---
name: category_design
description: "Design a new market category or reframe an existing one: run the go or no-go gate, build a point of view, name the category, map the ecosystem, and sequence the launch. Use when a product does not fit its current category, buyers compare it to the wrong alternatives, or a founder asks about category creation, category design, Play Bigger, naming a category, or a lightning strike launch."
---

# Category Design

## Purpose

Help a team decide whether to create a category, and if the answer is yes, produce the actual work: a tested point of view (POV), a defensible category name, an ecosystem map, a category blueprint, and a launch sequence with dates and owners.

Category design here means the discipline described in Play Bigger (Ramadan, Peterson, Lochhead, Maney, 2016): align product design, company design, and category design at the same time. One leg alone does not hold.

Two research claims explain why teams try this, and both need limits stated out loud:

- The Play Bigger authors report that category leaders in their sample of US venture-backed tech startups founded 2000 to 2015 captured 76 percent of their category market capitalization. That is a sample result with survivorship bias, not a promise for any new attempt.
- Yoon and Deeken (Harvard Business Review, March 2013) found that 13 category-creating firms on the Fortune 100 fastest-growing list for 2009 to 2011 accounted for 53 percent of incremental revenue growth and 74 percent of incremental market cap growth in that list. A small, selected sample.

April Dunford supplies the counterweight: work positioning in order (competitive alternatives, unique attributes, value, best-fit customers, market category, trends) and choose category last. Create a category only when the existing frame hides the value and the team can teach a new frame. That gate runs first in this skill.

## Workflow

### 1. Gate: should this be a category at all?

Score the four gate questions 0 to 5. Use evidence, not hope.

1. Existing frame hides value: buyers file the product on the wrong shelf, compare it to the wrong alternatives, or route the deal to the wrong owner.
2. Structurally different: the difference is architecture, business model, or workflow, not a feature list.
3. Budget can move: a buyer can fund this from a named budget line, or create one, within one planning cycle.
4. Teaching stamina: leadership will fund market education for roughly two years without demanding short-term pipeline from it.

Gate rule: any question at 0 or 1, or a gate total below 10 of 20, stops the project. Fall back to one of these, in order: win head to head in the existing category, own a named subsegment, reframe the existing category. Record the fallback. A stopped category project is a good outcome.

Run the full scorer early and again before launch: `python3 bin/category_score.py scores.json` (see references for the input format, or `--example`).

### 2. Frame the problem before the product

Write the problem in the buyer language found in interviews, support tickets, and lost-deal notes. No product name in this step.

Deliver a one-page problem brief:

- Problem name in plain words, 2 to 5 words if possible.
- Who feels it first, and what triggers it (growth, scale, regulation, a new channel, a new buyer).
- Cost of inaction: time, money, risk, or missed outcome, with the source for each number. Cut any number you cannot source.
- Old way: how people cope today, stated fairly. Include doing nothing and spreadsheets as real alternatives.
- Why the old way cannot fix it: a structural reason tied to how it was built, not a claim that rivals are lazy or stupid.

Test: read the brief to five target buyers. At least three should name the problem back in their own words without prompting. If they nod politely and change the subject, the problem frame is weak.

### 3. Build the POV

A POV is a short story the whole company tells the same way. Draft it in this order, one short paragraph each:

1. The world changed: the shift that made the old way fail.
2. The problem that shift created, using the name from step 2.
3. What happens if nobody solves it.
4. The new way of thinking the category requires.
5. The future outcome for a buyer who adopts the new way.
6. Proof available today: one customer, one dataset, or one working workflow. No roadmap items.

Rules for the draft: plain words, short sentences, no product feature list until paragraph 6, named owner at executive level. The CEO or founder owns the POV in the original Play Bigger formulation. If leadership will not repeat it in public for a year, it is a campaign theme, not a POV.

Pressure tests:

- Swap test: replace your company name with a rival. If the POV still fits them, it is generic.
- Enemy test: the POV names an old way or an idea as the opponent. Attacking a named rival company reads as fear and ages badly.
- Two-minute test: a salesperson tells it in two minutes and the buyer asks a question about their own situation. Questions about your product come later and count less.

### 4. Name the category

Generate 20 to 30 candidate names before judging any of them. Aim for 2 to 3 words. Good category names usually name the problem, the outcome, or the new method. They do not contain the company name.

Score each finalist with this battery (pass or fail each):

- Say it once on a call: can the listener repeat it and spell it?
- Shelf test: can a buyer guess what belongs in it and what does not?
- Adjacent test: is it distinct from the two nearest existing categories, in writing, in one line each?
- Search test: does the phrase already mean something else dominant in search or in an analyst taxonomy?
- Extension test: can a second company join the category without sounding like a copy of you? A category only one firm can occupy is a brand tagline.
- Legal screen: flag obvious trademark conflicts for counsel review. This skill does not give legal advice and never clears a name legally.

