Skip to content
busyless
Book a call

Startup SEO strategy

The category you are inventing has no search volume yet

Written by Eugene SuslovLast reviewed 28 August 2026No affiliate links
Sector
Software and tech
Model
B2B SaaS, National service
Competition
High
Time to results
6 to 18 months, against a runway that may be shorter
Typical monthly
$0 to $8,000

Key takeaways

  1. 1Nobody is searching for your category, because you are still naming it. They are searching for the spreadsheet, the manual process or the incumbent you intend to replace, and those queries have volume today.
  2. 2Search can be the wrong investment for a startup, and saying so is part of the job. If the runway is shorter than the payback, the honest answer is a narrow slice of this work rather than a programme.
  3. 3The founder is the only person who can write the content that matters, and has the least time. Every plan that ignores this fails in month three when the drafts stop arriving.
  4. 4A pivot leaves a domain full of content arguing for a positioning the company has abandoned. Deal with it deliberately, because the old pages usually outrank the new ones.
  5. 5Generating ten thousand pages from a database is cheap, fast and the exact shape search engines treat as scaled content abuse. Forty pages that could only have been written by you is the better trade.

SEO for startups begins with a fact everybody skips past. Search is a demand capture channel, and a startup building something genuinely new has no demand to capture. The queries do not exist because the category does not exist, and no keyword tool will tell you that because it can only report what people already type.

What does exist is the problem. Somebody is solving it today with a spreadsheet, a manual process, an intern, four tools stitched together or an incumbent product they dislike. Those things have names, and those names have search volume.

That gives a startup SEO strategy an odd shape. You are not ranking for what you sell. You are ranking for the thing your buyer is doing instead, and using the page to explain that there is another way to do it.

The second thing that makes this different is the clock. Organic search compounds over quarters and a seed-stage company is often funded for a number of months you can count on your fingers. Recommending a programme that pays back after the money runs out is not advice, it is a sales pitch.

There is a third difference and nobody enjoys it. The only person who can write the content that ranks is usually the founder, because the credibility comes from having done the thing. That person also has to raise money, hire and ship product.

The exclusion matters as much as the definition. A startup in an established category is really running the playbook of that category, which is what the rest of this directory is for. This page is for the case where the thing you sell does not have a name yet.

Who already ranks in startups

The first job in startup SEO is a subtraction rather than an addition, because the results pages you would most like to win are the ones that describe a market you are trying to make obsolete. Look at what actually ranks for the problem, not for the product.

What is on the results page

  • Incumbent brand pages and their comparison content, which occupy every query naming the way things are currently done.
  • Listicles from established publishers on "best [old category] tools", frequently written by companies in the category.
  • Reddit, Hacker News and specialist forum threads on the problem, which are often the only honest discussion of it anywhere.
  • Tutorials and templates for doing the job manually, which is what somebody searches immediately before they go looking for a tool.
  • Nothing at all on the queries describing your actual product, which reads as opportunity and is usually just absence.
  • AI-generated answers that have never heard of you, and which draw their vendor lists from review platforms you are not on.

  • The incumbent you are replacing

    varies

    Not a directory, and the most important entry here. They own every query naming the current way of working, including the ones about its limitations. Their comparison and alternatives pages are the first thing your buyer reads about you, and they were written before you existed.

  • Product Hunt

    producthunt.comClaim it

    Ranks well on product and category queries and gives a new company a page with real authority behind it on launch day. Worth doing properly once. Worth knowing that the launch spike is traffic rather than demand, and that it does not repeat.

  • Hacker News

    news.ycombinator.com

    Threads rank durably and the comments are frequently the most useful public analysis of a new product. Not a channel to optimise. A place where a founder posting honestly, including about what does not work yet, is treated very differently from a marketing account.

  • Crunchbase

    crunchbase.comClaim it

    Ranks for the company name almost immediately, often above the company's own site in the first year. The profile is claimable, most founders never look at theirs, and it is one of the sources an answer engine uses to decide what your company does.

  • G2 and Capterra

    g2.comClaim it

    The gatekeeper you are not yet eligible for. Category placement is driven by review volume, so a company with eleven customers cannot compete there. Worth setting up early and building slowly rather than launching at it.

  • Reddit and specialist communities

    reddit.com

    Where the problem is discussed by the people who have it, in language no marketing team would use. The single richest source of query and objection material available to an early company, and it costs nothing but attention.

  • LinkedIn

    linkedin.com

    Not a search surface in the ordinary sense, and for many early companies it is where the audience actually is. It matters here because founder-led writing there is frequently the source of the demand that later shows up as branded search.

  • The AI assistants

    varies

    They will be asked to recommend tools in your space and they have no information about you at all. What they draw on is review platforms, listicles and forum threads. That is an argument for existing on those surfaces early rather than for optimising anything.

