If you run a small business website, you have probably been pitched a link-building package. The pitch is almost always the same shape: a guaranteed number of backlinks per month, from a network of "high-authority" sites, for a flat monthly fee. Some of these packages cost real money, and the SEO backlink-building services category is a market with agencies charging serious retainers for exactly this. Before paying for one, it is worth knowing what actually happens to sites that buy links without the traffic or substance to back them up.
The number that should change how you read the pitch
Stackra tracked Common Crawl's link data against Chrome traffic data (CrUX) across roughly 9.7 million domains over three years. Inside that data, we found the exact group most likely to prove a link-buying pitch right: 60,291 sites that spiked their referring domains while sitting at zero real traffic to begin with. We tested the outcome 12 different ways, three gain multiples times four ending link-count bands, so no single cutoff could be quietly doing the work.
| Cohort | Reached real traffic |
|---|---|
| Full 60,291-site cohort, every one of 12 threshold variants | 0.4% or less |
| Single best-performing cell in the whole table | 0.26% |
| Tier-stratified rerun, 118,660 sites, every starting tier below the top | flat, no relationship |
Source: Stackra analysis of Common Crawl webgraph joined to CrUX traffic rank, cross-validated against DataForSEO's commercial backlink index (0.94 rank correlation on a 500-site sample).
What actually did work in the same data
This is not an argument that links never matter. A separate panel isolated sites with no visible traffic at the start and tracked what happened as their referring-domain count grew: the rate of breaking out into real traffic rose in a clean line from 0.82% at zero link growth up to 7.49% at +1,000 referring domains gained, a real, roughly 9x improvement. But 92.5% of even the biggest link gainers stayed stuck. For sites that already had real traffic and real, organic links, more referring domains genuinely correlated with better outcomes, up to a point. At the bottom traffic tier, the sweet spot sat at 101-300 referring domains, with a 34% quarterly improvement rate. Past that, more links started correlating with worse outcomes, not better. The difference between that cohort and the 60,291-site failure group was not the tactic. It was whether there was a real, visited business behind the links in the first place.
Questions worth asking before you pay for a link-building package
A few direct questions separate real digital PR from the volume-selling pattern that failed in the data above.
- •Where specifically are the links coming from? "A network of high-authority sites" with no named examples is a red flag. Ask for real, live examples of placements on other client sites.
- •Is the volume guaranteed regardless of relevance? A guaranteed link count per month, independent of whether a link makes sense for your industry or content, is closer to the manipulation pattern in the data above than to real digital PR.
- •Does the pitch mention your actual content or audience at all? Real link building (digital PR, guest posts, broken-link outreach) requires something worth linking to. If the pitch never asks what that is, it is likely selling placement volume, not relevance.
- •What happens to your traffic if you stop paying? If the answer is nothing changes either way, that is consistent with links that were never doing real work in the first place.
When outsourcing link building can make sense
Digital PR agencies that pitch journalists on genuinely newsworthy angles, or writers who place real guest content on relevant, actual-audience sites, are doing legitimate work that this data does not argue against. The distinguishing feature is always the same: the link exists because someone with a real audience decided your content or business was worth mentioning, not because a fee was paid for placement regardless of relevance.
The takeaway
If a package is selling you a guaranteed volume of backlinks with no connection to your actual business or content, the closest match in this data is the 60,291-site cohort that failed 99.6%+ of the time. Spend that budget on the thing that makes someone want to link to you in the first place, and the links tend to follow.
Frequently asked questions
Common questions from businesses evaluating a link-building purchase.
Does buying backlinks work?
Based on data from 5 million websites, buying links or spiking your link count without real traffic or substance behind it fails 99.6%+ of the time. Among 60,291 sites that tried it, tested 12 different ways, the single best-performing variant still only succeeded 0.26% of the time.
Is it worth paying an agency for link building?
It depends entirely on what they are actually doing. Digital PR and guest placements on genuinely relevant, real-audience sites can be worth it. A guaranteed volume of backlinks from an unnamed network, independent of relevance, matches the pattern that failed almost universally in this data.
How much should link building cost?
This data does not evaluate specific pricing, but it does show that the value of a link depends entirely on whether it comes from a real, relevant site that chose to link to you, not on how many links a fee buys. Price a link-building service on the quality and relevance of placements, not the guaranteed quantity.
What is a safer alternative to buying backlinks?
Original research, genuinely useful guides, and outreach to real, relevant sites (digital PR, guest posts, broken-link outreach) all produce links because someone with an actual audience chose to link to you. That is the pattern behind the sites in this data that saw real, lasting benefit from their link profile.