Category positioning for B2B SaaS: when not to do it

Most B2B SaaS companies should position inside an existing category or carve a subcategory of one. Creating a new category is correct in a narrow case: you have a budget for demand creation, a buyer who cannot describe the problem in existing words, and years of patience. Everyone else is using category creation to avoid a comparison they would lose. The tell is search volume. If nobody types the category name, nobody is looking for it, and you have bought yourself a market of one. I know this from my own site, which runs 60 industry landing pages that produced one click between them in a quarter.

The three moves, and what each actually costs

There are three positions available and only three. The choice is a budget decision dressed as a brand decision.

The row that decides it is the last-but-one. An existing-category failure is loud and cheap to detect. A subcategory failure shows up in five months. A new-category failure looks exactly like a new category that is still early, which is why companies keep funding them past the point of sense.

The one number everybody quotes, and why I am not quoting it

Category design writing leans hard on a Play Bigger statistic that category kings capture around 76% of their category's market cap. I went to check it. The report URL, playbigger.com/time-to-market-cap-report, now returns a 404 page — "Oops! Page not found". I could not read the methodology or the sample, so I am not using the number, and neither should you until someone produces the underlying document. It is repeated across dozens of agency blog posts with no primary source attached. That is the evidence base under the most confident advice in B2B marketing.

What has to be true before you name a category

Demand you can point at

A category name is a search query or it is a slogan. Before committing, pull the head term into Keyword Planner or Search Console and look at it honestly. Zero volume is not "untapped". Zero volume means every person who ever arrives at that page has to be sent there by you, at your cost, forever. If you are trying to forecast how that number moves across a year, the mechanics are in keyword seasonality and forecasting demand.

A budget for the 95%

Category creation is demand creation, and demand creation is a brand-advertising problem. John Dawes of the Ehrenberg-Bass Institute, in work for the LinkedIn B2B Institute, put a number on the size of that problem: with a typical five-year switching cycle, only 20% of business buyers are in the market in a given year and just 5% in a given quarter. The other 95% are not shopping.

Read that against category creation. You are teaching the market a new word, and teaching it to an audience of whom 95% will not act on the lesson for years. That is a sustained media budget. If your marketing plan is a founder posting on LinkedIn plus a blog, you do not own the instrument this strategy requires.

A name your buyer will repeat

The common failure is a name that only makes sense after the explanation. If a prospect cannot say it to their CFO without a slide, what you have is internal vocabulary that escaped.

The site that took 60 industry positions

Here is the cautionary example, with real numbers, because I own the failure.

This site defines 60 industries in lib/industries/*.ts, each rendered as its own landing page at /industries/[slug], and a seeded blog layer repeated the move by title: healthcare content marketing, legal content marketing, hospitality, fintech, beauty, pet care. Category positioning by breadth: be plausible everywhere.

Search Console, 85 days to 2026-09-03: 22 clicks from 4,553 impressions, a 0.48% CTR across roughly 390 indexed URLs. Russia sent 1,320 impressions and zero clicks. The United States sent 970 impressions at an average position of 38.8 and zero clicks. Nobody clicks position 38.

Now split the same domain by what kind of post it was:

• —Build-log and worked-arithmetic posts — narrow, specific, unglamorous — ranked between positions 2.8 and 15.5.
• —Generic "[industry] content marketing" posts clustered between 46 and 81. The best-placed of them — legal at 22.9, franchise at 26.9 — still earned zero clicks.

Same author, same domain, same quarter, same publishing process. The only variable was how specific the claim was. Even the brand query "santosh paudel" averaged position 8.2 across 51 impressions — my own name, ranked eighth, which is a separate humiliation. And 111 of 182 ranking pages earned five or fewer impressions in the quarter.

What narrowing did

A later audit classified 279 posts across the seed files. Roughly 150 of them averaged about 380 words with no table, no code, no internal link and no image. Pruning cut the published count from 267 to 103, a 61% cut. The full autopsy, including what I found broken along the way, is in 389 pages, 22 clicks: a site teardown, and the scoring method I used to decide what died is in the content pruning scoring model.

What survived was two clusters, both narrow: quantitative marketing arithmetic, and build logs from shipping this stack. Those are the pages sitting at 2.8 to 15.5. Sixty industry landing pages produced one click between them. Two subcategory positions produce ranking.

If you suspect your own site is spread across positions it cannot defend, the diagnosis is page by page — that is what a content audit is for, and it is the cheapest step before any repositioning decision.

