By 
Shreyas M V · Published · 12 min read
Zapier's Integration Pages: How App-to-App Pages Became a Search Moat
Zapier's integration pages worked because each one pairs a task people search for with data only Zapier holds, the triggers and actions of the apps involved, and a database lets it publish that template for every app and app pair. The sources document the page families and the traffic, but none of them measures how much traffic the integration pages alone produce. This teardown sorts the families, compares the templates, reconciles figures that disagree and ends with a usefulness test. It did not crawl Zapier, so it gives no page counts of its own.
- Sources describe four page families: app profiles, app-to-app pages, three-app workflow pages and editorial list posts.
- Each family carries data Zapier says is unique to its site: the triggers and actions of the apps involved.
- The traffic figures in the sources cannot be reconciled, because they measure different things on different dates.
- A six-line usefulness gate at the end tests whether your own integration pages are real pages or thin ones.
Why did Zapier's integration pages work?
Zapier's integration pages worked because each page answers a specific connection task with data that only Zapier has, and a database let the company publish that page for every app and every pair.
That reading comes from three sources: the Zapier blog's own programmatic SEO guide, the Practical Programmatic case study and the Salt.agency case study. They agree on the structure and disagree on the numbers.
The word moat is an interpretation. The sources document pages and traffic, not defensibility, and none separates the integration pages from the rest of Zapier's search traffic.
What this teardown measures and what it cannot
This teardown measures only what the three sources give, and it adds arithmetic on top of their figures. It did not crawl Zapier's site, so it contains no page counts, no template screenshots and no link graph of its own.
The three pages the research package analysed run 2,342, 470 and 1,423 words, and its page analysis found no table and no FAQ on any of them. They describe the pages in prose, with a few traffic examples.
What this page adds is organisation and testing: the families side by side, the templates compared point by point, the figures lined up with their dates, and a gate built from the criteria the sources imply. Each statement below is labelled as documented, reported, derived by arithmetic or inferred.
The page families
Salt.agency describes three tiers of integration page, and a fourth family of editorial posts sits beside them: app profiles, app-to-app pages, three-app workflow pages and list posts.
The first tier is a landing page for each app, such as Trello, Slack or WooCommerce. The second is a page for every app-to-app pairing, and the third is a page for a three-part workflow, for example WooCommerce to Slack plus Trello.
| Family | What it targets | Example query or title | Described by |
|---|---|---|---|
| App profile page | One app and everything it connects to | slack integrations; google sheets integrations | Practical Programmatic, Salt, Zapier |
| App-to-app page | One specific pair of apps | how to connect {app 1} + {app 2}; trello slack integration | Practical Programmatic, Salt, Zapier |
| Three-app workflow page | A workflow across three apps | woocommerce slack trello | Salt only |
| Editorial list post | A category query that is not about Zapier | best calendar apps | Salt only |
The Zapier blog gives the generating rule in plain terms: a head term of the form [Tool 1] and [Tool 2] integration, with every supported app as a modifier. Its example is a person searching for Google Sheets and Notion who lands on a page about that exact connection.
What each template contains
Every template the sources describe has a short description, a call to action and a list of triggers and actions, and the families differ in which extra blocks they add.
Practical Programmatic lists the data points of profile pages and pair pages separately, and Salt.agency gives one combined list. The table below puts the three lists side by side so the differences are visible.
| Data point | Profile page, Practical Programmatic | Pair page, Practical Programmatic | Salt.agency list |
|---|---|---|---|
| Short description | Listed | Listed | Listed |
| Call to action | Listed | Listed | Listed |
| Supported triggers and actions | Listed | Listed, for both apps | Listed |
| Popular integrations or workflows | Listed | Listed, for both apps | Not listed |
| Other apps this app connects to | Listed | Not listed | Listed |
| Related blog posts or tutorials | Not listed | Listed | Listed |
The pattern in the table is that a profile page leans toward navigation, pointing to other apps, while a pair page leans toward the specific job, listing workflows for both named apps and tutorials about them. That is an inference from the lists, not an observation of the live pages.
