Category Design for Startups: Create a Category or Borrow One?

Category design for a startup is the choice between inventing a new market category the buyer does not yet carry and differentiating inside a category the buyer already uses. The default is borrow: April Dunford’s Obviously Awesome argues that category creation is expensive and usually the wrong first move, and Geoffrey Moore’s Crossing the Chasm separates the early-market buyer (evaluating against status-quo or DIY) from the mainstream buyer (evaluating against the incumbent category leader). Creation is warranted only when three tests pass: no existing category actually fits the buyer’s job, the founding team has budget to educate the market for as long as adoption takes, and the ecosystem supports the frame the founder wants to introduce. The peer-reviewed record — Taeuscher and Rothe’s 2024 SMJ paper on framing and category dynamics, Erik Stam’s 2015 European Planning Studies paper on entrepreneurial ecosystems, and Saras Sarasvathy’s 2001 AMR paper on causation and effectuation — gives founders academic ground for treating category creation as the exception, not the default.

What “category design” actually means for a startup

Category design is not branding and it is not positioning; it is the founder-level choice about which market category the buyer should use to place the product. Branding is what the buyer feels when they encounter the visual and verbal identity. Positioning is the choice, within a category, of which alternative you replace. Category design sits above both — it decides which mental container the buyer opens before positioning ever gets to work. When the container is wrong, both positioning and branding are downstream investments in the wrong lens.

The reason the distinction matters in 2026 is that “invent a new category” has become the default advisor recommendation for founders under pipeline pressure, and the borrow-first alternative rarely gets its fair hearing. The editorial reading is that category is the container decision, positioning is the alternative-choice inside that container, and branding is the surface treatment; treating them as one decision produces refresh programmes that touch the surface without ever settling the container.

Why borrowing the category is usually the right first move

Buyers evaluate products against alternatives they already know how to compare. Moore’s chasm is the well-known illustration: an early-market buyer evaluates against status-quo or a DIY workaround, while a mainstream buyer evaluates against the incumbent category leader. Borrowing an existing category means the buyer’s comparison lens is already loaded — the founder is competing on unique attributes and value inside that container, not on the meta-question of what the container is. That is a strictly narrower and more winnable conversation.

Creating a new category means the founder has to teach the buyer three things in sequence — the container exists, the container matters, and the founder belongs in it — before any positioning claim registers. Dunford’s argument is that this three-step teach is expensive and slow, and the founding team almost always underestimates both the budget and the time it takes to move a buyer’s category vocabulary. The safer default is to borrow the container, differentiate inside it, and revisit category creation only when the three tests below are clearly true.

The three tests that justify creating a category

Test 1 — No existing category actually fits the buyer’s job

Ask twenty buyers what they searched for the last time they had the problem your product solves. If the vocabulary is consistent and matches an existing category name — “CRM”, “observability tool”, “customer data platform” — the container already exists in the buyer’s head and the correct move is to borrow it. Only when buyers use language that does not map to any existing category — a novel workflow they describe in narrative rather than category terms — has the first test genuinely passed. Buyer language is the evidence, not the founder’s belief that the category is new.

Test 2 — The founding team has budget to educate the market

Category creation is a multi-quarter education programme. The founding team needs runway to publish the primer content, run the analyst briefings, absorb the “we don’t understand what this is” objections in sales calls, and keep positioning stable while the buyer’s vocabulary catches up. If the runway comfortably supports the education spend on top of product and go-to-market, the second test has passed. If the education spend is being funded by narrowing product investment or by pushing the next round earlier, the test has not — borrow the category and revisit later.

Test 3 — The ecosystem supports the frame

A category name that no analyst covers, no adjacent vendor references, and no community discusses is a founder invention living in one deck. Before creating a category, the founder should be able to point to signals in the ecosystem — a related community conversation, an adjacent-vendor product page that gestures at the same job, an analyst who has begun clustering vendors similarly, a peer-reviewed paper that names the mechanism — that show the frame is already in the air. When those signals exist, the founder is naming something that is emerging; when they do not, the founder is inventing something the market cannot yet hear.

