQEmotional Quotient — emotional intelligence
KSTKorea Standard Time (UTC+9)
carouselHorizontally scrollable UI element
modalAn overlay dialog that appears above the main content
toastA temporary notification that auto-dismisses (typically bottom or top of screen)
placeholderTemporary value standing in for a real one (e.g., __PROD_URL__)
prepaidPayment collected before service is rendered

How to Use This

This glossary is meant to be a quick lookup, not a reading exercise. In practice:

  • Metrics conversations: anchor everything to MRR or your NSM. Don’t mix in ARR until MRR is meaningful.
  • Strategy calls: state kill criteria before committing. “We’ll kill this if D7 is 0/5” is a decision; “let’s see how it goes” is not.
  • Bug triage: label every open issue P0/P1/P2 before a launch. Unlabeled issues always inflate scope.
  • User research: confirm ICP and WTP before building. One endorsed user beats ten cold-outreach contacts.

Keep the glossary alive — add terms as they appear in your work, and note why a term was added so future-you has context.

이 기사는 저작권자의 CC BY 4.0 라이센스를 따릅니다.