The keyword ladder from app to workflow
Search demand falls steeply as the page gets more specific, from 1,400 monthly searches for a single-app query to 10 for a three-app workflow, according to Salt.agency's sample.
| Keyword | Page family it fits | Monthly searches |
|---|---|---|
| slack integrations | App profile | 1,400 |
| trello integrations | App profile | 400 |
| trello slack | App-to-app | 400 |
| trello slack integration | App-to-app | 300 |
| woocommerce slack trello | Three-app workflow | 10 |
The derived ratio is 1,400 divided by 10, or 140 to 1, between the top and bottom rungs. Salt.agency also reports that Zapier's own category term, automation platform, draws about 400 searches a month, which is why it says the team targeted what customers search for instead.
The takeaway is a shape, not a forecast. Each pair page sits on a small query, so the strategy works only when there are very many of them.
Why the traffic figures disagree
The traffic figures disagree because they count different things on different dates, and only one pair of them, both from Salt.agency, can be compared directly.
| Figure | What it counts | Date | Source | Status |
|---|---|---|---|---|
| 1.19 million monthly visitors | Zapier's website | November 2020 | Salt.agency | Documented by Salt |
| 4.8 million monthly visitors | Zapier's website | November 2023 | Salt.agency | Documented by Salt |
| 2.6 million organic visits a month | Pages the source calls programmatic | Not dated | Practical Programmatic | Documented by the source, method not stated |
| 7.3 million visitors a month | Not stated in the snippet | Not dated | Reforge (search snippet only) | Pointer only, page not read |
| 2 million monthly unique visitors | Not stated in the snippet | Not dated | buildd (search snippet only) | Pointer only, page not read |
| 9M | Title cuts off, so the metric is unknown | Not dated | demanzo (search title only) | Pointer only, page not read |
Salt.agency's two figures give a growth rate. The inputs are 4.8 million in November 2023 and 1.19 million in November 2020, so 4.8 divided by 1.19 is 4.03 times, or growth of about 303%, which matches the source's claims of quadrupling and of over 300%.
Salt.agency also reports 6 million backlinks, 1.4 million keywords in the top 100 and 42.8 thousand in the top 10. The top-10 share is 42,800 divided by 1,400,000, or about 3.1%.
The 2.6 million figure from Practical Programmatic cannot be set against the 4.8 million, because one is dated and one is not, and neither states a method. The honest summary is that the sources give multi-million figures for Zapier's search traffic and none of them states what share belongs to the integration pages.
What the page counts and samples imply
The sources imply hundreds of thousands of integration pages, far fewer than the theoretical number of app pairs, and a sample that suggests most pair pages earn very little on their own.
Practical Programmatic says Zapier created hundreds of thousands of pages, and the Zapier blog says it rolls out a page for every combination of apps. Salt.agency says Zapier claims to connect to more than 6,000 apps.
Taking 6,000 apps, the number of unordered pairs is 6,000 times 5,999 divided by 2, which is 17,997,000, about 18 million. That is a ceiling on pair pages, not a count, and it sits well above hundreds of thousands, so either not every pair has a page or the app figure is looser than it sounds. No source says which.
| Sample page | Family | Monthly organic traffic |
|---|---|---|
| Gmail integrations | App profile | 60,000 |
| Google Docs integrations | App profile | 42,000 |
| Google Sheets integrations | App profile | 34,000 |
| How to connect Descript + Notion | App-to-app | 1,100 |
| How to connect Steam + Twitter | App-to-app | 900 |
| How to connect Google Calendar + Notion | App-to-app | 800 |
The three profile samples sum to 136,000 (60,000 plus 42,000 plus 34,000), and the three pair samples sum to 2,800 (1,100 plus 900 plus 800). The profile sample is about 49 times the pair sample, and the three profile pages alone equal about 5.2% of the 2.6 million figure.
Now take the 2.6 million and read hundreds of thousands as at least 100,000 pages. The mean is then at most 2,600,000 divided by 100,000, or 26 visits per page per month, far below the 933 mean of the three pair samples. The listed examples are presumably the strong ones, so the typical pair page earns little, and the moat is the sum of a very long tail. That last step is inference.