Creation test	Founder question	Evidence that satisfies it	Failure mode
No existing category fits	What do buyers say they searched for when they had this problem?	20 buyer conversations, none of which name an existing category	Founder projects “no category” from their own vocabulary, not the buyer’s
Education budget on top of runway	Can we fund a multi-quarter education programme without pulling product spend?	Runway comfortably covers the education programme in addition to product + GTM	Education spend is funded by pulling product headcount
Ecosystem signals present	Which analysts, communities or adjacent vendors are already gesturing at this frame?	Named community thread + named adjacent vendor + named analyst clustering	Founder cites their own deck as the ecosystem signal

The ecosystem context — Stam and Startup Genome as background

Erik Stam’s 2015 European Planning Studies paper argues that entrepreneurial ecosystems are best analysed as regional configurations of resources — talent, capital, mentorship, adjacent firms, policy — that jointly shape how categories emerge and how adoption spreads. Inside its own ecosystem-policy scope, the paper says the ecosystem is not a passive backdrop; it constrains which categories can plausibly be created and which categories are already primed. For a founder deciding whether to borrow or create, the practical takeaway is that the local ecosystem’s fluency with the proposed frame is part of Test 3 and is not a variable the founder alone controls.

Startup Genome’s Global Startup Ecosystem Report 2025 is a macro reference for that same picture at the global level: the report describes hundreds of startup ecosystems and their comparative dynamics, and is useful for orienting the founder to what “an ecosystem that supports category creation” looks like at scale. Read it as background, not as a benchmark for any single startup; a founder’s category decision is answered by their specific buyers and their specific ecosystem, not by an aggregate ranking.

What the peer-reviewed evidence adds

Karl Taeuscher and Hannes Rothe’s 2024 Strategic Management Journal paper (“Entrepreneurial framing: How category dynamics shape the effectiveness of linguistic frames”) studied ventures in the MOOC market from 2012 to 2021 and argues that framing effectiveness depends on category dynamics — higher category salience amplifies social-impact framing while dampening innovativeness framing. Read inside its own MOOC-market sample, the paper is the most direct peer-reviewed evidence that a category’s dynamics govern which frames work. For a founder deciding to borrow or create, the editorial reading is that the effective frame is a property of the category the buyer already carries, not of the founder’s copy.

Saras Sarasvathy’s 2001 Academy of Management Review paper contrasts causation (goal-first) and effectuation (means-first) and argues that experienced founders often shape markets through effectual action from means they already control. For category design, the editorial reading is that a founder who lacks the budget for full-scale category creation is often better served by effectually iterating inside a borrowed category — testing, learning, and only naming a new container once the buyer conversations themselves start to demand it. Effectual category creation is emergent; deck-first category creation is aspirational.

Dean Shepherd and Marc Gruber’s 2020 Entrepreneurship Theory and Practice paper synthesises the lean-startup framework into five interconnected components — opportunity identification, hypothesis formulation, hypothesis testing, iteration, pivoting — within the studies they consolidated. For category work, the paper’s contribution is procedural: the workflow that safely tests a category hypothesis is the same as the one that tests any other hypothesis. Category design is a series of hypotheses about buyer vocabulary, not an act of naming.

Category discoverability in 2026 — what Google actually says

A category decision is also a discoverability decision, and 2026 buyers do not only reach a category page through a classic organic ranking. Google Search Central’s AI features documentation states that no special optimisation is required to appear in AI Overviews or AI Mode — pages must be indexed and eligible to be shown with a snippet, and the same helpful-content practices apply. What this means for category design is direct: an invented category name that no one searches for is not going to be surfaced by AI features, because the query pattern that would trigger the surface does not yet exist in the buyer’s vocabulary. Borrowed categories, by contrast, already carry query patterns Google recognises.

Google Search Central’s introduction to structured data reinforces the point: structured data helps Google understand what a page is about, but Google explicitly frames it as an aid to understanding, not as a ranking or rich-result guarantee. A category-creation founder who marks up their invented category with heavy schema is still asking the buyer to search for a name they do not yet carry; the structured data is doing the technical work but the semantic work still belongs to the market. The discoverability implication of category choice is downstream of the buyer’s vocabulary, not upstream of it.

Discoverability path	Borrow-first category	Create-first category	Google Search Central alignment
Organic query patterns	Established; buyers already run them	None; the query pattern does not yet exist	Same indexing + snippet-eligibility rules apply per AI features docs
AI Overview / AI Mode exposure	Eligible once page is indexed and helpful	Absent until buyer vocabulary catches up	No special optimisation; helpful-content practices apply
Structured data role	Describes an already-known category clearly	Cannot manufacture a query pattern the buyer does not carry	Structured data is an aid to understanding, not a ranking guarantee per structured-data intro
Time to first surfaced impression	Weeks (existing queries, existing category pages)	Multi-quarter (buyer vocabulary must accumulate)	Discoverability follows buyer vocabulary, not markup depth

A 90-day borrow → prove → create sequence

Assume an early-stage technology startup with roughly twenty design partners, no dedicated category page, and pressure from an advisor to “own a category”. The scenario is illustrative — no real startup or founder is named, no follower / pipeline / conversion / funding-round numbers are invented, and every outcome expectation is stated as a band rather than a target number.

