Startup glossary

An experiment that shows two versions of something to comparable groups and measures which one performs better on a single metric.

An A/B test splits traffic or users at random between two versions of a page, message or price and compares one predefined metric, such as sign-ups or paid conversions. It only tells you something if the sample is large enough for the difference to be unlikely to be chance, so early-stage founders with little traffic often get noisy results. Decide the metric and the sample size before you start, and do not stop the test the moment one version pulls ahead.

The first narrow, reachable segment you aim to win completely before expanding into adjacent markets.

A beachhead market is a small, well-defined group of customers who share the same problem, buy in the same way and talk to each other, so winning a few of them helps you win the rest. It is chosen for reachability and urgency, not size. For example, rather than 'restaurants', a beachhead might be 'independent restaurants in one city that take phone bookings'. Once you dominate it, the product, references and word of mouth carry you into neighbouring segments.

Building a business with your own money and customer revenue instead of outside investment.

Bootstrapping means funding a business from personal savings, a day job or early customer revenue rather than raising venture capital or angel money. It keeps ownership and control with the founders and forces discipline: the idea must reach paying customers quickly because there is no cushion. The trade-off is slower growth and less room for ideas that need years of development before they earn anything, which makes bootstrapping a useful filter when comparing ideas.

Estimating a market by multiplying the number of reachable buyers by what each would realistically pay.

Bottom-up market sizing starts from countable units: how many specific buyers exist, how many you could plausibly reach, and what each would pay per year. For example, if there were 4,000 independent veterinary clinics in your region and each paid 1,200 a year, the market would be 4.8 million. It is more credible than top-down sizing, which takes a large industry figure and assumes a percentage, because every input can be checked and challenged.

The sales volume at which total revenue covers total costs, so the business neither loses nor makes money.

The break-even point is where revenue equals fixed costs plus variable costs. You find it by dividing fixed costs by the contribution margin per unit. As an illustration, if monthly fixed costs are 3,000 and each sale contributes 30 after variable costs, you need 100 sales a month to break even. Knowing this number early tells you whether an idea needs a realistic or an heroic level of demand to survive.

How much cash a business spends each month beyond what it brings in.

Burn rate is the net amount of cash a business loses per month. Gross burn is total monthly spending; net burn is spending minus revenue. If you spend 10,000 a month and earn 4,000, your net burn is 6,000. Burn rate matters because it sets how long you can keep going, and it is the number to watch when deciding whether to hire, cut costs or change direction. Divide cash in the bank by net burn to get runway.

The share of customers or revenue lost over a given period, usually measured monthly.

Churn rate measures how many customers cancel or stop buying in a period, divided by how many you had at the start. Revenue churn does the same with recurring revenue. If you start a month with 200 customers and 10 leave, monthly customer churn is 5 percent. High churn means you are refilling a leaking bucket: acquisition spend is wasted and growth stalls. Persistent churn early on usually signals a weak problem, the wrong customer, or a product that does not deliver.

Delivering the promised result to early customers by hand, openly, before building any software.

A concierge MVP delivers the outcome your product would provide manually, with the customer aware that a person is doing the work. A founder planning meal-planning software might first build weekly plans by hand for ten paying customers. It tests whether people value the result enough to pay and teaches you exactly which steps matter before you automate anything. It does not scale, and that is the point: it is cheap learning, not a business model.

The tendency to notice and remember evidence that supports what you already believe and to discount the rest.

Confirmation bias is the habit of seeking, interpreting and recalling information in a way that supports an existing belief. For founders it shows up as hearing polite interest as demand, asking leading interview questions, and dismissing the one prospect who said the problem was minor. The defence is structural: write down in advance what evidence would change your mind, ask about past behaviour rather than opinions, and have someone else read your interview notes.

Revenue from a sale minus the variable costs of making that sale, the amount left to cover fixed costs.

Contribution margin is the price of a unit minus all costs that rise with each unit sold, such as materials, payment fees, delivery and usage-based hosting. If a product sells for 50 and variable costs are 20, the contribution margin is 30, or 60 percent. It is the money each sale contributes towards rent, salaries and other fixed costs. Unlike gross margin, it explicitly includes every variable cost, which makes it the better input for break-even maths.

The percentage of people at one step who complete the next desired action, such as signing up or paying.

Conversion rate is the number of people who take a defined action divided by the number who had the chance to. If 500 people visit a landing page and 25 join a waitlist, the conversion rate is 5 percent. In validation it only means something when the traffic is the right audience and the action has a cost, such as entering a card or booking a call. Compare rates between your own tests rather than against generic benchmarks.

The total sales and marketing spend needed to win one new paying customer.

