SEO for SaaS: what moves the needle for a small product

SEO for a small SaaS is not the same problem as SEO for a content site or a store. You have a handful of pages, no link profile, and a product that solves one narrow thing. Most SEO advice is written for sites with 500 pages and a team. This is the version for a product with twelve pages and one person.

The short answer: fix the technical basics once, then spend every remaining hour on pages that answer a question your buyer actually types. Everything else is a rounding error until you have links.

The order that matters

Do these in order. Skipping ahead is how people spend six months writing posts that never get indexed.

1. Make every page indexable and canonical. If Google cannot decide which URL is the real one, nothing else you do counts.
2. Give each page one job and one keyword. Two pages targeting the same phrase cannibalise each other and both lose.
3. Publish pages that answer questions, not pages that describe your product. Your features page ranks for your brand name and nothing else.
4. Earn links. Below roughly 20 referring domains you are competing with one hand tied behind your back, and no amount of on-page work substitutes.

Steps 1 and 2 take an afternoon. Step 3 is months. Step 4 never ends.

Step 1: the technical floor

There are exactly five things that matter here, and they take about an hour to verify across a small site.

A canonical on every page. Without one, /pricing, /pricing/, /pricing?utm_source=x and https://www.yoursite.com/pricing are four pages competing with each other, splitting whatever link signals you have four ways. Set an absolute self-referencing canonical on every page.

A title under 60 characters that leads with the keyword. The title is still the single strongest on-page signal. "Invoice software for freelancers | Acme" beats "Acme - the best way to get paid faster" every time, because the second one is about you and the first one is about what someone searched.

A meta description between 120 and 160 characters. Google rewrites it roughly 70% of the time, but when it uses yours, a clear one wins clicks. Write it as a promise, not a summary.

One H1 per page. Not zero, not three. Multiple H1s are not a penalty, but they are a reliable symptom that the page does not know what it is about.

A sitemap and a robots.txt that agree with each other. A URL that is in your sitemap but blocked in robots.txt sends a contradictory signal and wastes crawl budget.

Step 2: one page, one job

The most common mistake in small-SaaS SEO is publishing five pages that all target variations of the same phrase. "Time tracking software", "best time tracking software", "time tracking app", "time tracking tool" - Google sees one intent, picks whichever page it likes least, and buries the rest.

Map it explicitly. One row per page, one target phrase per row, and no phrase appearing twice. If two pages want the same phrase, merge them. A single 1,800-word page beats three 600-word pages on the same topic, reliably.

Step 3: pages that answer questions

Your homepage and features page will rank for your brand name. That is all they will ever do, and that is fine - brand searches convert at 30%+ and cost nothing.

Non-brand traffic comes from pages that answer a question. The pattern that works for small SaaS, in rough order of return:

Free tools. A tool that solves one small job in the browser earns links, gets bookmarked, and ranks for a keyword with real volume. It is the only content type that reliably outranks bigger sites, because most incumbents have not bothered to build a good one.

Comparison pages. "X vs Y" where at least one of X and Y is not you. Third-party comparisons rank better than self-serving ones because Google reads intent, and someone searching a competitor comparison does not want a sales page.

"How to" for the job your product does. Not "how to use Acme" - how to do the underlying task, with your product as one of several ways to do it. The traffic is qualified because the searcher already has the problem.

Original data. If you can publish a number nobody else has - a benchmark, a survey, an aggregate from your own usage - that is the one content type that reliably earns links without asking.

What does not work, despite being the most common advice: generic listicles ("10 tips for productivity"), keyword-stuffed glossary pages, and AI-generated posts about topics adjacent to your product. All three are now actively filtered.

Step 4: links, and being honest about them

Everything above assumes you eventually get links. At DR 0 with no referring domains, a perfect page ranks around position 30 for anything competitive. The same page at 40 referring domains ranks around 8. That gap is roughly 20x the traffic.

The link sources that actually work for a small SaaS, in order of effort-to-return:

• A tool other people cite. Passive, compounds, and the only one that scales without you.
• Original data. Publish a number, pitch it to three people who write about your space.
• Directories and launch boards that are real. Most are worthless - a page with 400 outbound links and no editorial standard passes nothing. A handful with genuine editorial review and dofollow links do count.
• Integration and partner pages. If you integrate with a tool, most have a directory. That link is easy and permanent.

What to measure, and when

Do not check rankings weekly. New pages on a new domain do essentially nothing for four months, then move. Checking early only makes you rewrite pages that were fine.

Look at these instead:

• Indexation. In Search Console, what percentage of submitted URLs are indexed? Below 90% means thin content, and adding more pages will make it worse.
• Impressions, not clicks. Impressions rise months before clicks do. Rising impressions at position 25 means the page is working and needs links, not a rewrite.
• Queries you did not target. Long-tail queries you never wrote for are the strongest signal that a page has topical authority.

A realistic timeline

For a new domain with a dozen pages, publishing two good pages a month:

• Months 1–3: near zero. Pages get indexed, nothing ranks. This is normal and is not a signal to change strategy.
• Months 4–8: long-tail queries land at positions 8–20. First meaningful traffic, usually a few hundred sessions a month.
• Months 9–14: if referring domains grew, the best pages move to positions 4–10 and traffic roughly triples. If they did not, everything stays at position 15 forever.

That last sentence is the whole game. On-page work has a ceiling you hit in month six. Links are what raise it.