This is an argument we have an obvious commercial interest in making, so treat it accordingly. What follows is the version we would write if we were on the client side, which is where both founders were when they learned it.
The case against cheap links is usually made as a warning about penalties. That is the least interesting part, because penalties are rare. The stronger case is arithmetic, and it holds even if nothing ever goes wrong.
Bill 1 — the placements do nothing
Arrives: immediately, invisibly.
A link on a site with 200 monthly visitors produces no referral traffic, no brand exposure, no credibility with a buying committee, and no chance of being a source an answer engine draws on. Its entire theoretical value is a ranking effect.
And that effect is small, because the site passing it has little to pass. You have bought forty links and closed zero of the gap that mattered, because none of the forty were in it.
Cost: the full amount paid. Not a discount — a purchase of nothing.
Bill 2 — the quarter
Arrives: month four to six.
The larger and less visible cost. Six months spent on inventory is six months during which the four competitors above you kept acquiring properly. You have not stood still; the gap has widened while you were paying to close it.
In a category where competitors add roughly 20–30 referring domains a year, a wasted year is a real setback rather than a neutral pause.
Cost: the spend, plus the movement you did not make, plus the additional gap to close afterwards.
Bill 3 — attrition
Arrives: months six to twenty-four.
Low-quality sites disappear at markedly higher rates than genuine publications. They get sold, deindexed, abandoned, or pruned when the operator rebuilds the network.
Nobody offers a warranty at this end of the market, so replacement is a repurchase. And because nobody is monitoring, the losses are usually discovered at an annual audit rather than at the time.
Cost: a material share of the original spend, again.
Bill 4 — remediation
Arrives: when someone finally audits.
This is where it stops being a matter of wasted money and becomes a project. Classification of every referring domain, footprint clustering, removal outreach, a disavow file, and a quarter with no acquisition while the profile contracts.
| Remediation item | Typical cost |
|---|---|
| Full profile classification | $2,000–$6,000 |
| Removal outreach | $1,500–$4,000 |
| Disavow file and monitoring | $1,000–$2,500 |
| A quarter of paused acquisition | Three months of lost progress |
In our own engagements, roughly one client in five arrives needing this. It is the least popular invoice we send and the one that most reliably precedes a good year.
Bill 5 — the human one
Arrives: rarely, expensively.
Backlink profiles are public. In enterprise sales, a competitor forwarding your prospect a screenshot of yours is a thing that happens — not often, and it is not recoverable in the deal where it happens.
The same applies during due diligence. Acquirers and investors run technical audits, and a profile full of network inventory is a finding that has to be explained.
The comparison people never run
Option A: 40 links at $100 = $4,000. Roughly 8 in the gap, most gone within two years, remediation likely later.
Option B: 10 links at $400 = $4,000. All 10 in the gap, ~96% still live at twelve months, nothing to unwind.
Option A looks like four times the volume. It is one fifth of the outcome, and it carries a liability the other does not.
Why the cheap end exists
Not because those suppliers found efficiencies. Because they removed the two most expensive parts of the job.
Work through a single placement at realistic rates: prospecting and scoring ≈$35, pitching ≈$25, writing ≈$90, verification and monitoring ≈$15. That is $165 in delivery cost before any publisher fee, margin or overhead.
A $99 placement has not compressed that. It has skipped prospecting and skipped vetting, and sourced from inventory that requires neither — which is the entire business model.
Where cheap is genuinely fine
Three cases, because the argument should not be absolute.
- Directory and partner listings. Effectively free, entirely legitimate, and almost nobody claims them. Cost per link under $50 in staff time.
- Broken link replacement. Low yield per attempt, negligible cost, produces a steady trickle.
- A genuinely small publication with the exactly right readers. Low price because the audience is small, not because the site is inventory. These are frequently the best links available in a niche B2B category.
The distinction is not price. It is whether the publication would exist without link buyers.
If you already bought them
Do not panic and do not mass-disavow on a tool's toxicity score. Sequence:
- Classify. Export every referring domain and score it manually. You need a count before a plan.
- Assess proportion. Twelve network links in nine hundred is different from three hundred in nine hundred.
- Request removal first, particularly if a previous agency arranged the placements — they have the relationships.
- Disavow what remains at domain level, if the proportion is material.
- Pause acquisition while the profile contracts, then ramp gently.
The most expensive sentence in this business is "we'll start cheap and upgrade later". You do not upgrade later. You remediate later, and then you start.