Customer acquisition cost is everything you spend to acquire customers in a period, including ads, tools, content and the cost of sales time, divided by the number of new paying customers in that period. If you spend 2,000 in a month and gain 40 customers, CAC is 50. Founders often understate it by leaving out their own time. CAC only means something next to what a customer is worth, which is why it is paired with lifetime value.

Structured conversations with potential customers to learn how they experience a problem before you build a solution.

Customer discovery is the process of interviewing people who might have a problem to understand how they currently deal with it, what it costs them and what they have tried. Good discovery asks about specific past events rather than pitching an idea or asking for opinions. The output is evidence: patterns in behaviour, workarounds, spending and urgency across several conversations. It comes before building and should be allowed to kill or reshape an idea.

The total gross profit you expect to earn from an average customer over the whole relationship.

Customer lifetime value estimates how much profit a typical customer generates before leaving. A simple version is average monthly revenue per customer, multiplied by gross margin, divided by monthly churn. As an illustration, 40 a month at 80 percent margin with 4 percent monthly churn gives an LTV of 800. Early-stage LTV is mostly a guess because you have not seen customers stay long enough, so treat it as a range and revisit it as cohorts age.

An early customer who commits time, feedback and often money to help shape a product before it is finished.

A design partner is an early customer, usually a business, who agrees to use an unfinished product, meet regularly and give detailed feedback in exchange for influence over the roadmap and favourable terms. The best design partners have the problem acutely and resemble the customers you want later. A partnership is stronger evidence when it involves a commitment such as a paid pilot or a letter of intent, rather than goodwill alone.

The repeatable way a product reaches the people who will buy it.

Distribution is how customers find, try and buy your product: search, communities, partnerships, marketplaces, outbound sales, referrals or an existing audience. Many ideas fail not because the product is poor but because there is no affordable, repeatable route to buyers. When evaluating an idea, ask where your first hundred customers already gather and how you would reach them without large ad budgets. A clear distribution advantage can matter more than a clever feature.

A customer who feels a problem so sharply they will try an unfinished product to solve it.

Early adopters are the first customers willing to tolerate rough edges, missing features and an unknown brand because the problem hurts enough, or because they enjoy being first. They have often already built a workaround, such as a spreadsheet or a patchwork of tools. Finding them is the practical goal of early marketing. Be careful not to over-fit to their requests, since later mainstream customers may need something simpler and more reliable.

Offering a feature or product that does not exist yet to measure how many people try to use it.

A fake door test places a button, menu item or pricing option for something not yet built and counts how many people click it. After the click, visitors see an honest message explaining it is coming and, ideally, an option to join a list. It measures interest in context, which is more reliable than asking. Use it sparingly and transparently, since misleading people repeatedly damages trust, and pair the click with a costlier signal where possible.

How well a founder's experience, access and motivation suit the specific market they are entering.

Founder-market fit describes whether the people building a business have an unfair advantage in that market: domain knowledge, a network of likely buyers, credibility, or a personal reason to persist through the slow years. A former logistics manager building software for warehouses has more of it than a generalist who read about the sector. It does not guarantee success, but it shortens customer discovery, makes early sales easier and reduces the odds of quitting when progress stalls.

A pricing model where a basic version is free forever and advanced features or higher usage are paid.

Freemium gives away a useful core product at no cost and charges for more capacity, features, seats or support. It can lower the barrier to trying a product and spread it through word of mouth, but free users still cost money to serve and support, and many never upgrade. It works best when the marginal cost of a free user is low and the paid tier solves a clearly bigger need. It is not a substitute for testing willingness to pay.

The plan for which customers you will sell to first, how you will reach them and how they will buy.

A go-to-market plan ties together the target segment, positioning, pricing, channels and sales motion for launching a product. It answers practical questions: who exactly buys first, what message makes them care, where you will reach them, whether they self-serve or need a sales call, and what it costs to win each one. A weak GTM plan is a common reason good products stall, so it should be drafted while the idea is still being validated.

Revenue minus the direct cost of delivering the product, expressed as a percentage of revenue.

Gross margin is revenue minus cost of goods sold, divided by revenue. For software, cost of goods sold usually includes hosting, third-party services and customer support; for physical products, materials, manufacturing and shipping. If you sell 10,000 worth in a month and delivery costs 3,000, gross margin is 70 percent. Higher margins leave more room to spend on acquisition and still make a profit, which is why investors and founders scrutinise it early.

Software built for a common function, such as invoicing or project management, used across many industries.