In the first month the founder inventories every existing category the buyer might use, runs twenty buyer conversations that ask what the buyer searched for the last time they had the problem, and records the actual vocabulary. In the second month the founder ships positioning and copy inside the most-matched borrowed category, publishes the primer content that helps the buyer differentiate inside that category, and instruments discovery so the query patterns can be observed. In the third month the founder evaluates the three creation tests against the fresh evidence: if all three are clearly true (buyer language now diverges from the borrowed category, education budget is real, ecosystem signals have emerged), a category-creation programme is a defensible next step; if any test is still aspirational, the correct answer is another quarter inside the borrowed category.

The point of the scenario is not a promised discoverability or pipeline outcome. The output at the end of month three is a written category decision — borrow-and-continue or create-with-evidence — with observable signals attached to each of the three tests, and a category page that reflects the decision rather than the advisor’s default.

What I would skip

The skip list follows the three tests. Each entry below either fails one of the three tests, replaces the founder-owned category decision with an advisor default, or spends scarce runway on category infrastructure the buyer will not use. “Skip” here means “for an early-stage technology startup with a small founding team”; a later-stage company with a proven borrowed category and real education budget could defensibly do some of these.

• Category invention as the first move: the borrow-first default exists for a reason; skipping it commits multi-quarter runway to teaching before earning the right to name.
• Analyst-report-driven category naming: if the analyst’s report labels a category the buyer does not use, the founder is inheriting an analyst artefact rather than the buyer’s language.
• Category-refresh masquerading as positioning fix: renaming the category does not settle competitive alternatives, unique attributes or value; the underlying positioning problem persists.
• Structured-data-first category launches: schema markup does not manufacture buyer vocabulary; borrowed categories already carry it.
• Founder-only ecosystem signals: citing the founder’s own deck as evidence that the category exists is a Test 3 failure with a fancy graphic.
• “Own a category” advice without the education budget: aspirational category ownership without runway to teach ends in half-baked category pages that neither borrow nor create effectively.
• Category page as the first marketing artefact: buyers land on solution and comparison pages first; the category page is downstream, not upstream, of the borrow-or-create decision.

Common mistakes I still see

Patterns that repeat around every category conversation

The mistakes below show up in nearly every category conversation I have with an early-stage technology founder. They persist because advisor decks reward “category ownership” as a narrative and because the borrow-first default sounds less ambitious than the invent-a-category pitch.

Mistake	Answering source	Safer alternative	Signal that catches it
Inventing a category before running Test 1	Dunford — Obviously Awesome	Interview 20 buyers on the actual search vocabulary first	Category name appears only in the founder’s deck, not in buyer transcripts
Skipping the education-budget test	Moore — Crossing the Chasm	Model the multi-quarter education spend before naming the category	Education spend is funded by pulling product headcount
Treating one analyst’s category label as ecosystem consensus	Stam 2015 (European Planning Studies)	Require three independent ecosystem signals before treating the frame as emerging	Only one named source uses the label; no community or adjacent-vendor uptake
Assuming schema markup creates category discoverability	Google — Structured data intro	Prioritise buyer vocabulary in on-page copy; use schema to describe, not to invent	Category page has heavy schema but no organic queries with the category name
Reading Startup Genome ecosystem rank as permission to create	Startup Genome GSER 2025	Use the report as background context; rely on local buyer conversations for the decision	The category rationale references only the ecosystem ranking, not local buyer signals

Read the table above with the current category rationale open on the second screen. The mistakes that look serious against the rationale are different from the ones the founding team would name from memory — and the safer alternative already sits inside the three-test sequence for a founder who has committed to it.

Macro context — background, not benchmark

Two macro references belong in the founder’s mental model while a category decision is being scoped — but only as background, never as a benchmark for a single startup. The World Bank’s New Business Density indicator measures new registrations per 1,000 people aged 15-64, and the underlying World Bank Entrepreneurship Database provides downloadable trend data from 2006 to 2024 for new limited-liability companies. Use it to orient to ecosystem competition for the founder’s frame; do not use it to justify the category decision itself.

The OECD’s Financing SMEs and Entrepreneurs 2024 report covers around 50 countries and documents a sharp increase in SME financing costs in 2022 with a decline in SME lending. The macro reading is that runway is tighter than in the previous cycle — a specific reason to default to borrowing the category and to reserve category creation for the case where all three tests are already true and the education budget is genuinely in hand.

