Domain authority from zero: the new-SaaS playbook

A brand-new SaaS does not start at the bottom of the rankings. It starts off the map entirely.

When you ship a product nobody has heard of, the problem is not that search engines rank you low. It is that nothing on the web confirms you exist. No other site links to you. No directory lists you. No human has written your name in a sentence that a crawler later read. From the outside, your domain is a single unverified claim: a website that says it is a company. The rest of the web has not agreed yet.

This post is about what domain authority actually is once you strip the jargon off it, why a new product is invisible by default rather than by penalty, and the concrete steps to build domain authority from zero without a budget or a launch spike to lean on.

What domain authority really is

Domain Authority, with the capital letters, is a third-party score invented by an SEO tool vendor. Google does not use it. It is a useful rough proxy and a terrible north star. The thing it approximates is real, though, and it is worth being precise about what that thing is.

Search engines and AI answer engines both run on the same underlying bet: a page is more trustworthy when other trustworthy sources point at it, mention it, and agree about what it is. That is the entire mechanism. Authority is not a number you earn by writing more. It is the accumulated weight of the rest of the web vouching for you, directly through links and indirectly through mentions, references, and consistent descriptions across independent sources.

A new domain has none of that weight because it has no history. Nothing points at it. Nothing describes it. There is no corroboration to lean on, so the engines have no reason to surface you over the thousand established pages that already answer the query. This is not a punishment. It is the absence of evidence, and the fix is to manufacture evidence the honest way.

Two layers: links and entities

There are two distinct layers stacked here, and founders usually only think about the first one.

The first is the classic link layer. Other sites link to your pages, those links pass signal, and pages accumulate standing. This still matters for traditional search and it is the part everyone already knows about.

The second is the entity layer, and it is the one that decides whether you show up in AI answers. An entity is a thing the web has agreed exists: a company, a product, a person, with a stable identity that systems can attach facts to. Google maintains a Knowledge Graph of these. Large language models absorb the same kind of structure from their training data. When ChatGPT or Gemini or Perplexity recommends a tool, it is reaching for entities it recognises and has facts about, not crawling and ranking pages in real time. If your product is not a recognised entity, it is not in the candidate set at all.

Why a new SaaS starts invisible

Picture the chain of evidence an engine needs before it will confidently say your product exists and does what you claim. Walk through it for a week-old SaaS and you can see exactly where it breaks.

Does another independent site name the product? Usually no. Does any structured database hold a record for the company, with a founding date, a category, and a link back to the site? No. Do the descriptions of the product match across the places it does appear, so a system can reconcile them into one consistent entity? There is only one place it appears, so there is nothing to reconcile. Has anyone in a community written the product name next to the problem it solves, in a sentence a crawler indexed? Not yet.

Every one of those is a missing corroboration. The engines are not refusing you. They simply have a single unverified source, your own marketing site, claiming things about a product, and one self-interested source is the weakest possible evidence. The work of building authority from zero is the work of turning that single claim into a web of agreeing, independent references.

The self-reference trap

The instinct most technical founders have is to write more on their own blog. More pages, more keywords, more posts. This is not wrong, but on its own it does almost nothing for authority, because it is all the same source talking about itself. You can publish a hundred excellent pages and still be a single unverified claim, just a longer one. Content compounds only once other surfaces start pointing back at it. The early work is not writing more. It is getting named elsewhere.

Step one: make yourself a recognisable entity

Before chasing links, give the machines something to attach facts to. This is the cheapest, highest-leverage work for a new product, and almost nobody does it deliberately. The goal is consistency: the same name, the same one-line description, the same category, the same founding details, repeated across a handful of structured sources so that systems can reconcile them into one confident entity.

Get into the structured databases

A few sources carry disproportionate weight because other systems read them as ground truth. Create real, accurate records in them.

• Crunchbase: a company profile with founding date, category, founders, and the canonical website URL. It is widely scraped and frequently referenced as a source of company facts.
• Wikidata: the open, structured knowledge base that feeds the wider knowledge-graph ecosystem. A clean Wikidata item for your company, linked to its official site and described in one consistent sentence, is one of the most direct ways to declare an entity. Only create one if the facts are real and verifiable. It is a factual database, not a marketing surface.
• LinkedIn company page: a populated page with a consistent description and the website link, treated by many systems as an identity anchor.
• Your own site's structured data: Organization and Product schema in JSON-LD on your homepage, stating the name, the description, the logo, and your profiles on the sources above. This is you telling the crawler, in machine-readable form, which entity this site belongs to.

