How do startups do SEO with no budget or authority?

Does low domain authority really stop a new startup from ranking?

Low domain authority isn't what's keeping your startup off page one — keyword choice is. Authority scores are third-party estimates built by SEO tool vendors, not inputs Google uses. A six-week-old site can absolutely rank for a question nobody has answered properly; it cannot rank for a term defended by a dozen established competitors with a decade of links behind them. The realistic sequence looks much the same for every new domain: indexing within one to two weeks, your own brand name inside a month, long-tail and comparison terms somewhere between three and six months, and genuinely competitive head terms at six to twelve months or longer.

Google has never confirmed a sandbox, and its own search advocates say there's no switch holding new sites back. What exists is slower and less dramatic: a trust-building period where the algorithm has almost no behavioral data about you, so it assigns an early ranking estimate and revises it as clicks, links and crawl history accumulate. Practitioners put that phase at roughly three to nine months. It stretches in money-and-health categories — finance, legal, medical — where quality thresholds are stricter, and it shortens when you earn a few relevant links early, publish consistently, and look like a real company with named humans behind it.

So pick a nine-to-twelve-month horizon before you judge the channel — but don't fly blind until then. In month one you should see pages indexed and impressions appearing in Search Console. By month three, average position on your target queries should be climbing even while clicks stay thin. By month five or six, the first long-tail pages should be producing signups or enquiries you can trace. If impressions are flat at week twelve across everything you've published, the problem isn't patience. Either you chose keywords you can't win, or the site hasn't given Google a single reason to trust it yet.

Which keywords should you target first with zero authority?

Target bottom-of-funnel queries first: comparison terms, "alternatives to X", integration and use-case pages, pricing questions, and the awkwardly specific problems your support inbox answers all day. These searches are low in volume and low in competition precisely because every content team in your category is chasing the 5,000-search educational term instead. They also convert several times better, because the person typing them has already decided to buy something and is only choosing between options. A page with 150 monthly searches and a 10% signup rate is worth far more to you than a guide with 5,000 searches converting at 1%.

Difficulty scores are a starting filter, not a verdict. New domains generally do well under a difficulty rating of about 20, and the lowest band — roughly zero to ten — is where your first quarter belongs. Then open the results page and actually look at it. Thin pages, posts last updated in 2021, forum threads and unknown blogs on page one are your real opening. A difficulty score of 45 with a weak results page is a better bet than 30 with four strong brands occupying it. If Reddit or Quora ranks for a commercial term in your category, treat that as an open invitation.

Here's the unpopular part: for your first three months, don't start a blog. Build comparison pages, alternatives pages, integration pages and one honest page per core use case, then fill the top of the funnel after those are indexed and moving. Mine the keywords from places that cost nothing — recorded sales calls, support tickets, the search terms report if you've ever run paid ads, and the exact wording customers use during onboarding. And write the comparisons honestly. A page that admits where a competitor is genuinely better is the one buyers trust, and it's the one AI assistants quote back.

How do you earn links and mentions without spending money?

You earn links without a budget from three places: your own data, your own face, and other people's audiences. Original numbers from inside your product — anonymized benchmarks, pricing teardowns, survey results from your users — give writers something they can't get anywhere else, which is the only dependable reason anyone links to a company nobody has heard of yet. Founder bylines in niche trade publications, podcast appearances, and real answers in the communities your buyers already read cover the rest. None of this is actually free, mind you. It costs your calendar instead of your card.

In practice, the cheapest wins come from relationships you already have. Partner and integration pages trade links with the tools you connect to. Customers and suppliers will often add you to a case study or a "who we work with" page if someone simply asks. Industry directories, accelerator portfolios, review platforms and local business associations are unglamorous but entirely legitimate. Conference and meetup speaker pages link out as a matter of course. A handful of topically relevant links from sites your buyers genuinely visit will move rankings more than fifty generic ones, and it won't put the domain at risk.

Skip anything sold to you as a link package. Bought links, private blog networks and mass guest-post farms are the quickest route to a manual action, and cleaning one up costs far more than the links ever did. Put that energy into entity signals instead: the same company name, description, location and founding details on your site, your LinkedIn page, funding databases and review profiles, so search engines and AI models can connect scattered mentions to one business. Unlinked brand mentions matter here too — being named in a trade article or a podcast transcript feeds the models even with no hyperlink attached.

What gets a startup cited by ChatGPT, Claude and Perplexity?

AI assistants cite pages that answer a question in one quotable passage, back claims with specifics, and are technically reachable by their crawlers. Research into generative engine optimization found the largest visibility gains came from adding relevant statistics, credible quotations and clear source attribution — and that the effect was strongest for lower-authority sites, which is exactly where a startup sits. Keyword stuffing performed worse than changing nothing at all. The practical translation is blunt: replace "growing fast" with a number, attribute the number, and make every section readable and extractable on its own without the paragraphs around it.

Before any of that matters, check that the machines can actually see you. ChatGPT's search layer leans on Bing, so a page missing from Bing's index gives it nothing to quote — verify that first in Bing Webmaster Tools. Then confirm your robots.txt and your firewall allow GPTBot, OAI-SearchBot, ClaudeBot, Claude-SearchBot and PerplexityBot; bot-management rules drop them silently far more often than anyone expects. Render content server-side, add Article and FAQPage markup, show a visible author and last-updated date, and consider publishing an llms.txt map of your most citable pages. It helps to check how quotable your pages look before you publish.

Structure beats volume here. Citations cluster heavily in the opening third of a page, so lead each section with a forty-to-sixty-word answer and expand underneath, phrase headings as the questions people actually ask, and keep one idea per paragraph so a model can lift it cleanly. Freshness carries real weight too, particularly with Perplexity, which favors a genuinely updated page over an identical stale one. And don't build a separate "AI version" of your site. The page quoted in an AI answer is almost always the same page ranking well in Google — one asset, two channels, no duplicate work.

What should your first 90 days of startup SEO look like?

Your first ninety days should produce a small set of high-intent pages, a clean technical base, and the start of a measurement habit — not a content library. Weeks one and two: set up Search Console, Bing Webmaster Tools and GA4, submit a sitemap, clear indexing errors, and check Core Web Vitals on the pages that carry revenue. Weeks three to six: publish four to six bottom-of-funnel pages, each answering one question properly. Weeks seven to twelve: pitch three founder bylines, record two podcast appearances, and internally link every new page to your signup or contact page.

Keep the toolkit free until the data justifies paying for more. Search Console already tells you which queries you appear for, and positions eleven to twenty are the cheapest wins you will ever get — a page stuck on page two usually needs a sharper title, a tighter answer in the first paragraph and two internal links, not a rewrite. Autocomplete, People Also Ask and related searches are a perfectly serviceable keyword research substitute for a quarter. Add a paid research tool once you have enough pages for the data to change an actual decision, which is rarely before month four.

Measure signups, enquiries and revenue per page rather than sessions. Then compare the channels honestly: paid search delivers traffic this afternoon and stops the afternoon you pause it, while an organic page keeps earning for years and keeps getting quoted in AI answers long after you published it. If you want the numbers behind that trade-off, our breakdown of whether SEO or Google Ads costs less per lead runs the twelve-month comparison, and the traffic projection calculator will model what your own pages could return. If you'd rather not run all of this solo, book a short strategy call.

Frequently asked questions