Discussion starter: a B2B and SaaS SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- SaaS SEO
Product-led SEO like public templates, calculators and user pages: where did it drive signups, and where did it just inflate traffic?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
Product-led pages drive signups when the searcher’s job continues inside the product after they land. They inflate traffic when the job ends on the page, or when the people arriving would never buy the product. That one test explains most of the difference between templates, calculators and user pages, and it tells you more than the traffic curve does.
A caution before any examples. Public product-led SEO case studies mostly report traffic, often estimated by third-party SEO tools rather than taken from the company’s own analytics. Very few report signups or activation by page type. So the public record mostly answers “where did it drive visits”. Whether it drove signups on your site is something you have to instrument, and the way you instrument it decides whether the answer is believable.
The test that separates signup surfaces from traffic surfaces
Searcher lands on the page↓Does their job end here? (answer read, file downloaded, person found)↓Yes: a traffic and links asset. Judge it on cost, links earned and reach, not signups.↓No: does finishing the job require your product?↓Yes: a genuine signup surface. No, any tool will do: expect weak conversion.Compare “notion weekly planner template” with “weekly planner template”. The first searcher already uses the tool, and the next step is duplicating the template into their workspace. Many people behind the second query want a printable file or a Google Doc. A SaaS template page can rank for both, but only the first audience has a reason to create an account.
How each asset type tends to behave
Asset Tends to drive signups when Tends to inflate traffic when Main risk Public templates The template only works inside the product (duplicate or import into a workspace) and the query names the product or a tool category The query is generic and a document or PDF satisfies it, or the template downloads without an account Near-duplicate template pages at scale, and cannibalisation with feature pages Calculators and free tools The output is the first step of recurring work the product does, and saving or sharing the result needs an account The calculation is a one-off, Google shows its own calculator or an AI answer, or the audience sits outside your customer profile (students, consumers on a B2B tool) Maintenance cost, and accuracy liability for finance or health calculations User pages (profiles, portfolios, public boards, booking pages) The page shows the product’s output to peers who could make their own, ideally as part of normal use The traffic is navigational for a person’s or company’s name, or most profiles are empty User-generated spam and index bloat Integration pages (“X + Y”) The searcher already uses one of the tools and connecting them requires your product Pages exist for integrations with little real functionality Thin pages nobody needs Integration pages were not in the question, but they are worth including because they usually pass the job test most cleanly. The searcher is already a user of an adjacent tool, the query is specific, and the only way to finish the job is to set up the connection.
Why user pages carry the most risk
User pages scale automatically, which is their appeal and their problem. Most search traffic to profiles is someone looking for the person or business on the profile. That visitor wanted them, not you, and the user who owns the page is often already a customer. The exception is where the page shows the product in use to exactly the people who would use it next, such as a public booking page or a portfolio that other practitioners browse. Those can create a loop where viewers become users.
The index-quality risk is real. Every signup can create a page, including signups from spammers who want a link or a hosted page on a domain with some authority. Google’s spam policies define user-generated spam as “spammy content added to a site by users through a channel intended for user content”. They list “sites designed to allow user-generated content, such as a forum website or comment sections” among things not considered site reputation abuse. So the practical danger is less a site reputation problem and more a directory full of thin or spammy pages.
Google’s own guidance on preventing abuse is directly usable here. It suggests “adding the
noindexrobotsmetatag on posts that come from new users that don’t have any reputation on your platform” and “adding anofolloworugcrelattribute to all links in untrusted content”. In practice I would keep new profiles noindexed and out of sitemaps until they pass a threshold: verified account, a minimum amount of published work, some activity, and no abuse reports.Watch for this:
Template and calculator variants generated in bulk carry their own risk. Five hundred “[profession] invoice template” pages where only the title changes fit Google’s scaled content abuse examples, which include “using generative AI tools or other similar tools to generate many pages without adding value for users”. A variant earns a page when its content actually differs for that profession, not when the keyword does.