Two overlapping circles. One holds the product vocabulary: the category name, the feature names, the positioning. The other holds what people actually type: the spreadsheet, the manual process, the incumbent's name. The overlap is small and contains only the company's own name.
The small overlap is not a failure of keyword research. It is what being early looks like, and the right-hand circle is where the work goes.

What people actually search

Every cluster below describes something other than your product, which is the point. They are ordered from the ones that have volume today to the one that will not have volume for a year, and the last is the only one that names your category.

Doing it manually

Informational

> [job] spreadsheet template

The page that wins it: The template itself, published free, plus the honest limits of it

The highest-volume cluster available to most early companies and the one they refuse to build, because it helps somebody not buy. It reaches exactly the person with the problem, at the moment they are feeling it, and it costs a weekend.

The incumbent's name

Commercial investigation

> [incumbent] alternatives

The page that wins it: An alternatives page that is genuinely useful, including about competitors

Real demand, real intent, and the searcher has already decided they want something different. It is also the only cluster where you can compete on day one, because the query is about them rather than about a category you have not established.

The limits of the current approach

Informational

> why does [current process] break at scale

The page that wins it: A problem page that describes the failure precisely

The bridge cluster. Somebody searching this has outgrown the manual method and does not yet know a category exists. If your page names the pattern well, you get to be the one who names the solution too.

The adjacent tool people are misusing

Informational

> how to use [popular tool] for [your job]

The page that wins it: An honest guide to doing it in the other tool, and where it stops working

Counterintuitive and effective. People are already forcing a general-purpose tool to do this job, they search for help constantly, and nobody else is willing to write the guide because it recommends a competitor first.

Your own company name

Navigational

> [your company] pricing

The page that wins it: Real pages for pricing, security and what the product does

Small volume and disproportionate value, because it is the query somebody runs after hearing about you somewhere else. In the first year Crunchbase, LinkedIn and a launch listing frequently outrank you for it.

The role rather than the tool

Informational

> how to run [process] as a small team

The page that wins it: Founder-written material about the operating problem

Where founder-led content actually lands. The reader is looking for a way of working rather than a purchase, and the credibility comes from having done it, which is why it cannot be outsourced to a content agency.

Integration and stack queries

Commercial investigation

> [tool a] [tool b] integration

The page that wins it: One real page per integration that genuinely works

The most tempting cluster to generate programmatically and the place most startups get it wrong. Two hundred pages for integrations that do not exist yet is the scaled-content shape. Twelve for ones that do is a legitimate asset.

The category you are naming

Informational

> what is [your category]

The page that wins it: The definitional page, published early and updated as the language settles

No volume today and possibly none for a year. Publish it anyway, early, because if the category does take hold the definitional page written first tends to keep the position. Just do not fund the programme on it.

A linear sequence: somebody builds a spreadsheet, hits its limits, searches for a better way to do the manual job, discovers the approach has a name, and only then searches the category. Four steps happen before the category is ever typed.
Every step before the last one has search volume today. The last one is the only step most startups write for.

What the rules change

No regulator governs startups as such. What does apply is the ordinary law on advertising claims, and early companies breach it more often than established ones because the claims are bolder and nobody has been told to check them.

1

An AI claim is an advertising claim and it is being enforced

Federal Trade Commission, Operation AI Comply, announced 25 September 2024, and Section 5 of the FTC Act

What it means

The Commission brought five actions in that sweep against companies whose claims about artificial intelligence outran what their products did, and the enforcement focus has continued since. Describing a rules engine as AI, or claiming autonomy the product does not have, is a deceptive claim regardless of how common it is in the category.

So do this

Write down what the product actually does and check every page against it. If a claim needs a qualifier to be true, put the qualifier in the sentence rather than in a footnote. Anybody raising money should also read the securities regulators' position on the same claims.

2

First, only and leading all require substantiation

Federal Trade Commission Act Section 5, and the advertising substantiation doctrine

What it means

A superiority or uniqueness claim has to be supported by evidence the advertiser holds at the time it is made. "The only platform that" and "the fastest" are objective claims. So is a comparison against a named competitor, and so is a performance number in a headline.

So do this

Keep a substantiation file: one line per claim on the site, with the evidence and the date it was checked. It takes an hour a quarter and it is the difference between a defensible claim and a hopeful one.

3

Your investors and advisors leaving reviews is the named case

Federal Trade Commission Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465, effective 21 October 2024

What it means

The rule prohibits undisclosed insider reviews and testimonials, and reviews given in exchange for compensation conditioned on a particular sentiment. Investors, advisors, employees and their relatives are insiders. A launch-day push where the network posts enthusiastically without disclosure is precisely what the rule describes.

So do this

Ask supporters to disclose the connection when they post, and say so in the message you send them. Never offer anything in exchange for a review. Keep a record of who you asked, because the failure here is almost always that nobody thought about it.

