How to read these numbers
The three sites below are real, publicly visible businesses. The authority and traffic figures are third-party estimates from public SEO tools, rounded, and plotted month by month.
They are estimates, not analytics exports. Domain Rating is Ahrefs' 0-100 logarithmic authority metric, and organic traffic estimates are modelled from keyword rankings, not measured.
None of these sites ran cloud stacks in isolation. Each ran content work and other link building at the same time. Treat the charts as context for a method, not as proof of causation.
We include one case where the link layer did nothing. A case study set with three wins in it is a brochure.
Correlation is all a case study can honestly give you. Anyone promising causation from one campaign is selling something.
Case 1: a UK trades site (Checkatrade-listed contractor)
A small contractor site with fewer than 40 referring domains at the start. The link layer was mostly volume: cloud stacks pointed at service-area pages, plus a handful of local citations.
Profile at the start: DR 8, 37 referring domains, 14 indexed service-area pages with no external links at all.
What ran: roughly 40 stack links a month for six months, region-matched, pointed at service-area pages rather than the homepage, plus NAP-consistent citations.
Authority moved slowly and predictably. Traffic moved later, which is the normal lag — links get crawled in days, rankings settle over months.
The useful detail: pages that received stack links picked up long-tail impressions first. Head terms barely moved, and the map pack position did not change at all. That matches what we describe in cloud stacking for local SEO — links feed prominence, and prominence is only part of the local picture.
Third-party estimates, rounded. Not analytics exports.
Case 2: a SaaS knowledge-base tool (eniston.io)
A small European SaaS with a strong product page and almost no links to its documentation and comparison pages. Stacks went to those supporting pages, not to the homepage.
Profile at the start: DR 21, a few hundred referring domains concentrated on the homepage, and a documentation section with effectively zero external links.
Domain Rating barely moved, which is expected — low-authority links do not shift a logarithmic metric much. Traffic to the linked subpages is where the change showed up.
This is the pattern we see most often with software companies: the metric stays flat while the internal competitiveness of specific URLs improves. If you only watch DR, you will conclude nothing happened.
What we would do differently: more of the budget on two or three reseller guest posts aimed at the comparison pages, with the stacks pointed at those posts instead of directly at the site.
Third-party estimates, rounded. Not analytics exports.
Case 3: an established ecommerce brand (Gymshark-tier competitor set)
A mid-size clothing retailer competing against brands with tens of thousands of referring domains. Here the stack layer was a rounding error.
The chart shows the honest outcome: the trend follows the brand's digital PR and seasonal demand, not the volume links. The November spike is Black Friday, not link building.
Why it failed: at this profile size, adding 300 low-authority links changes the referring domain count by a fraction of a per cent. There was nothing for the links to move.
We include it because the null result is the point. If your competitor set looks like this, spend on placements people actually read. Volume links belong underneath them, not instead of them.
Third-party estimates, rounded. Not analytics exports.
What the three have in common
Small profiles moved. Large profiles did not. That is the whole lesson, and it matches what the most underrated affordable link building strategies show about indexing and batch quality.
Indexed links moved things; unindexed ones did nothing at all. In every batch, roughly one link in five never entered the index and can be treated as spend with no return.
Supporting pages beat homepages. In both cases where something happened, the movement was on URLs that had no external links before the campaign started.
Attribution stays messy. Use Google Search Console's performance report on the specific linked URLs rather than sitewide traffic if you want the cleanest read available.
How to run this test yourself
You do not need our data. Run a controlled version on your own site in one quarter, for the price of a small batch.
- Pick ten pages with similar traffic and no external links. Link five, leave five alone.
- Record impressions and average position for each in Search Console the week before you start.
- Order a small batch of cloud stacking links pointed only at the five, spread over three weeks.
- Check index status at day 14 and discard unindexed links from the analysis.
- Compare the two groups at week 12, not week 4.
Marcus Scott
Head of fulfilment, Cloud Stacking
Marcus has built links for a living since 2011. He now runs the team that publishes every stack we deliver, and he still checks the index reports himself.
Ready to test a batch?
Ten links, built and indexed the same day, with the full CSV report attached.
Create an account