Borrow the category first; earn the right to create one

Before the next category meeting, write down the current answer to each of the three tests with a piece of evidence next to it. If Test 1 is answered with “we think buyers don’t use an existing category”, the evidence is founder belief and the test has not passed — schedule twenty buyer interviews and rerun. If Test 2 is answered with “we’ll figure out the education budget as we go”, the evidence is a spreadsheet the founder has not built — model the spend before naming the category. If Test 3 is answered with “the analyst says so”, the evidence is one source — find two more independent ecosystem signals or treat the frame as not-yet-emerging.

Publish the three-line answer as an internal document with a date. If the honest reading is “borrow-and-continue”, the next quarter’s marketing plan reflects it — a category page that lives inside the borrowed frame, positioning that differentiates inside it, discoverability language that matches queries buyers already run. If the honest reading is “create-with-evidence”, the plan includes the education programme, the primer publishing cadence, and the analyst outreach that the multi-quarter category launch requires. Either outcome is defensible; the failure mode is the third one, where the founder claims category creation without either the evidence or the education spend.

Two quarters of this discipline separate startups that get pulled into every advisor’s category-invention pitch from startups that make category decisions from evidence. Borrow the category first; earn the right to create one — the borrow decision is the discoverability decision the buyer will reward first, and the create decision is one the founder earns only after two quarters of buyer conversations that demand it.

Sources and evidence

• April Dunford — Obviously Awesome (book)
• Geoffrey A. Moore — Crossing the Chasm 3rd Edition (HarperCollins)
• Startup Genome — Global Startup Ecosystem Report 2025
• Erik Stam — “Entrepreneurial Ecosystems and Regional Policy: A Sympathetic Critique”, European Planning Studies, 2015 (DOI 10.1080/09654313.2015.1061484, peer-reviewed)
• Steve Blank — Why the Lean Start-Up Changes Everything (Harvard Business Review, May 2013)
• Shepherd & Gruber — Entrepreneurship Theory and Practice, 2020 (DOI 10.1177/1042258719899415, peer-reviewed)
• Sarasvathy — “Causation and Effectuation”, Academy of Management Review, 2001 (DOI 10.5465/amr.2001.4378020, peer-reviewed)
• Taeuscher & Rothe — “Entrepreneurial framing: How category dynamics shape the effectiveness of linguistic frames”, Strategic Management Journal, 2024 (DOI 10.1002/smj.3554, peer-reviewed)
• Google Search Central — AI features and your website
• Google Search Central — Introduction to structured data
• World Bank — New business density (indicator IC.BUS.NDNS.ZS)
• World Bank Entrepreneurship Database
• OECD — Financing SMEs and Entrepreneurs 2024

FAQ

Is category design the same as positioning?

No. Category design is the choice of which market container the buyer should use; positioning is the choice, inside that container, of which alternative you replace. Both matter, and they sit on top of branding, but treating them as one decision produces the “rebrand as positioning fix” failure I see most often.

Should I invent a new category if my product is genuinely new?

Only if the three tests are all clearly true: no existing category actually fits the buyer’s search vocabulary, the runway comfortably supports the multi-quarter education spend, and the ecosystem shows independent signals of the frame emerging. If any test is aspirational, borrow the category and continue; the product being new is not, on its own, a category-creation trigger.

What if analysts are already using a name for the category?

A single analyst report is one ecosystem signal, not consensus. Look for two more independent signals — a community conversation, an adjacent vendor gesturing at the same job, or a peer-reviewed paper naming the mechanism. When three independent signals exist, the frame is emerging and creation is defensible; when only one exists, borrow and revisit next quarter.

Does AI Overview change the category-creation calculus?

Not in the direction founders often assume. Google Search Central states that no special optimisation is required to appear in AI features and that the same indexing and snippet-eligibility rules apply. An invented category name that no one searches for is still absent from AI Overviews because the query pattern that would trigger the surface does not yet exist in the buyer’s vocabulary. AI features reward borrowed categories with established query patterns first.

How long should I stay inside a borrowed category before revisiting creation?

Two quarters of visible traction inside the borrowed frame is the working rule I use. That gives the founding team enough evidence to re-run the three tests with fresh data — buyer conversations, observed discovery queries, ecosystem signals — rather than the founder’s initial hypothesis. Category decisions get sharper as buyer vocabulary accumulates, and two quarters is roughly the minimum window in which the accumulation stops being noise.

Do the peer-reviewed papers really apply to a technology startup today?

Within their limits, yes. Taeuscher and Rothe’s 2024 SMJ paper is bounded to the MOOC market 2012-2021 sample but its mechanism (framing effectiveness depends on category dynamics) is what generalises. Stam 2015 is a sympathetic critique of ecosystem policy; the founder implication is the mechanism (regional resources shape category emergence), not the policy conclusions. Sarasvathy 2001 is a theoretical contribution on effectuation. Read each paper inside its own scope.