Indie founder SEO: what building in public can and can't do for a new domain

What should a solo founder with a months-old domain and no links do about search? Not chase head terms. A new site will not rank for "best CRM" or "what is AEO" in its first months, however good the writing, and AI answers mostly cite pages that already rank. What you can control is three things: an identity chain that makes you verifiable, public posts specific enough that other people reference them, and a handful of deep pages on the narrow problem your product solves. Building in public feeds all three. It just works slower than most build-in-public threads suggest.

What a new domain is up against

Start with the honest constraints, because they decide where the effort goes.

Retrieval comes before citation. ChatGPT, Perplexity and Google's AI features cite from pages their retrieval step already surfaced. AirOps looked at 548,534 pages ChatGPT retrieved across 15,000 prompts and found 15% of them cited; pages ranking first on Google were cited at 43.2% (AirOps). For Google's own AI Overviews, a page must be indexed and eligible for a snippet before it can be a source (Google). A domain with no authority enters that funnel late.

Content tweaks have a low ceiling. The C-SEO Bench study found most content-rewriting methods largely ineffective, and found their gains shrinking as more competing pages adopt them (Puerto et al., 2025). There is no formatting trick that lifts a new site past established ones.

Re-evaluation is slow. Google says some changes take effect in days, while a site-wide reassessment "could take several months" (Google on core updates). A new site is effectively being assessed for the first time.

Engines change underneath you. When Gemini 3 became the default model for AI Overviews on 27 January 2026, SE Ranking found 42.4% of previously cited domains replaced, while the top 500 cited domains were almost untouched (SE Ranking). Churn like that falls on small sites. The upside is that a solo founder can react the same week.

None of this means a new site can't win. It means the wins come from long-tail queries, from places other than Google, and from being referenced by people, before they come from rankings.

Mentions, and the caveat that comes with them

Ahrefs compared 75,000 brands' signals with their visibility in AI Overviews and found web mentions correlating at 0.664, branded anchors at 0.527, brand search volume at 0.392 and backlinks at 0.218 (Ahrefs). Two cautions apply. It's a correlation, and Ahrefs says so. And it measures brands. Applying it to a founder's personal name is a reasonable extrapolation, not a finding.

The second caution comes from Google. Its May 2026 AI optimization guide says "seeking inauthentic 'mentions' across the web isn't as helpful as it might seem." Since 15 May 2026 its spam policies also explicitly cover attempts to manipulate AI answers. So the mention advantage only holds for mentions you earn by being worth referencing: a number someone quotes, a decision someone links to in a newsletter. Seeding your own brand into forums and listicles is the version Google is now telling you not to bother with.

The identity chain

This is one-time work, about an afternoon, and it is the part a founder has for free that a corporate content team has to build.

Google's own question for authorship is simple: "Do bylines lead to further information about the author?" (Google). The chain is how you answer yes. If you have a sizable following on a major social or video platform, Google's Search profiles for creators (US first, from June 2026) are worth checking as well.

The build-in-public loop

Posting publicly is how other people come to reference you, and the replies tell you which topics deserve a long page.

What makes the loop turn is specificity. "Shipped a feature today" gives nobody anything to reference. A post like "moved our background jobs to a queue and cut p95 response time from 9 seconds to 2, here's the trace" gets quoted, because it's a fact someone else can use. (That one is a made-up illustration. Use your own numbers, and only real ones.)

The same property makes a page citable to an engine. Writing to be referenced and writing to be cited turn out to be one habit.

Where the hours go

Retrievability, first and once. Confirm OAI-SearchBot, PerplexityBot, Claude-SearchBot and Googlebot can reach you, at the CDN as well as in robots.txt. Fifteen minutes, and nothing else matters if it fails.
The identity chain above. One afternoon.
Your product page. It's usually the page with the most relevant queries. Give it SoftwareApplication and Organization markup that points at your founder identity.
Five to ten deep pages, not fifty. Each on a problem you understand because you built the product. Length is whatever the answer needs; Google says it has no preferred word count.
Dogfood write-ups. Use your own product in public and publish what you learn, with screenshots and real numbers. It's the only original data a small site is guaranteed to have.
A small free tool or two. Tools get linked and mentioned in a way articles rarely are, and they can't be answered away by an AI summary, because the user has to arrive to use them.

Free tool · No signup

Free Schema Validator

Paste any URL → find the structured data your page is missing, with ready-to-paste JSON-LD fixes.

Check your schema →

What not to do

Publish under the company name. It gives up the one advantage that costs nothing.
Scale with AI generation. Thirty mediocre articles in three months will do worse than eight good ones on a site with no authority to carry them, and mass-produced pages are what Google's scaled content abuse policy describes.
Write the fiftieth generic article on your category. Write what only you know because you built the thing.
Target head terms. "Best CRM" belongs to budgets you don't have. "Best CRM for solo developers on Postgres" is a fight you can win.
Buy links or manufacture mentions. Link spam has been a spam policy for years, and since May 2026 attempts to manipulate AI answers are explicitly covered too.
Chase engine-specific tricks. Per C-SEO Bench, their gains shrink as others adopt them.

Measuring when there's almost nothing to measure

For the first months, traffic is the wrong scoreboard. Watch these instead:

Indexing. In Search Console's Page indexing report, how many of your pages are indexed, and how many sit in "Crawled – currently not indexed"? That second bucket is the clearest sign Google doesn't yet see enough value to index them.
Long-tail impressions. Impressions on specific queries usually show up well before clicks do.
AI Overview impressions. Search Console's generative AI performance report shows when your pages appear in AI Overviews and AI Mode (impressions only, data from 18 May 2026).
Citations elsewhere. Pick ten queries you care about and check ChatGPT and Perplexity for them now, then again in a month.
A mentions log. Each week: what you posted, who referenced it, and where.

On timing, what can honestly be said is that structural changes don't register until pages are re-crawled and re-evaluated. That takes weeks, not days, and Google says broader reassessment "could take several months." How much traffic you'll have in month six isn't knowable for a site nobody has seen, and any article that gives you a month-by-month table is making it up.

FAQ

Should I publish under my real name or a pseudonym?

Your real name, if you can. It's the simplest way to make bylines lead somewhere verifiable. If you have a genuine safety reason for a pseudonym, a consistent pseudonym with its own bio page and public profiles that match works much the same way. What doesn't work is no author at all, or a company account.

How many hours a week does this take?

After the one-time setup, a pace many solo founders can sustain is one substantive article a week or every other week, plus regular posting and a short weekly check of Search Console and a few AI answers. That's an opinion, not a benchmark. The real limit is quality: if a faster pace means weaker pages, go slower.

Can I outsource the writing?

Partly. It works when you supply the research, the numbers and the opinions, and a writer turns them into clean prose that you edit and publish under your name. It fails when you hand over a topic and publish what comes back, because the specific perspective that makes a founder's page worth citing is exactly what gets lost.

This week

Check your robots.txt and CDN for the four retrieval crawlers. Then write the bio page and point your byline at it. Then post one specific, true number from your own product, and write it down in your mentions log. The rest of the strategy builds on those three.

If you're a team rather than a solo founder, see AEO for SaaS startups; for the pre-publication list, the AEO content checklist.

Sources cited in this piece

Updated October 1, 2026: reframed around honest expectations for a new domain; added Google's May 2026 caution on inauthentic mentions; replaced timing ranges that had been labeled as documented with Google's own wording.