The point is not any single listing. It is that five independent sources now say the same true thing about the same product, with matching names and matching links. That cross-source agreement is what lets a system collapse the scattered mentions into one entity it can recommend. The matching matters as much as the listing. If your product is described five different ways under three different names, you have given the engines five weak signals instead of one strong one. Pick one name and one sentence and use them everywhere, exactly.

Claim the directories and review sites

Category directories and review sites are doing double duty. They are themselves surfaces where users decide what software to use, and they are high-trust pages that name your product in context. A G2 or Product Hunt or Capterra page is an independent source describing what you do, in the same category language your users search with. Claim the profiles, fill them out accurately, and keep the description consistent with everywhere else. You are simultaneously appearing where people choose tools and adding another corroborating reference to your entity.

Step two: earn mentions, then links

Once you exist as an entity, the next job is volume and variety of independent references. Note the order: mentions first, links second. A plain-text mention of your product name, with no hyperlink, still teaches the web that the name exists and what it sits next to. AI answer engines in particular pick up unlinked brand mentions, because they are reading sentences, not just following anchors. Links are a stronger version of the same signal, not a different one.

Be useful where your users already are

The durable way to earn mentions with no budget is to be genuinely useful in the places your users already gather, and to do it under your own name. Answer a hard question in a relevant subreddit. Write a comment on Hacker News that actually helps, then mention what you built only when it is the honest answer to the thread. Publish a teardown or a build-log on Indie Hackers. Each time, you leave behind a sentence on an independent, indexed, high-trust surface that names your product next to the problem it solves. That is the exact shape of evidence the engines are looking for.

This is slow and it does not feel like growth work, which is precisely why it compounds and why most founders skip it. The reader who lands on your comment is one outcome. The crawler that indexes your product name in a credible context is the other, quieter one, and the second outcome keeps paying out long after the thread goes cold.

Write the thing other people cite

There is a specific kind of content that earns references on its own: the page that is genuinely the best explanation of a narrow, real question your users have. Not a keyword-stuffed listicle. The actual reference. A clear methodology, a worked example, an original framing, a small piece of data you gathered yourself. People link to and quote the page they wish they had written. For a new domain this is the only kind of content that pulls its weight early, because it is the only kind that gives an independent source a reason to point back at you.

Step three: consistency, repeated, over time

Authority is not a project you finish. It is a slope you hold. The mechanism rewards steady, consistent presence over time, because trust is partly a function of persistence: an entity that keeps appearing, keeps being described the same way, and keeps being referenced is one the engines grow more confident about. A burst of activity around a launch and then silence reads, to a system, like a product that may no longer exist.

So the cadence matters more than any single action. A consistent name and description everywhere. A handful of structured records kept accurate. A steady trickle of useful public contributions under your own name. The occasional genuinely-good reference page. Done for months, not weeks. None of it is clever. All of it compounds, because each new corroboration raises the confidence of every system that has already seen the others.

What to watch while it builds

You cannot watch authority directly, so watch its shadows. Are you starting to appear, by name, in AI answers for the questions your users actually ask? Are new referring domains showing up in Search Console over time? Does a search for your exact product name return a clean, consistent picture across the web, or a muddle of mismatched descriptions? Those are the leading indicators that the entity is forming and the references are accumulating. They shift slowly, and that is the honest shape of the work.

The whole playbook in one breath

A new SaaS is invisible because it is a single unverified claim. The fix, in order: become a recognisable entity by getting consistent, accurate records into the structured databases and directories that systems read as ground truth. Then earn independent mentions, plain-text or linked, by being useful in public under your own name and by writing the pages other people choose to cite. Then hold the line, with the same name and the same description, for long enough that the web grows confident you are real. There is no shortcut in that sequence, but there is also nothing in it that requires a budget.

If you want a read on where you stand today, the free Growth Snapshot scores your discoverability across seven dimensions, and entity presence and authority is one of them. It will tell you which structured sources already name you, where your description drifts, and which surfaces are quietly leaving you out. AfterLaunch then monitors those same surfaces and drafts the work in your voice for you to approve, so the slow, consistent part of this playbook actually gets done instead of sitting on a someday list.