Horizontal SaaS serves a business function that exists in almost every company, regardless of industry, such as email marketing, accounting, scheduling or team chat. The potential market is large, but so is the competition, and products tend to compete on breadth, polish and brand. For a new founder, a horizontal idea usually needs a sharp angle, such as a neglected user type or a radically simpler approach, to avoid being one more option in a crowded list.

A precise description of the type of customer who gets the most value from your product and is easiest to win.

An ideal customer profile describes the organisation or person most likely to buy, succeed and stay: their role, size, industry, tools, triggers and constraints. 'Small businesses' is not an ICP; 'operations leads at 20 to 50 person logistics firms still scheduling drivers in spreadsheets' is. A sharp ICP tells you whom to interview, where to find them and what to say, and it makes it easier to say no to customers who will distract you.

A way of understanding demand by asking what progress a customer is trying to make, not what features they want.

Jobs to be done is a framework that treats a purchase as 'hiring' a product to make progress in a particular situation. Instead of asking what people want, you ask what they were trying to get done, what triggered the search, and what they used before. Someone buying a meal kit may be hiring it to avoid deciding what to cook on weeknights. Framing ideas around the job reveals unexpected competitors and the outcome customers actually judge you on.

A one-page template for summarising a startup idea's problem, customers, solution, channels, costs and revenue.

The lean canvas is a single-page business model template adapted from the Business Model Canvas for early-stage ideas. It asks you to fill in boxes for problem, customer segments, unique value proposition, solution, channels, revenue streams, cost structure, key metrics and unfair advantage. Its value is forcing explicit, testable assumptions onto one page so the weakest box is obvious. It is a thinking aid rather than a plan, and should be rewritten as evidence arrives.

A short written statement from a prospective customer saying they intend to buy if the product meets agreed conditions.

In early-stage validation, a letter of intent is a brief document in which a prospective customer states that they intend to purchase or pilot a product on outlined terms, such as price, scope and timing, once it is available. It is usually non-binding, so its value is as a signal of seriousness: a named person at a real organisation put their intent in writing. It is weaker than a paid pre-order but far stronger than verbal enthusiasm.

Customer lifetime value divided by customer acquisition cost, a quick check on whether growth pays for itself.

The LTV to CAC ratio compares what a customer is worth over their lifetime with what it costs to acquire them. If LTV is 900 and CAC is 300, the ratio is 3 to 1. A ratio below 1 means each new customer destroys value; a very high ratio can mean you are underinvesting in growth. At the idea stage both inputs are estimates, so use the ratio to see which assumption the idea depends on most rather than as proof.

The smallest version of a product that lets real customers get value and lets you learn whether the idea works.

A minimum viable product is the least you can build to test your riskiest assumption with real users, ideally paying ones. It is minimum in scope, not in quality: it should do one job well enough that people would be disappointed to lose it. An MVP can be software, a spreadsheet, a manual service or a single landing page with a checkout. The common mistake is treating it as version one of the full product and spending months on it.

A durable advantage that makes it hard for competitors to take your customers or copy your business.

A moat is whatever protects a business's margins and customers from competition over time: network effects, switching costs, proprietary data, brand, regulatory licences, exclusive distribution or cost advantages from scale. Features are rarely moats because they can be copied. Most new ideas start with no moat at all, which is fine; the useful question is whether the business could plausibly build one as it grows, or will always compete on price.

When a product becomes more valuable to each user as more people use it.

Network effects occur when each additional user makes the product more useful for others. Direct network effects appear in messaging apps, where more contacts means more value. Indirect or two-sided effects appear in marketplaces, where more buyers attract more sellers and vice versa. They can create a strong moat once established but make the start very hard, because the product is weak while the network is small. Ideas that rely on them need a plan for the empty early phase.

The single measure that best captures the value customers get from your product and predicts long-term growth.

A North Star metric is one number a team agrees best reflects customers receiving real value, chosen so that improving it tends to improve revenue and retention too. For a scheduling tool it might be appointments booked per week, not sign-ups. A good one is measurable, moves within weeks, and is hard to inflate without helping customers. It keeps a small team focused, but should be supported by a few guardrail metrics such as churn and margin.

The value of the best alternative you give up when you commit time or money to one option.

Opportunity cost is what you forgo by choosing one path over another. For founders the scarcest resource is usually time: a year spent on a mediocre idea is a year not spent on a better one, a job or a different market. It is why comparing several ideas side by side matters more than judging one in isolation. When an idea is merely 'not bad', ask whether it beats the next best thing you could be doing.

A shorthand for whether a product solves an urgent problem people must fix or offers a nice-to-have improvement.