4

The waitlist is a database with rules attached

California Consumer Privacy Act as amended, the GDPR where any user is in Europe, and CAN-SPAM for the follow-up

What it means

A pre-launch signup form collects personal information and creates obligations that do not wait for a product to exist. A privacy notice that describes a company you no longer are, or an email flow with no working opt-out, is a live problem rather than a future one.

So do this

Publish a privacy notice that matches what you actually collect, keep it current through the pivot, and make sure the opt-out works in every system you send from. It takes a morning and it is much harder to retrofit.

Proving expertise

A new company has no history, no links and no reviews, which are the usual proxies for trust. What it does have is people who genuinely know the problem, and almost every early company hides them behind a brand voice.

  • Named founders with real profiles, real histories and a stated reason they are working on this problem.
  • Content published under a person's name rather than the company's, with that person's own experience visible in it.
  • A changelog with dates on it, which is the cheapest possible proof that the product is being built rather than announced.
  • Public documentation, outside any login, which is the strongest evidence a stranger has that the product does what the marketing says.
  • Named early customers, with permission, rather than a wall of logos from a pilot that ended.
  • An honest limitations page saying what the product does not do yet, which converts better than most people expect.
  • Public participation by the founders in the communities where the problem is discussed, predating the launch.

How to build a startup SEO strategy

Before anything below, answer the runway question honestly, because it decides how much of this list applies. A startup SEO strategy for a company with eight months of funding is three pages and a decision to revisit in a year, not a programme.

  1. 1

    Week 1

    Decide whether to do this at all

    • Write down the runway in months and the realistic payback period for organic search, side by side
    • Check whether the problem you solve has search volume even though the category does not
    • Decide who is writing. If the answer is the founder, get the hours committed in a calendar rather than in principle
    • If the runway is short and the founder has no hours, pick the narrow version: the alternatives page, the free template, the pricing page. Stop there

    You end up with
    A written decision about scope, with the reason recorded, so nobody relitigates it in month four

  2. 2

    Weeks 1 to 6

    Build the pages that work with no authority

    • Publish the free template or tool for doing the job manually, with no gate
    • Write the incumbent alternatives page, honestly, naming who each option suits
    • Write real pages for the company name: what it does, pricing, security, who it is for
    • Claim the Crunchbase, Product Hunt and review platform profiles even though there are no reviews yet
    • Publish the definitional page for the category you are naming, and date it

    You end up with
    A small set of pages that can rank without link authority, because they answer queries nobody else wants to answer

  3. 3

    Weeks 4 to 16

    Get the founder on the record

    • Publish under named authors, starting with whoever has genuinely done the job before
    • Write the problem pages describing where the current approach breaks, in the words practitioners use
    • Write the guide to doing it in the adjacent tool people are already misusing, including where it stops working
    • Open the documentation and treat it as content rather than as support material
    • Participate in the communities where the problem is discussed, under real names, without linking anything for a month

    You end up with
    A domain with a voice attached to it, and coverage of the problem rather than of the product

  4. 4

    Weeks 12 to 40

    Build only what compounds, and survive the pivot

    • Add integration pages one at a time, only for integrations that genuinely work
    • Get the first named customers on the record with checkable detail
    • Review every page against the current positioning after any material change of direction, and redirect rather than delete
    • Report impressions on the problem clusters, not sessions, and state plainly that branded search is being created elsewhere

    You end up with
    A compounding asset that still describes the company you have become rather than the one you were

Four stacked layers, each resting on the one below. Founder hours at the base, then pages that need no authority, then problem content, then the category page at the top with nothing resting on it.
Each layer holds the one above it. The top layer is the one everybody wants to start with, and it is the only one that carries nothing.

Technical fixes with the best payoff

Startup websites break in ways established sites do not, because they are built fast, by whoever was free, on whatever tool got it live that week. None of the items below is expensive to fix and several of them are invisible until somebody looks.

  • The staging site is indexed

    An hour

    A preview deployment on a public host, with no crawl protection, duplicating the whole site. It happens on almost every project that uses preview URLs, and it can end up outranking production for the company's own name.

    Block indexing on every non-production environment at the platform level rather than per project, and search for your own copy in quotes to find the ones already out there.

  • The coming-soon page is still the strongest page on the domain

    An afternoon

    The launch page collected the links from the announcement and was never redirected. A year later it has more authority than anything on the live site and it says the product is coming soon.

    Redirect it to the page that inherited its job, keeping the URL if it holds links. Check what points at it before deciding where it goes.

  • The documentation lives on a different domain

    A day if done early

    Documentation platforms default to their own subdomain or a hosted address. The docs are usually the most specific and most linked content the company owns, and none of that reputation reaches the marketing site.