How to measure whether it drove signups
This is where most product-led SEO reporting goes wrong. Last-click analytics credits the page that preceded the signup session, and product-led visitors often sign up days later through a branded search or a direct visit. Meanwhile, existing customers using templates and integration pages inflate the session count. Both errors point the same way: traffic looks big, and attribution looks either invisible or inflated depending on the model.
- Store the first landing page and its page type at signup, server-side, alongside the last-touch source. Report both.
- Split logged-in visitors out. Existing users on template and integration pages represent retention or expansion value, not acquisition. Mixing them in makes conversion rates look worse and new demand look bigger.
- Track the in-page action: template duplicated, calculation run, result saved, booking made. A high action rate with low signups points to the gate. A low action rate points to the audience.
- Report activation, not just signup. People who sign up only to save a calculator result often never reach the product event that signals real value.
- Check customer-profile fit: role, company size, and business versus free email domains, by page type.
- Normalise per 1,000 organic sessions by page type, and keep absolute numbers next to the rates so small page types do not look heroic on tiny volumes.
Hypothetical example:
A scheduling SaaS runs three product-led surfaces. Over a quarter, calculators bring the most non-logged-in sessions, templates fewer, and public booking pages fewer still. Per 1,000 sessions, calculators produce the fewest signups and the lowest activation, because most visitors are employees working out a one-off cost. Templates convert better, mostly among people already searching with a tool name. Booking pages convert modestly, but those signups activate at the highest rate, because the visitor has just used the product as a guest and wants their own page. Ranked by traffic, calculators win. Ranked by activated accounts, booking pages win. (Invented scenario, for illustration only.)
When “just traffic” is still worth keeping
A calculator that never produces signups can still earn links that support commercial pages through internal linking, and it can put the brand in front of future buyers. Both effects are plausible but hard to prove, so I would treat them as reasons to keep an existing asset, not as reasons to build more of them. The comparison that matters is cost: engineering and maintenance time, plus accuracy risk, against what the same effort would produce on pages that pass the job test.
Exposure to results-page answers also differs. Common calculations are often answered on the results page itself, by Google’s own tools or by AI answers. Tools that need personal inputs, produce a result worth saving, or connect to the user’s own data are less exposed than tools doing arithmetic anyone can reproduce.
Signals that a surface is not earning its place
- Growing counts of “Crawled - currently not indexed” or “Discovered - currently not indexed” in a template or profile directory in Search Console’s page indexing report
- High sessions with low in-page action rates, meaning the audience is wrong rather than the page
- Signups from the surface that rarely activate, or that sit mostly outside the customer profile
- Template or tool pages ranking for queries your feature pages should own
- Moderation workload on user pages rising faster than the signups they produce
What to test
Product-led pages lend themselves to page-group tests better than most SEO changes, because there are many similar pages. Two tests I would run:
- Gate placement on templates. Split template pages into matched groups by traffic and topic. In one group, make “use in product” the primary action and the download secondary. Primary metric: activated accounts per 1,000 non-logged-in sessions. Secondary: organic clicks, since changing what the page offers can change how well it satisfies the query. Run it for six to eight weeks and watch seasonality around planning periods such as the new year.
- Indexing threshold on user pages. Noindex profiles below a quality threshold for one cohort of new accounts. If organic sessions drop but signups and activation from user pages barely move, that traffic was inflating the chart.
Neither test proves causation cleanly, because organic traffic reacts to many things at once. They are still far better evidence than a traffic chart annotated after the fact.
The decision rule
If I were deciding where to invest, I would rank existing product-led surfaces by activated accounts per 1,000 non-logged-in sessions, not by sessions. Surfaces where the searcher’s job continues in the product get more investment. Surfaces where the job ends on the page stay only if their cost is low and they earn links. User pages stay indexed only when they pass a quality threshold. When someone presents a product-led SEO win, the question to ask is “activated accounts from which page type, attributed how?”, because without that answer the win is probably traffic.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.