The short version, if you only have a minute
- They never chased traffic. Every decision was scored on signups and paying customers.
- Over 1,000 template pages turned every use case of the product into a search entry point.
- The product built its own backlinks. Every embedded form carried a link home.
What $1M ARR and no SEO programme looks like
June 2015. Typeform has just crossed $1M in recurring revenue with about 5,000 customers. A growth marketer joins as one of the first thirty employees. There is no SEO programme, no content team, and a budget of 5,000 euros a month.
Three and a half years later the company is at $13M ARR and search is its most reliable acquisition channel. The budget was never the interesting part. The interesting part is what they refused to measure.
Traffic was never the target. The team tracked signups and paying customers, and they tracked them separately, because a keyword that produces signups is not always a keyword that produces revenue. That one rule quietly killed off most of the keyword list a normal agency would have handed them.
boltDo this on your site
Name the one number your SEO is accountable for before you open a keyword tool. If it is signups or revenue instead of sessions, half your planned content dies on the spot, and that is the point.
The first SEO win took one afternoon and zero words
It was a footer.
The early audit turned up a set of product landing pages Google could not reach. They had been built for paid campaigns, they converted well, and nothing on the site linked to them. As far as a crawler was concerned, they did not exist.
Adding them to the sitewide footer took an afternoon. They became indexable immediately, and they were already good pages.
That tiny fix sets the pattern for everything after it. Internal links are how search engines find pages and how authority travels between them. A page nobody links to is invisible, no matter how good the writing is. So Typeform built the hierarchy to say out loud which pages mattered most.
How the URL structure encoded commercial priority
Distance from the root is a signal. Typeform used it deliberately.
typeform.com/quizzesMoney pages, one click from the homepage/surveys/nps-guideCluster content feeding authority to the parent/templates/{use-case}1,000+ template pages, sandboxed in one directoryboltDo this on your site
Before you commission another article, crawl the site and pull every page that converts but has fewer than three internal links pointing at it. That list is usually the fastest ranking movement available to you, and it costs nothing to fix.
Nobody searches for "conversational data collection platform"
They search for "event RSVP form". Or "customer satisfaction survey sample". Or "registration form online". Typeform's product had thousands of use cases, and every single one was already a query.
So the strategy wrote itself: one page per use case, with a live working template sitting right there on the page, free to try. The distance between a Google result and actually using the product was close to zero.
The registration form template ranked first for “registration form online”, a term doing over 8,000 searches a month. Nobody landing on that page needed convincing that they wanted a registration form. They were already building one.
Then came the test that doubled the business case for the whole library. The “use this template” button originally dumped people into a blank builder. The variant dropped them straight into the template they had just been playing with.
The activation test
One change, to where a single button led.
CONTROL
Click the button, land in an empty generic builder, rebuild from scratch the thing you just saw working. Most people left.
VARIANT
Click the button, land inside that exact template, already populated, ready to edit and send.
boltDo this on your site
Ranking a page and activating a visitor are two different jobs. Watch what happens in the ten seconds after the click on your best page. Fixing that handover is usually worth more than another 2,000 words above the fold.
Unlimited ideas, six slots a month
The template library could have grown forever. Engineering capacity was five or six new templates a month. That gap is where most SEO programmes quietly fail, because they fill the slots with whatever has the biggest search volume.
Typeform scored every candidate on four things instead, and built toward the best combined score rather than the biggest number.
A single keyword that will not rank is usually an authority problem that a cluster would have solved.
The four-factor scoring model
Search volume was the entry ticket, never the decision.
| FACTOR | QUESTION IT ANSWERS | WHERE THE DATA COMES FROM |
|---|---|---|
| Traffic potential | How many people search this each month? | Keyword tools, the base layer only |
| Ranking difficulty | How many links before we can compete? | Referring domains of the current top results |
| Conversion estimate | Will these people actually sign up? | Paid search data and comparable live pages |
| Retention signal | Do customers from this use case stay? | Customer success and churn data by use case |
boltDo this on your site
Group your research into topics before you sort it into a list. Decide which topics you can credibly own this year, then pick terms only inside those. Everything else is a page you will publish and quietly forget.
2.3 million backlinks and nobody sent an outreach email
Typeform reached DR 93 and a backlink profile among the largest on the web. There was no link building team behind it. There was a product decision.
Forms were built to be embedded on other people’s websites. Every embedded form carried a small “powered by Typeform” line, and that line was a link. Hundreds of thousands of customers, each one publishing a link on their own domain, every time they shipped a form.
The upgrade is the part worth stealing. Once the homepage had far more authority than it could use, the team rewrote the embed logic so the attribution pointed to the relevant page instead. Authority stopped pooling at the front door and started landing on the pages that had to win competitive terms.
boltDo this on your site
Ask one question in your next product meeting: where does our product already show up on a customer's own website? Embeds, shared reports, public profiles, exported PDFs, invitations. That is distribution you already own and probably are not using.
How "form builder" went from nowhere to position two
Around 175,000 visitors a month sit behind that term. Typeform was not ranking for it. They did not fix it by writing a better page. They fixed it by pointing the site at it.
-
Find the pages that deserve the equity
Some pages bring traffic. Some convert. The overlap is a short list, and it is the only list that should receive links.
-
Get it into the navigation
Header or footer placement means consistent crawling and homepage authority flowing down. Necessary, but not sufficient on its own.
-
Mine the existing content for contextual links
This is the step that actually moved rankings. They found each natural mention of the phrase and linked from there, varying anchor text on purpose.
-
Make it part of publishing, not an annual audit
The durable version is a rule: nothing publishes without ten internal links to priority pages on day one.
boltDo this on your site
Pick your single most commercially important page this quarter. Run a site search for its main phrase and link from every relevant mention you find. Most sites are sitting on forty unused internal links and a ranking they could have had months ago.
Make SEO arguable in the language your CFO uses
Everything above only holds together because SEO was measured in revenue. A growth model is what makes that connection explicit, so you can defend a content decision in a board meeting instead of showing a traffic chart and hoping. Four layers, built in this order.
- Map every page that has SEO value. Primary keyword for each, and a check that no two pages are fighting over the same term. Cannibalisation splits authority and holds both pages down.
- Add keyword research as page-level decisions. Every keyword that matters gets a page of its own. Pull candidates from customer interviews, paid search data, and sales-call objections.
- Layer on performance data. Clicks in the last 30 days, total traffic potential, the gap between them, topic cluster, referring domains, search intent.
- Close it with conversion maths. Visit-to-signup rate for comparable pages, expected signups, expected paying customers, and the link investment needed to reach the target position.
Sort by upside against achievability against conversion rate, and the model tells you what to build next. The debate ends, and the work starts.
Three rules worth stealing
None of this came from publishing volume. It came from a system.
Buying intent beats search volume
A term with 1,000 searches and a buyer behind it outperforms one with 100,000 searches that converts at a fraction of a percent and churns in ninety days. Every time.
The product can be the strategy
Every use case is a keyword. Every embed is a link. You do not need a forty person editorial team to own a category, you need to look closely at what your product already does.
Plan topics, then keywords
Keywords chosen in isolation fail on authority, not on copy. Build the cluster that makes the competitive term winnable before you go after it.