    Serve the docs from a subdirectory on the main domain. Do it early, while there are twelve pages, rather than later when there are six hundred.

  • The site is a single JavaScript page with no URLs

    Two days now, a quarter later

    A no-code builder or a single-page app where the whole marketing site is one address and the sections are anchors. There is nothing to rank because there is only one page, and no way to add anything without a rebuild.

    Give each real topic its own URL with its own title and description. This is a two-day fix in the first year and a three-month migration in the third.

  • The pivot left the old positioning ranking

    A week after any material pivot

    Eighteen months of content argued for a product the company no longer sells. Those pages have the links and the history, so they outrank the new ones, and a visitor arrives at a confident description of something that is not for sale.

    Audit every page against the current positioning. Redirect what maps, rewrite what nearly maps, and remove what does not, keeping the URLs that hold links pointed somewhere sensible.

  • Programmatic pages were generated before there was anything to say

    A day of deleting

    A database of integrations, locations or use cases turned into thousands of near-identical pages. It is cheap, it looks like progress, and it is the textbook shape of scaled content abuse.

    Keep only the pages with a genuinely different answer on them, and remove the rest. Where the set is large, publish in small batches as each one becomes true.

  • Nobody set up Search Console for the first year

    An hour

    It is a ten-minute task with no immediate payoff, so it is skipped, and the data does not backfill. A year later there is no record of what the domain was doing before the pivot or the rebrand.

    Set it up today for every property and subdomain, including the docs host, and export monthly from the start. Add Bing as well, since its index feeds surfaces beyond Bing itself.

  • The rebrand dropped the old domain

    A day, plus a renewal fee

    A name change moves the site to a new domain and the old one lapses or is left to redirect to the home page. Every link, mention and citation earned in the first two years quietly stops passing anything.

    Keep the old domain registered indefinitely and map the redirects page by page rather than to the root. This is cheap insurance against losing the only authority you had.

Structured data that applies here

The types below fit this industry specifically. Most of them earn no rich result on their own, which is worth knowing before anyone sells the work on that basis. What they do is describe the entity precisely, which matters for how search engines and answer engines resolve who you are.

  • Organization

    The home page, referenced by everything else

    The most valuable node for a new company, because there is nothing else telling an engine what you are. The identifiers matter more here than the description: sameAs pointing at Crunchbase, LinkedIn and the repository is how a search engine joins your company to the mentions it finds elsewhere.

    Organization.jsonld
    {
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://[YOUR-DOMAIN]/#organization",
  "name": "[COMPANY NAME]",
  "legalName": "[REGISTERED NAME]",
  "url": "https://[YOUR-DOMAIN]",
  "logo": "https://[YOUR-DOMAIN]/[LOGO.PNG]",
  "description": "[ONE SENTENCE ON WHAT THE PRODUCT DOES, NOT ON THE VISION]",
  "foundingDate": "[YYYY-MM]",
  "founder": [
    {
      "@type": "Person",
      "name": "[FOUNDER NAME]",
      "url": "https://[YOUR-DOMAIN]/about#[SLUG]",
      "sameAs": ["[LINKEDIN URL]"]
    }
  ],
  "sameAs": [
    "[CRUNCHBASE URL]",
    "[LINKEDIN COMPANY URL]",
    "[GITHUB ORG URL]",
    "[PRODUCT HUNT URL]"
  ],
  "contactPoint": {
    "@type": "ContactPoint",
    "contactType": "sales",
    "email": "[EMAIL]"
  }
}
  • SoftwareApplication

    The product page, where the product is software

    Use it for your own product only. Keep offers honest: a free tier is an offer at zero, a trial is not. Leave aggregateRating out entirely until real, verifiable ratings exist somewhere you can point at, which for a new company is usually not yet.

    SoftwareApplication.jsonld
    {
  "@context": "https://schema.org",
  "@type": "SoftwareApplication",
  "name": "[PRODUCT NAME]",
  "url": "https://[YOUR-DOMAIN]/product",
  "applicationCategory": "BusinessApplication",
  "operatingSystem": "Web",
  "description": "[WHAT IT DOES, IN THE WORDS A BUYER WOULD USE]",
  "publisher": { "@id": "https://[YOUR-DOMAIN]/#organization" },
  "softwareVersion": "[VERSION]",
  "releaseNotes": "https://[YOUR-DOMAIN]/changelog",
  "offers": {
    "@type": "Offer",
    "price": "[PRICE, OR 0 FOR A GENUINELY FREE TIER]",
    "priceCurrency": "USD",
    "priceSpecification": {
      "@type": "UnitPriceSpecification",
      "price": "[PRICE]",
      "priceCurrency": "USD",
      "unitText": "[PER USER PER MONTH, OR WHATEVER IS TRUE]"
    }
  }
}
  • Person

    Founder and author profiles

    The node that does most of the work for an early company, because the people are the only established entities you have. Link it from every article the person writes and point sameAs at profiles that already exist and already rank.

