First-party data · Updated September 6, 2026
What growth services do buyers actually click?
Every figure below is a filtered count of click-throughs recorded on our own outbound links, updated daily. It is what readers actually clicked on the pages we publish — which means it is shaped by our own layout as well as by reader interest, because we choose which services a page lists and in what order. Read it as the share of our outbound clicks, not as market demand.
30,578 click-throughs recorded between April 17, 2026 and September 6, 2026·13,530 in the last 30 days.
01 · By service
Where reader clicks go
Share of all tracked click-throughs, by service. The small figure is the raw click count; a green ▲ marks services winning a bigger share in the last 30 days than they hold all-time.
Showing the top 12 of 24 services that clear our publication floor of 100 clicks. Shares are of every counted click, so this table does not add up to 100%: 359 counted clicks sit below the floor and are not shown. The full list of published rows is in the dataset.
Why the top row is the top row: Likes.io is the service we list first on 34 of our 47 ranked buying guides. A service listed first on almost every page collects clicks that a service listed eighth cannot, so its share here measures our placement at least as much as it measures reader preference.
02 · By platform
Which platforms buyers grow
Share of tracked click-throughs by the platform the buyer was shopping for.
Platform is read from the page the click came from, so this table covers only clicks that started on a buying guide for a named platform. 2,698 of the 30,578 counted clicks started somewhere else — a review, a comparison, the home page — or fall below the publication floor. They are excluded from the rows below but still counted in the total each share is divided by.
How to read this
What this is: an aggregate, anonymized count of click-throughs recorded on our outbound links, refreshed daily. Exactly two filters apply: known bots, crawlers, and link scanners are excluded by user-agent, and a click only counts if the same session also loaded a page on this site. The second filter matters, because our redirect is a public URL that automation hits without ever reading anything. What does not apply: our internal analytics also drop visits whose browsing pattern looks automated — eight or more pages spread over four or more hours — and this public dataset does not, so these totals are looser than the verified figures we use internally. The counts are of clicks, not people: one reader clicking twice counts twice.
What this is not: it is a click, not a purchase or a confirmed sale, and it is not our editorial ranking. Popularity is not quality. Which service we actually recommend, and why, is decided by our published rubric and editorial scoring (plus a paid hands-on pilot of five Instagram services), not by clicks. See how we rank.
What we do not capture: impressions, how far down a page a service's card sat, click-through rate, and which page template the click came from. None of those are recorded against an outbound click anywhere in our logs, so no share on this page can be adjusted for how much exposure a service got. That is the single biggest limit of the dataset, and it is why the placement note above is a count of pages rather than a correction to the figures.
The underlying figures are available as a machine-readable dataset for anyone who wants to cite or verify them. It carries the same filters, floors and coverage figures as this page.
Corrections and changes
- August 26, 2026
- Removed the claim that the figures “reflect live demand instead of an opinion”, and the claim that they “track human reader interest as closely as server logs honestly allow”. Neither was supportable: the counts are clicks on links this site places in positions this site chooses, and a stricter automation filter exists in our own analytics code than the one applied here. Added the placement disclosure with its measured page counts, the publication floor, the display cap, the coverage figures that reconcile the tables to the total, the start of the recording window, and an explicit list of the things this dataset does not capture.
- July 28, 2026
- Tightened the click definition: a click now counts only when the same session also loaded a page on the site. The previous user-agent-only filter counted automation that hit the public /go redirect without ever reading anything, and every figure on this page dropped accordingly.
The figures themselves change every day as new clicks are recorded; this list covers changes to what the report claims, not to the numbers it reports. Spotted something wrong? Tell us.