Painkiller versus vitamin is a common way of describing urgency. A painkiller solves a problem that is costly, frequent or blocking, so people seek it out and budget for it. A vitamin offers a benefit people agree is good but can easily postpone. Vitamins are not doomed, but they are harder to sell and easier to cancel. A practical test is whether prospects are already spending money or time on a workaround today.

How many months it takes for the gross profit from a customer to repay what it cost to acquire them.

The CAC payback period is customer acquisition cost divided by the monthly gross profit each customer generates. If CAC is 240 and a customer pays 30 a month at 80 percent gross margin, the monthly gross profit is 24 and payback is 10 months. Shorter payback means growth needs less cash up front, which matters a great deal for bootstrapped businesses. If payback is longer than typical customer lifetime, the model loses money on every sale.

A deliberate change to one core part of a business, such as the customer, problem or model, based on evidence.

A pivot is a structured change of direction that keeps what you have learned while changing one fundamental element: the customer segment, the problem, the solution, the channel or the revenue model. Moving from selling to consumers to selling the same tool to businesses is a pivot. It differs from drifting because it is driven by specific evidence and tested like a new hypothesis. Pivoting repeatedly without that evidence is usually avoidance rather than strategy.

Taking payment or a firm commitment from a customer before the product is fully built.

A pre-sale is an agreement in which a customer pays, places a deposit or signs a purchase commitment before the product is delivered. It is one of the strongest forms of validation because money changes hands, which filters out polite interest. Common forms include discounted founding-customer plans, paid pilots and refundable deposits. Be clear about what is being sold and when it will arrive, and offer refunds if you cannot deliver as promised.

Evidence that a real group has a meaningful problem and that your proposed solution addresses it in a way they value.

Problem-solution fit is reached when you have evidence that a specific customer group experiences a painful, frequent problem and responds to your proposed solution with commitment, such as pre-orders, pilots or repeat use of a manual version. It sits before product-market fit: you have shown the problem and the approach are right, but not yet that you can deliver, retain and acquire customers profitably at scale. Skipping it is a common route to building the wrong thing well.

The point at which a product satisfies a real market so well that demand pulls it forward with little pushing.

Product-market fit describes a product that a well-defined market wants enough to buy, keep using and recommend. Signals include retention curves that flatten rather than fall to zero, customers who would be very disappointed to lose the product, organic referrals and sales cycles that shorten. It is not a single moment or a score; it is a pattern in behaviour. Before it, the job is learning; after it, the job shifts towards scaling what already works.

A group of customers who started in the same period, tracked over time to see how many keep using or paying.

A retention cohort groups users by when they started, for example everyone who signed up in March, and tracks what share are still active or paying in each following week or month. Plotting cohorts side by side shows whether retention is improving as the product changes and whether the curve flattens, which suggests a core of lasting users. Cohorts are more honest than total user counts, which can grow while most new users quietly leave.

How many months a business can keep operating before it runs out of cash at its current burn rate.

Runway is cash available divided by net monthly burn. With 60,000 in the bank and a net burn of 5,000 a month, runway is 12 months. It is the clock every early-stage business runs against, and it shapes decisions such as when to raise money, when to cut costs and how long to give an idea before pivoting. Recalculate it whenever spending or revenue changes, and plan major decisions well before it gets short.

The part of the total market your product can actually serve given its focus, geography and business model.

Serviceable addressable market narrows the total addressable market to the customers your product is built for and can reach. If the total market is all accounting software buyers worldwide, your SAM might be small UK-based construction firms that need job costing, because that is what you build and where you sell. SAM is more useful than TAM for planning because it reflects real constraints such as language, regulation, integrations and the segments you are choosing to serve.

The share of your serviceable market you could realistically win in the next few years.

Serviceable obtainable market is the portion of SAM you can plausibly capture in a defined period, given your channels, competition, sales capacity and budget. It should be built from the bottom up: how many customers you can reach per month, what share convert, and how many you keep. For example, reaching 300 prospects a month and converting 3 percent gives roughly 9 new customers a month. SOM is the figure that tests whether an idea can support you.

A low-cost experiment, often a landing page, that measures real demand for a product before it is built.

A smoke test presents a product as if it exists, usually through a landing page with a clear offer and price, and drives targeted traffic to it to see how many visitors take a meaningful action such as joining a waitlist, booking a call or entering payment details. Visitors should be told honestly that the product is not yet available. A smoke test is only as good as its traffic and its call to action: cheap clicks from the wrong audience prove little.

An idea that starts from a technology or product and then hunts for someone who needs it.

A solution in search of a problem describes an idea that begins with something the founder can build or finds exciting, such as a new technology, and only afterwards looks for customers. The warning signs are vague target users, difficulty naming what people do today instead, and interviews that keep drifting to the product rather than the pain. The fix is to reverse the order: find a specific, costly problem first and check whether your solution genuinely beats current alternatives.