    Person.jsonld
    {
  "@context": "https://schema.org",
  "@type": "Person",
  "@id": "https://[YOUR-DOMAIN]/about#[PERSON-SLUG]",
  "name": "[FULL NAME]",
  "jobTitle": "[ROLE]",
  "url": "https://[YOUR-DOMAIN]/about#[PERSON-SLUG]",
  "image": "https://[YOUR-DOMAIN]/[PHOTO.JPG]",
  "worksFor": { "@id": "https://[YOUR-DOMAIN]/#organization" },
  "knowsAbout": ["[THE PROBLEM DOMAIN]", "[THE PROBLEM DOMAIN]"],
  "alumniOf": {
    "@type": "Organization",
    "name": "[PREVIOUS COMPANY OR INSTITUTION]"
  },
  "sameAs": [
    "[LINKEDIN PROFILE URL]",
    "[GITHUB PROFILE URL]",
    "[PERSONAL SITE URL]"
  ]
}
  • Article with a named author

    Founder-written problem and craft content

    The author node is the reason to bother. A startup publishing under the company name throws away its only differentiator, which is that a person who has actually done the job is writing. Point the author at the Person node above rather than restating a name.

    Article with a named author.jsonld
    {
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "[THE PAGE TITLE]",
  "url": "https://[YOUR-DOMAIN]/[PATH]",
  "datePublished": "[YYYY-MM-DD]",
  "dateModified": "[YYYY-MM-DD]",
  "author": { "@id": "https://[YOUR-DOMAIN]/about#[PERSON-SLUG]" },
  "publisher": { "@id": "https://[YOUR-DOMAIN]/#organization" },
  "about": "[THE PROBLEM, NOT THE PRODUCT]",
  "wordCount": "[NUMBER]"
}
  • WebApplication for a free tool

    The free template, calculator or utility you publish ungated

    Where the tool is genuinely free to anybody, say so with an offer at zero and isAccessibleForFree. Do not use this on something behind an email wall: that is a different claim and marking it up as free is inaccurate rather than clever.

    WebApplication for a free tool.jsonld
    {
  "@context": "https://schema.org",
  "@type": "WebApplication",
  "name": "[TOOL NAME]",
  "url": "https://[YOUR-DOMAIN]/tools/[TOOL-SLUG]",
  "description": "[WHAT IT CALCULATES OR PRODUCES]",
  "applicationCategory": "BusinessApplication",
  "operatingSystem": "Web",
  "browserRequirements": "Requires JavaScript",
  "isAccessibleForFree": true,
  "publisher": { "@id": "https://[YOUR-DOMAIN]/#organization" },
  "offers": {
    "@type": "Offer",
    "price": "0",
    "priceCurrency": "USD"
  }
}
  • BreadcrumbList

    Anything below the top level, once there is more than one level

    Skip it entirely while the site is six pages, which is the honest answer for most companies in their first months. Add it when a real hierarchy appears, usually when the docs, the blog and the integration pages arrive at once.

    BreadcrumbList.jsonld
    {
  "@context": "https://schema.org",
  "@type": "BreadcrumbList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "name": "Home",
      "item": "https://[YOUR-DOMAIN]/"
    },
    {
      "@type": "ListItem",
      "position": 2,
      "name": "[SECTION]",
      "item": "https://[YOUR-DOMAIN]/[SECTION-SLUG]"
    },
    {
      "@type": "ListItem",
      "position": 3,
      "name": "[PAGE NAME]",
      "item": "https://[YOUR-DOMAIN]/[SECTION-SLUG]/[PAGE-SLUG]"
    }
  ]
}

What it costs

The bands below are editorial estimates rather than quotes, and the first one is not a fallback. For a company with a short runway, doing this yourself in a few hours a week is frequently the correct answer rather than the poor one.

Founder-led, no budget

$0 to $200
  • The alternatives page, the free template and real pages for the company name
  • Founder-written problem content, which is the only version that carries any weight
  • Profile claims on Crunchbase, Product Hunt and the review platforms
  • Search Console and Bing set up, so a year from now there is a record

Who it suits

Pre-seed and seed companies where the founder is still doing sales

Where it stops

It cannot survive the founder getting busy, which is certain. The realistic failure mode is four good pages and then eleven months of nothing.

One writer, part time

$1,500 to $4,000
  • A cadence the founder can sustain by being interviewed rather than by writing
  • The problem and adjacent-tool clusters built out properly
  • Documentation opened and maintained as content
  • Integration pages added one at a time as each integration becomes real

Who it suits

A company with product-market fit signals and twelve months or more of runway

Where it stops

It does not create demand for a category. If nobody is searching for the problem either, this is a distribution problem and search is downstream of it.