The arithmetic, in nine lines

Breadth loses on arithmetic: a click-through curve multiplied by a volume number, and the curve is brutal past position 10.

type Move = { name: string; monthlyVolume: number; attainablePosition: number; monthsToRank: number };

// Assumed CTR curve. Two anchors from my own Search Console (85 days to 2026-09-03):
// Nepal, the only market that converts, ran 10.8% CTR at position 9.2; the United States
// returned 0% over 970 impressions at position 38.8. Both are country rows, not single queries.
// The curve below is deliberately harsher than that Nepal point and hard-zeroes past 20.
const ctr = (pos: number): number => (pos > 20 ? 0 : 0.28 * Math.pow(0.72, pos - 1));

const clicks = (m: Move): number => m.monthlyVolume * ctr(m.attainablePosition);

const moves: Move[] = [
  { name: "Existing category", monthlyVolume: 8100, attainablePosition: 25, monthsToRank: 18 },
  { name: "Subcategory",       monthlyVolume:  320, attainablePosition:  4, monthsToRank:  5 },
  { name: "New category",      monthlyVolume:   10, attainablePosition:  1, monthsToRank: 24 },
];

for (const m of moves) {
  console.log(`${m.name.padEnd(18)} vol ${String(m.monthlyVolume).padStart(5)}/mo  pos ${String(m.attainablePosition).padStart(2)}  -> ${clicks(m).toFixed(1)} clicks/mo after ${m.monthsToRank} months`);
}

console.assert(clicks(moves[1]) > clicks(moves[0]), "subcategory should beat existing category");
console.assert(clicks(moves[1]) > clicks(moves[2]), "subcategory should beat new category");
console.assert(clicks(moves[0]) === 0, "position 25 is zero clicks");

Run it with npx tsx moves.ts:

Existing category  vol  8100/mo  pos 25  -> 0.0 clicks/mo after 18 months
Subcategory        vol   320/mo  pos  4  -> 33.4 clicks/mo after 5 months
New category       vol    10/mo  pos  1  -> 2.8 clicks/mo after 24 months

Every input there is an assumption I have labelled as one — the volumes and attainable positions are illustrative, not benchmarks. Swap in your own Search Console figures and the shape holds. A market 25 times smaller that you can rank in beats a large one you cannot reach, and beats owning position one on a term with no searchers by an order of magnitude. That is the case against category creation for anyone without a media budget, expressed as multiplication.

The trap in the third row is that it looks like success. Position 1, brand-defining, no competition. Also 2.8 clicks a month.

Drift, and the half-life of a category name

Drift is the canonical category-creation story in B2B SaaS: launched 2015, named "conversational marketing", built a movement around replacing lead-capture forms with chat.

On 13 February 2024, Salesloft acquired it. Read how the acquisition was described at the time. Salesloft Acquires Drift in CRM Magazine, dated 14 February 2024, calls Drift "a buyer experience and conversational artificial intelligence company". The category Drift spent years teaching the market had already been relabelled by the market, in the announcement of its own acquisition.

I am not claiming the strategy failed — a company was built and sold. The narrower point is worth internalising: a category you create is an asset you rent. It gets renamed by analysts, absorbed into platforms, or swallowed by whatever the current AI framing is. If your positioning depends on a word you invented, you are exposed to everybody else's vocabulary decisions.

What I would do instead

Position in a subcategory of a category that already has demand, and win it on specificity rather than on naming rights.

Pick the segment before the noun

"CRM" is contested. "CRM" plus a segment whose workflow the generic product handles badly is not. The segment does the differentiating work that a new name was supposed to do, and it arrives with search demand already attached.

Make the claim testable

Vague positioning and invented categories fail for the same underlying reason: a reader cannot check either one. A claim that survives is one a prospect can verify in the first paragraph. The mechanics of that are in the specificity principle, and the system view of running content this way is in what is a content system.

Set a five-month clock when you commit

Write the review date down on the day you choose the position. If a subcategory position has produced no inbound using your segment phrasing in five months, the position is wrong or the segment is too small to matter. Which signals are worth watching at small scale is covered in the five SEO metrics that matter.

FAQ

Should a B2B SaaS startup create a new category?

Almost certainly not. Creating a category means funding demand creation for a term nobody searches, aimed at an audience of whom roughly 95% are not in the market this quarter. It fits a funded company with a multi-year media budget and a buyer who genuinely lacks language for the problem. For everyone else, a subcategory of an existing category buys a defensible position in months rather than years.

How do I know if my category positioning is working?

Watch three signals: impressions on the category term rising from a base you can see, inbound enquiries that use your exact positioning phrase unprompted, and average position improving on the terms that describe your segment. On my own site the difference was stark — positions 2.8 to 15.5 for the specific claims against a 46-to-81 band for the broad ones, in the same quarter, from the same publishing process.

What is the difference between category creation and category positioning?

Category creation means inventing a new market name and teaching buyers to use it. Category positioning is the wider decision about which of three slots you occupy: an existing category, a subcategory within one, or a new one. Creation is the most expensive of the three and the one most often chosen for the wrong reason, which is avoiding a comparison you would lose.

Does category positioning still matter with AI search?

More, if anything. Language models answer by summarising what multiple sources say about a named thing. A category with one publisher gives them nothing to synthesise, so an invented term with a single site behind it is invisible to an assistant in a way it was not to a search index. A specific subcategory that several independent sources discuss is legible to both.

Is it too late to position in a crowded category?

No, but go two levels down. Crowded categories stay crowded at the head term and are usually thin at segment level, which is where a small team can rank inside a quarter.

Spread thin across positions you cannot defend? A page-level audit shows which claims are ranking, which are dead weight, and what a narrower position would actually cost you. Get a content audit or get in touch.