Continuing with something because of what you have already invested rather than what it is likely to return.

The sunk cost fallacy is the tendency to keep investing in a decision because of the time, money or effort already spent, even when the evidence says to stop. Founders fall into it when they keep building an idea because they have spent a year on it, not because customers are responding. Past investment cannot be recovered either way, so the only relevant question is whether the next six months are the best use of your time.

The time, money, risk and effort a customer faces when moving from one product to another.

Switching costs include data migration, retraining staff, rebuilding integrations, contract exit fees and the risk that the new product disrupts daily work. High switching costs protect incumbents, which means a new product must be clearly better, not slightly better, to win customers away. Once you have customers, the same costs help you keep them. When evaluating an idea, ask what a customer has to give up to adopt it and how you could make switching easier.

The total annual revenue available if every possible customer for a product category bought it.

Total addressable market is the maximum revenue a product category could generate if it captured every potential customer. It indicates the ceiling of an idea rather than a forecast. Large TAM figures quoted from industry reports are easy to find and rarely meaningful for an early business, because no new product serves an entire category. Use TAM to check that an idea is not capped too low, then focus on the serviceable and obtainable markets.

Pulling one function out of a large, general product or platform and building a focused product just for it.

Unbundling is a source of ideas in which a founder takes one feature or use case from a broad product, such as a spreadsheet, a classifieds site or an all-in-one suite, and builds a dedicated product that does that one job much better for a specific group. The sign of an unbundling opportunity is many people bending a general tool to fit a specialised need with templates, workarounds and add-ons. The opposite, rebundling, combines several narrow tools into one.

The revenue and costs attached to a single unit of the business, usually one customer or one sale.

Unit economics describes whether one unit of business, typically a customer or an order, makes money on its own once all directly attributable costs are counted. The core inputs are price, variable costs, contribution margin, acquisition cost, retention and payback period. If each unit loses money, more growth only grows the losses. At the idea stage you can rough out unit economics with illustrative numbers to see which assumption the whole model rests on.

A clear statement of who the product is for, the problem it solves and why it is better than the alternatives.

A value proposition explains in a sentence or two what specific outcome a product delivers, for whom, and why it beats what that customer does today. A useful one is concrete and comparative: 'Reconcile your shop's card payments in minutes instead of an afternoon each week' says more than 'Smart finance for small business'. If you cannot write it without vague words, the idea or the target customer probably needs sharpening before you build.

Software built for the specific workflows of one industry, such as dental practices, construction or logistics.

Vertical SaaS serves a single industry end to end, handling its particular workflows, regulations, terminology and integrations. Examples of categories include practice management software for clinics or job management tools for trades. The total market is smaller than for horizontal software, but customers are easier to find, the product can fit tightly, and switching costs tend to be high. It suits founders with industry knowledge or access to a community of practitioners.

A narrow, compelling first product or use case that gets you into customers before expanding into a broader offering.

A wedge is a small, sharp entry point: one feature, one workflow or one segment where you can be clearly better than anything else and win customers quickly. Once inside, you expand into adjacent needs. A tool that starts by solving one painful report for finance teams, then grows into wider planning software, is using a wedge. A good wedge is easy to adopt on its own and creates a natural reason to buy the next thing.

The specific recent change that makes an idea possible or necessary today when it was not a few years ago.

Why now is the argument for timing: what has changed in technology, regulation, cost, behaviour or the market that opens a window for this idea at this moment. Examples of such changes include a new rule that creates a compliance burden, an infrastructure cost that has fallen, or a shift in how a profession works. A strong answer explains why incumbents have not done it yet. A weak answer usually means the idea has been tried and failed for reasons still present.

The highest price a customer would accept for a product, revealed by behaviour rather than stated opinion.

Willingness to pay is the maximum a customer would spend for a given outcome. People are poor at predicting it when asked directly, so it is best revealed through behaviour: what they currently pay for workarounds, how they respond to real price points on a checkout or proposal, and whether they accept a pre-order. Testing a few concrete prices with real offers tells you more than surveys, and helps decide whether an idea can sustain a business.

A test where customers use what looks like a working product while a person secretly performs the work behind it.

A Wizard of Oz test presents users with an interface that appears automated while a human handles the requests behind the scenes. A founder testing an automated bookkeeping assistant might process each uploaded receipt by hand. Unlike a concierge MVP, the user experiences it as a product, which tests whether the product experience itself is wanted. It is useful before investing in hard engineering, but think carefully about disclosure, privacy and data handling before running one.