Funded programme

$4,000 to $8,000
  • Coverage of the full problem space rather than the three cheapest clusters
  • Named customer stories, which cost time and approvals rather than money
  • A proper technical foundation before the site gets big enough to make it expensive
  • Reporting that separates demand you created from demand you captured

Who it suits

A Series A company with a category that is starting to have a name

Where it stops

Spending more does not make a category exist sooner. That happens through customers, press and other people using the words, and search reflects it rather than causing it.

How to do it with no budget

Almost all of the highest-value work here is free, which is unusual and is a genuine advantage of being early. What it costs is founder attention, which is the scarcest resource in the company.

  1. 1

    Write down the runway and the payback period next to each other

    An hour

    A document

    The most valuable hour on this list. It decides whether the rest of the page applies to you at all.

  2. 2

    Publish the spreadsheet or template people use to do this manually

    A weekend

    Whatever you already built for yourself

    You almost certainly made one while researching the problem. Publishing it ungated reaches the exact person you want.

  3. 3

    Write the alternatives page for the incumbent

    A day

    Your own knowledge of the market

    Name who each option genuinely suits, including when the incumbent is right. That is what makes it get linked rather than dismissed.

  4. 4

    Claim Crunchbase, Product Hunt and the review platform profiles

    3 hours total

    The platforms

    These outrank you for your own name in the first year and they are what an answer engine reads about you.

  5. 5

    Set up Search Console and Bing Webmaster Tools

    An hour

    Both are free

    The data does not backfill. Doing this today is the only way to have a record of the year before your pivot.

  6. 6

    Read the problem's subreddit or forum for a fortnight without posting

    20 minutes a day

    Reddit, Hacker News, a Slack community

    It produces a better query list than any tool, in the words practitioners actually use rather than the words in your deck.

  7. 7

    Write the substantiation file for every claim on the site

    2 hours

    A spreadsheet

    One line per claim, the evidence, the date checked. Cheap now, and much harder to assemble after somebody asks.

The tool stack

Buy almost nothing. The tools below are chosen for being cheap or free and for answering a question a founder actually has, rather than for producing a dashboard nobody will open twice.

  • Check whether the problem has volume even though the category does not

    Keywords EverywhereSEO APIRead the review

    Credit-based rather than a subscription, which suits somebody who needs an answer twice a quarter rather than a seat every month. The question to ask it is about the manual process, not about your product.

    Free routeSearch Console once you have any impressions, plus the autocomplete suggestions

  • See what the incumbent already ranks for

    SpyFuSEO APIRead the review

    The useful export is their pages that mention alternatives or comparisons, because those are where your buyer is already reading about the category. Cheap, and you need it once rather than continuously.

    Free routeReading their site map and their comparison pages, which takes an afternoon

  • Publish a site that can grow without a rebuild

    Sveltia CMSHeadless CMSRead the review

    The requirement is a real URL per topic and an editing route that does not need a developer. Git-based tooling suits a team where the founder writes and an engineer deploys, which is most companies at this stage.

    Free routeIt is open source and runs on your existing repository

  • Find out whether the assistants have heard of you

    OtterlySEO APIRead the review

    Early on the answer is no, and the useful part is seeing which sources the assistant does cite for your category. That list is where to exist next.

    Free routeAsking the assistants directly, on a schedule, and saving the answers

  • Get indexed on the surfaces beyond Google

    Bing Webmaster ToolsSEO APIRead the review

    Worth ten minutes because several answer engines and browsers draw on that index, and a new domain is frequently missing from it entirely. Submit the sitemap and check the coverage report once a month.

    Free routeIt is free

  • Make the runway argument numerically rather than by instinct

    SEO Forecast ToolFree toolOpen the tool

    Use it to show what the channel could return by a given month, then compare that month to the date the money runs out. If the second is earlier, that is the answer and it is better to have it on paper.

    Free routeIt is free

Take it from here

Everything below is yours to take. Fill the [BRACKETS] and it is ready to use. Start with the runway decision, because it is the one that tells you whether to use any of the others.

Brief

The one-page document that decides whether search is a rational investment for your company right now, and what to do if it is not.

RUNWAY AND PAYBACK DECISION
Company: [COMPANY NAME]
Prepared by: [NAME]     Date: [YYYY-MM-DD]
Decision owner: [FOUNDER NAME]

WHY THIS COMES FIRST
Organic search compounds over quarters. A lot of early companies are
funded for a number of months smaller than that. This document makes the
comparison explicit before anybody signs a contract, because the answer
"not yet" is a legitimate outcome and is much cheaper than finding out
in month nine.