What made the pages useful, not thin
The pages count as useful on the evidence available because they hold unique data, match a real task and offer a next step, while their upkeep is the one criterion no source can confirm.
The test was assembled in three steps. Each criterion comes from the Zapier blog's own advice on avoiding thin content, each is checked against what the sources say about Zapier's templates, and each reading is Kanso's judgement.
| Test | What the sources show | Label | Reading |
|---|---|---|---|
| The page holds data no one else has | Zapier says the triggers and actions of each app are dynamic and unique to its own site | Documented by Zapier | Strong for every family, because every template lists them |
| The page matches a task | Zapier says the pages catch users with an exact problem at the moment it can solve it. Practical Programmatic calls the queries transactional | Documented by sources | Strong on pair pages, whose query names both apps |
| The page differs from its siblings | Pair templates list workflows, triggers and actions for both named apps, plus related tutorials | Documented by Practical Programmatic | Likely, since the content changes with the pair, but no page was compared here |
| The page offers a next step | A call to action appears in every template described | Documented by two sources | Met |
| The page sits in a hierarchy | Profiles list connectable apps, pairs link tutorials and posts; Salt credits internal linking with hierarchy and authority | Reported by Salt | Plausible, but no link graph was captured |
| Someone can maintain it | Zapier warns that more pages mean more tech debt and content debt | Documented by Zapier | Unknown: no source says how Zapier refreshes the pages |
The Zapier blog says the unique data on its pages is the set of triggers and actions for each app, which are dynamic and completely unique to its website. It also warns that skipping the template planning step risks Google no-indexing the pages.
The plain conclusion is that the pages avoided thinness because each one carried a block of structured data that could not be written from outside the company. That is Kanso's reading of the Zapier blog's criteria, not a finding from inspecting the pages.
How the families link to each other
The sources show the families linking in two directions, profile pages to other apps and pair pages to related tutorials and blog posts, but none captured the actual link graph.
Practical Programmatic lists a block of other apps an app can connect to on profile pages, and related tutorials and posts on pair pages. Salt.agency names related blogs and articles as a data point and says internal linking gives the site hierarchy and authority.
Salt.agency adds that the editorial list posts, such as best calendar apps, carry calls to action for connecting those apps through Zapier, and says that article type ranks for over 400,000 keywords. The logic it gives is that someone choosing a calendar app is a plausible buyer for an automation involving it. That is Salt.agency's reasoning, not a measured conversion.
If you want a sibling teardown, Kanso's look at how Stripe's documentation doubles as marketing applies the same documented-versus-inferred method to a different company.
Is it really a moat?
The evidence supports calling it a moat only in a narrow sense: the data is hard to copy and the page count is hard to match, but no source tests defensibility.
- Secondary sources. The traffic figures come from an agency and a case-study site, neither of which states a method, and the Zapier blog gives no traffic numbers for its own pages.
- Maintenance cost. The Zapier blog says more pages mean more technical and content debt, and that the content must stay current.
- An uncertain future. The same article says generative AI may make programmatic SEO less valuable over time, and it was last updated in February 2025.
- Copyability. Anyone can copy the template; the difficulty is the per-app data. A founder with a large integration catalogue holds a version of the same asset.
Salt.agency and the Zapier blog also describe product-led growth, a free tier of 100 tasks a month and an app catalogue as part of the picture. So the pages are one lever in a company whose product is itself the catalogue.
A usefulness gate for founders with integrations
A founder with integrations should publish a page family only when it passes six checks, built from the criteria in the usefulness test above.
real task → unique data → visible variation → measured demand → linked → maintained
| Check | How to test it | Why it matters |
|---|---|---|
| 1. Unique data | Name the field a competitor cannot copy from your docs | A pair page with no unique field is a thin page |
| 2. Real task | Write the query a person would type and check the page answers it | No task, no page |
| 3. Pair variation | Open five random pairs and compare them side by side | If only the app names differ, do not publish that family |
| 4. Demand | Check search volume for the pattern, not one page | Zapier's third rung shows 10 a month, so volume is a long tail |
| 5. Links | Make sure each page is reachable from a profile or hub | A page nothing links to is hard to find and hard to rank |
| 6. Upkeep | Name who updates the data and how often | Stale triggers make a useful page wrong |
Run the gate on a sample of ten pairs before generating a thousand. If the pairs fail check three, the template is the problem, and more pages will multiply it.