Keep a naming log: candidate, battery results, buyer reactions from at least five interviews, and the reason the winner won. Adapted criteria from Marty Neumeier as commonly summarized: distinctive, brief, appropriate, easy to spell and pronounce, likeable, extendible, protectable.

Worked pattern: inbound marketing named the method (pull buyers in with useful content) and implied its opposite, outbound marketing. The name taught the POV every time someone said it.

### 5. Map the ecosystem

A category becomes real when other parties gain from its existence. Map five rings and name real organizations or roles for each, with what they gain and what you ask of them:

1. Buyers and superconsumers: the heaviest, most passionate users. Yoon's work notes a small share of buyers can drive a large share of sales, so find them first. They supply language, proof, and referrals.
2. Partners and integrators: firms that sell, implement, or bundle alongside you.
3. Builders: developers, creators, or consultants who can build on or around the category.
4. Curators: analysts, communities, publications, conferences, and award programs that sort markets for buyers.
5. Adjacent categories: neighboring categories that feed buyers into yours or receive them after.

For each ring, write one concrete first move with an owner and a date. A map with logos and no moves is decoration.

Salesforce pattern to copy in structure, not in scale: product first, then a partner and developer ecosystem around the platform in the mid-2000s (AppExchange, Force.com), so the category had more than one company selling it.

### 6. Write the category blueprint

One document, 3 to 5 pages, that holds the whole design:

- Problem brief (step 2) and POV (step 3).
- Category name, definition in one sentence, inclusion and exclusion rules: what counts as a member of this category and what does not.
- Category requirements: the 3 to 5 capabilities a real member must have. Write these so a fair rival could meet some of them. Requirements written so only you pass read as a rigged spec and buyers discount them.
- Magic Triangle check: what product must build, what the company must change (hiring, pricing, model, support), and what category work must happen, side by side. List any leg that lags and its fix date.
- Evidence bank: customer proof, data, and third-party voices, each with a source link and a date checked.
- Language guide: words to use, words to retire, and the old category terms to stop feeding.

### 7. Sequence the launch

Use phases, not a single announcement day.

Phase A, private pilot (weeks 1 to 6): test POV and name with 10 to 15 buyers and 3 to 5 partners. Track verbatim language adoption: do buyers use your problem name unprompted? Revise or stop here if they do not.

Phase B, internal mobilization (weeks 5 to 8, overlapping): train sales, support, and product on the blueprint. Every public surface gets the new language in the same window, or the market hears two stories.

Phase C, lightning strike (one concentrated window, usually 1 to 2 weeks): a single POV-driven moment that aligns the company and ecosystem at once, rather than thin spend across many channels. Pick one primary strike type per Category Pirates: Stunt (attention event), Fight (name the old way and challenge it in public), Scarcity (limited cohort or access), Functional (education event or transformation program), Culture (a community ritual or identity people join). Support it with the ecosystem moves from step 5 landing in the same window: partner posts, curator briefings, customer proof, one anchor asset (report, event, or dataset).

Phase D, develop (months 3 to 12): repeat strikes on a cadence, publish category proof quarterly, recruit members, and track share of search and share of voice for the category term itself, not just the brand.

Kill and pivot criteria, written before Phase C: for example, after two strikes and six months, if unprompted buyer use of the category term stays near zero in interviews, or pipeline still arrives only under the old category name, fold the work back into subsegment positioning and keep the POV as sales narrative. Dunford-style narrative still pays even when category creation does not.

## Output Contract

Deliver, in this order:

1. Gate verdict with scores and the fallback if stopped.
2. One-page problem brief.
3. POV draft (six paragraphs) with swap, enemy, and two-minute test results.
4. Naming log and winning name with battery results.
5. Ecosystem map table: ring, named parties, what they gain, first move, owner, date.
6. Category blueprint (3 to 5 pages) with requirements and Magic Triangle check.
7. Launch sequence: phases A to D with dates, owners, strike type, and kill criteria.
8. Scorer output file from `bin/category_score.py` attached or pasted.

## Operating Rules

- Every factual claim needs a source a reader can check. Market size, growth, and share figures get attributed to their author and sample. Unverifiable numbers get cut.
- Never present the 76 percent figure or the Yoon and Deeken figures as guarantees. State the sample and the limit in the same paragraph.
- Do not invent analyst categories, quotes, customer names, or dates. If a curator has not adopted the term, say the term is company-proposed.
- Attack old ways and ideas, not named rival firms or people.
- Category requirements must be fair on their face. If only one firm can pass, label it a product spec, not a category standard.
- No legal clearance from this skill. Trademark, naming, and comparative claims go to counsel.
- If the gate fails, ship the fallback positioning work instead and say plainly that category creation was the wrong move for now.
- Keep the POV in buyer language. Internal jargon in the POV is a defect, not a style choice.

See `references/cases-and-templates.md` for worked cases, the scorer input format, and copy-ready templates for the problem brief, ecosystem map, and strike plan.