THE TWO NUMBERS

  Cash in bank                                    [AMOUNT]
  Net monthly burn                                [AMOUNT]
  Runway, in months                               [N]
  Next raise expected                             [MONTH YYYY, OR "none
                                                   planned"]

  Realistic months before organic produces
  meaningful signups                              [N]
  Basis for that estimate                         [WHAT YOU BASED IT ON.
                                                   If it is a vendor's
                                                   promise, write that.]

  Does the payback land inside the runway?        [YES / NO / ONLY IF WE
                                                   RAISE]

DOES THE DEMAND EVEN EXIST?

  Our category name                               "[NAME]"
  Monthly searches for it                         [N]
  The manual process people use today             "[WHAT]"
  Monthly searches for that                       [N]
  The incumbent                                   "[WHO]"
  Monthly searches for their name plus
  "alternatives"                                  [N]

  If the second and third numbers are also near zero, this is a
  distribution problem rather than a search problem. Write that down and
  stop here: [CONCLUSION]

WHO IS ACTUALLY WRITING?

  Person                                          [NAME]
  Hours per week they will commit                 [N]
  Are those hours in a calendar?                  [YES / NO]
  What gets dropped to make room                  [WHAT]

  If the answer to the calendar question is no, assume zero hours. That
  is not cynicism, it is the observed failure mode.

THE DECISION

  [ ] FULL PROGRAMME
      Runway comfortably exceeds payback, demand exists on the problem,
      writing time is committed.

  [ ] NARROW VERSION
      Do only these, then stop and revisit:
        [ ] The incumbent alternatives page
        [ ] The free template or tool
        [ ] Real pages for the company name, pricing and security
        [ ] Profiles claimed, Search Console set up
      Revisit on: [YYYY-MM-DD or AFTER THE RAISE]

  [ ] NOT NOW
      Reason: [WRITE IT DOWN SO NOBODY RELITIGATES IT IN MONTH FOUR]
      Revisit trigger: [WHAT HAS TO CHANGE]

  Decided by: [NAME]      Date: [YYYY-MM-DD]

What to publish

Everything below is written for somebody who has the problem and has never heard of you. That rules out most of what early companies publish, which is written for people who already follow the company.

  • The free template or tool

    One, published early, then improved

    It reaches the person doing the job manually, at the moment they are frustrated by it, and it costs a weekend because you already built it for yourself.

  • The incumbent alternatives page

    One per serious incumbent, revised twice a year

    Real search demand from day one, from people who have already decided they want something else. The only high-intent query a new company can realistically compete for.

  • Founder-written problem content

    Whatever the founder will actually sustain, which is usually less than planned

    The credibility comes from having done the job, which cannot be outsourced. It is also the material that gets quoted, which is how a new domain earns its first real links.

  • The guide to doing it in the adjacent tool

    One per tool people genuinely misuse

    People are already forcing a general-purpose tool to do this. Nobody writes the guide because it recommends a competitor first, which is exactly why it works.

  • Open documentation

    Continuous, as part of shipping

    The most specific and most trusted content the company owns, and the strongest evidence to a stranger that the product is real. It also ranks for questions marketing pages cannot answer.

  • The limitations page

    Once, updated with the roadmap

    Saying what the product does not do yet converts better than most founders expect, and it is the page that makes every other claim on the site credible.

And what not to

  • Vision and manifesto posts. They reach investors and peers, which is a real audience and is not the one with the problem.
  • Funding announcements as content. The coverage is a day, the page is forever, and it tells a buyer nothing about whether the product works.
  • Thousands of generated pages for integrations, locations or use cases that do not exist yet. Cheap, fast, and the exact shape of scaled content abuse.
  • Ranking attempts on the category name before anybody searches it. Publish the definitional page and then leave it alone.
  • Ghostwritten founder content in a voice the founder does not have. The audience for early-stage writing is small and it notices.
A dense grid of identical generated page tiles with a small number of distinct tiles scattered among them. The distinct tiles are the only ones carrying anything a reader could not have guessed from the template.
The grid took an afternoon to produce. Removing it takes a day, and the scattered tiles are what should have been published on their own.

The expensive mistakes

Building the plan around the category name

Costs you A year of content aimed at a query with no searchers

Target the manual process, the incumbent and the adjacent tool. Publish the category page once and leave it.

Starting a content programme with eight months of runway

Costs you Money spent on a channel that pays back after the company runs out of it

Do the narrow version: alternatives page, free template, real company pages. Revisit when the runway extends.

Planning around founder writing time that does not exist

Costs you Four good pages and then eleven months of silence

Interview the founder and have somebody else draft. Book the hour weekly or accept that it will not happen.

Generating pages from a database because it is cheap

Costs you Thousands of near-identical pages, which is the shape search engines act against

Publish each page as its subject becomes genuinely true, and keep only the ones with a different answer on them.

Leaving the pivot's content in place

Costs you Confident pages arguing for a product you no longer sell, outranking the ones you do

Audit against the current positioning after any material change, and redirect rather than delete.