After publishing, check whether the pages reach the index. Kanso's decision tree for Search Console statuses separates pages that were never crawled from pages that were crawled and left out.
For another integrations-led case, how Razorpay grew with developers covers a company where integrations and documentation were part of distribution.
What is documented and what is analysis
This section separates what the sources document from what Kanso concludes, so a reader can tell evidence from interpretation.
Documented
- Zapier has pages for each app, each app pair and, per Salt.agency, three-app workflows (Salt.agency, Practical Programmatic).
- Pages are database-driven, and the triggers and actions are unique to Zapier's site (Zapier blog).
- Zapier's website had 1.19 million monthly visitors in November 2020 and 4.8 million in November 2023 (Salt.agency).
- Keyword volumes run from 1,400 to 10 across the tiers (Salt.agency).
Kanso's analysis
- The ceiling of about 18 million pairs is arithmetic on a claimed 6,000 apps, not a page count.
- Typical pair pages earn far less than the sampled ones, an inference from sums and an assumption of at least 100,000 pages.
- The pages were useful because of unique structured data, an inference from the Zapier blog's own criteria.
- Nothing here proves the pages are a moat, and nothing here is Kanso's own crawl or traffic data.
Frequently asked questions about Zapier's programmatic SEO
What is programmatic SEO?
Programmatic SEO is a way of creating many search-optimised pages at once from existing data and pre-programmed rules, according to the Zapier blog. The aim is to drive traffic and revenue by putting relevant pages in front of potential customers at scale.
How do you create programmatic SEO pages?
The Zapier blog describes an eight-step process: keyword research and content planning, a page template, collecting data, building a database, using AI to improve quality, building the CMS template, connecting the database to the site, and testing. It gives the head term plus modifier pattern, such as a two-app integration phrase filled with every app name. It also advises validating the idea before rolling it out and warns that skipping the template step risks Google no-indexing the pages.
What is Zapier?
Zapier is a web-based automation platform that connects apps so that an event in one app can trigger an action in another, without code. Practical Programmatic gives the example of creating a to-do item whenever an email arrives with a certain subject line, and Salt.agency says Zapier claims to connect more than 6,000 apps.
What are integration pages?
Integration pages are web pages that describe how a software product connects with other tools. In Zapier's case they come in tiers: one page per app, one per pair of apps, and one per three-app workflow, as described by Salt.agency.
What is Zapier's SEO strategy?
Zapier's documented strategy is to target what customers search for rather than what it calls itself, using database-driven pages for every supported app and app pair. Salt.agency says the term automation platform draws about 400 searches a month, which is why demand for the product category was too small to build on.
How did Zapier accelerate organic growth?
Salt.agency attributes the growth to a mix of product-led growth, tiered integration pages and blog content such as best-app lists, and reports site visitors rising from 1.19 million in November 2020 to 4.8 million in November 2023. No source in this research isolates how much of that came from the integration pages alone.
How much organic traffic does Zapier get?
The sources disagree because they measure different things on different dates. Salt.agency reports 4.8 million monthly visitors to Zapier's website in November 2023, while Practical Programmatic reports 2.6 million organic visits a month from programmatic pages, with no date or method.
How many companies use programmatic SEO?
No source in this research counts them. The Zapier blog names three examples, Zapier, Nomad List and Tripadvisor, and its author says the approach is not suitable for every business.
Sources
- Programmatic SEO: How to do it & if you should, Zapier blog
- Zapier: Programmatic SEO Case Study, Practical Programmatic
- How Zapier Quadrupled Organic Growth | SEO Case Study, Salt.agency
Note: Facts are attributed to the Zapier blog, Practical Programmatic and Salt.agency, as supplied in the research package and retrieved on 17 October 2026. The latter two are secondary sources whose methods are not stated, and their figures are not independently verified here. Kanso did not crawl Zapier or access its analytics, and has no affiliation with Zapier. Traffic and keyword figures change, so check them again before relying on them.
Teardowns