Letting the docs live on somebody else's domain

Costs you The most linked content you own building somebody else's reputation

Serve the documentation from a subdirectory. It is a day's work at twelve pages and a quarter at six hundred.

Asking investors to post supportive reviews without disclosure

Costs you The exact undisclosed-insider case the FTC's reviews rule describes

Ask them to disclose the connection, offer nothing in exchange, and keep a record of who you asked.

What to measure

The reporting problem is that the numbers are tiny and stay tiny for a long time, which makes month-on-month comparison meaningless. What is worth watching early is direction and composition rather than volume.

Leading indicators

Move first. They predict, they do not prove.

  • Impressions on problem queries, excluding brand

    Search Console, filtered to queries naming no company

    The earliest signal that any of this is working, and it arrives months before clicks do. Track it as a trend across a quarter rather than a change between two weeks.

  • Queries you rank for that you did not target

    Search Console, sorted by impressions and read manually

    The most useful early report and almost nobody looks at it. It tells you which words your buyers actually use, which is usually not the words in your deck.

  • Referring domains earned, not bought

    Any backlink tool, or Search Console links

    A new domain has none, and the first ten matter more than the next hundred. Worth reading individually rather than counting, because who links tells you which content landed.

  • Branded search volume over time

    Search Console, brand queries only

    The honest measure of whether anybody has heard of you. It moves because of launches, press and founder writing rather than because of search work, and the report should say so.

  • Free tool or template usage

    Analytics on the tool page

    A better early signal than traffic, because using it costs the visitor something. Watch repeat use in particular, which is the first evidence of a real problem being felt.

Business indicators

The ones a manager acts on.

  • Signups and activations from organic entry

    The product database joined to first-touch source

    Report the untracked share alongside it. Early companies rarely have the instrumentation to attribute properly and pretending otherwise sets a precedent you cannot maintain.

  • Revenue from accounts with an organic first touch

    Billing joined to whatever attribution you have

    Small numbers, so read them as case studies rather than as a trend. One customer who found you through a template page is information; a percentage of eleven is not.

  • Founder hours actually spent writing

    A calendar

    Unusual to report and it predicts the rest better than anything else on this list. When it hits zero, the leading indicators follow about a quarter later.

The verdict

The uncomfortable part of this page is that search may be the wrong channel for you right now, and a good adviser will say so. A channel that compounds over eighteen months is a poor fit for a company funded for ten.

Where it does fit, the work is unusual. You rank for the problem rather than the product, you publish the thing that helps somebody not buy yet, and you accept that the category page earns nothing for a year.

Anybody selling startup SEO services should be asked about runway in the first conversation. If the answer treats a ten-month company and a Series B the same way, the plan was written before they met you.

The other tell in SEO services for startups is what a provider proposes to publish first. A programmatic set of location or integration pages is fast to build, easy to invoice and exactly the wrong first move.

FAQ

Startups SEO questions

  • Should a pre-seed startup do SEO at all?
    Often not as a programme. Do the narrow version: an alternatives page, a free template and real pages for your own name. That is a week of work with a long tail, and it does not commit money you need for product.
  • What do we target when nobody searches for our category?
    The thing your buyer does today. The spreadsheet, the manual process, the incumbent's name, and the general-purpose tool people are already forcing into the job. Those have volume now.
  • Is it worth publishing content about the category we are naming?
    Publish the definitional page early and then leave it alone. If the category takes hold, the page written first tends to keep the position. Just do not build the plan or the forecast on it.
  • Can we outsource founder content?
    You can outsource the drafting, not the source. Interview the founder and have somebody else write it up under the founder's name with their approval. Fully ghostwritten thought leadership reads as generic because it is.
  • What happens to our content after a pivot?
    Audit it against the new positioning within a few weeks. Redirect what maps, rewrite what nearly maps, remove what does not. The old pages usually outrank the new ones, which is why leaving them is worse than it looks.
  • Are programmatic pages ever a good idea early?
    Only where each page has a genuinely different answer on it and the subject is already true. Twelve integration pages for integrations that exist is an asset. Two hundred for ones on the roadmap is not.
  • Are the budget figures on this page quotes?
    No. They are editorial estimates, and the zero band is a real recommendation rather than a placeholder. For a short-runway company, doing a small amount of this yourself is frequently the right call.
  • What is the one thing to do first?
    Write your runway in months next to the realistic payback period for organic search. If the payback lands after the money runs out, that decides the scope of everything else on this page.

Run it yourself, or have someone own it

Everything above is written to be run without us, and the free path is genuinely most of the value for a single-location business. Where these plans stall is almost never the plan. It is that nobody owns it after the first month. That is the job we do, with search as one distribution layer inside a wider system rather than the whole engagement.

See how that worksAll industries