Backlink profiles: what to do when there is no budget to buy links

Conversations about links almost always turn into conversations about buying them, and buying them runs into a budget that does not exist. Meanwhile the profile itself usually holds a month of work that costs nothing to place. This is about that work: what is worth reading in the report, what turns into tasks, and when disavow deserves any time at all.

Why look at the profile at all

A link exists independently of any search engine, so link data applies equally to Yandex and to Google. That is a rare property; almost everything else in search has to be worked through separately per engine.

Second, your own profile is the one part of the link picture you influence directly. Other people's links cannot be removed, while lost ones can sometimes be recovered, and that is cheaper than any purchase.

Four figures the analysis starts from

A link report looks intimidating, but decisions come from four numbers.

• Referring domains, not link count. A thousand links from one site weigh roughly what one does. Domains are the unit.
• The trend over the last year. The profile is growing, flat, or shrinking. A shrinking profile with stable content means you are losing links faster than you earn them.
• Anchor distribution. A healthy one is mostly the brand name and the bare URL. A tilt towards commercial phrases is the trace of a purchase, yours or someone's before you.
• The share pointing at the homepage. If ninety per cent land there, inner pages grow only through internal linking.

Everything else in the report is a breakdown of those four.

Record them quarterly. A single snapshot says almost nothing: a referring-domain count is never good or bad in itself, only larger or smaller than it was, and larger or smaller than the sites ranking beside you. Three or four readings in a row turn a report into a trend, and the trend is the only thing that shows whether what you are doing works.

Lost links, the cheapest work available

The lost-links report almost always contains something recoverable, and almost nobody opens it.

Losses come in four types, three of which are fixed without writing to anyone:

1. The linking page was deleted. Nothing to recover; move on.
2. The link was edited out of the text. The only case needing correspondence, and the outcome depends on the site.
3. The link points at a 404 on your side. Fixed with a redirect in five minutes. This is the most common and the most annoying type: the link is live, the site is willing, and the page is gone.
4. The link points through a redirect chain. Fixed by shortening the chain to one hop.

Types three and four are technical debt in its purest form. Where it comes from is covered in our technical audit checklist.

Unlinked mentions

The second source of free links. Look for pages where the company or product is named in text with no link. Those turn up through a brand search, and some of them sit on sites that already wrote about you voluntarily.

The approach is short and honest: you mentioned us in this piece, here is the page with more detail. Conversion is modest, the cost of asking is zero, and the site has already shown the topic interests it.

The link gap as a target list

The same arithmetic as in competitor work: take the domains linking to competitors and subtract those linking to you. The most useful entries link to several competitors at once, because they are willing in principle to link to companies like yours.

The list earns its place even without acting on it. It shows what class of site mentions your niche at all, and that has a sobering effect: if trade associations and industry press link to your competitors, buying from a marketplace will not move you towards their profile. The competitor side of this is covered separately.

Toxicity and disavow

This attracts the most myths, so in order.

A toxicity score in any tool is that service's formula, not the search engine's opinion. A high percentage on its own indicates neither a penalty nor a need to act.

Google ignores most link spam by default, and disavow is for cases where you have grounds to think the problem is real. In practice that means two scenarios: you know links were bought, or you can see a manual action in the console.

Two technical points, each of which has already cost somebody their list:

• The disavow tool does not accept Domain properties in Search Console. It needs a URL-prefix property, added separately.
• Uploading a file replaces the previous list entirely rather than adding to it. Read the existing file before uploading a new one, or you will quietly delete the old entries.

With no purchase history and no manual action, a month spent on disavow is almost certainly wasted. The same month spent on lost links and mentions returns more.

What to do about anchors

Anchors are the part you do not control directly, and that is as it should be. The controllable share is small: your own profiles, bylines in published material, and internal linking.

Internal anchors are worth tidying because it is free and takes a day. If every internal link says "read more", the engine gets no signal about what the destination is about.

On external ones: a tilt towards commercial phrases is corrected by dilution rather than by replacement. New links carrying the brand name and the bare URL gradually pull the distribution back towards normal. There is no fast route, and negotiating anchor changes on links already placed usually costs more than the result is worth.

What does the work, and what it costs

The webmaster console shows some of your links free, and for a small site that can be enough. A full report with history and gap analysis needs a paid index.

Prices from the gbseo.ru catalogue as of 19 September 2026; "retail" is the vendor's own price for the same subscription. The full set is 4 999₽ a month against 36 700₽ at retail. The calculator on the homepage prices any subset.

For link work, index coverage decides more than the interface does: the more links a platform sees, the longer the lost-links list and the more useful the gap. The difference between the two platforms is in our Semrush and Ahrefs comparison.

A month of work, in order

1. Record the four baseline figures so there is something to compare against next quarter.
2. Work the lost-links report and redirect everything pointing at a 404.
3. Shorten redirect chains to one hop.
4. Find unlinked mentions and write where it makes sense.
5. Build the link gap and look at what class of site dominates it.
6. Tidy the internal anchors.

Six items, none needing a placement budget. Items two and three usually show an effect on their own, because they recover weight that was already earned.

FAQ

How many referring domains count as normal?

There is no absolute figure; take the benchmark from competitors in your own results. The trend tells you more: a profile gaining a few domains a month is healthier than a static one, even at a lower absolute count.

Do directory links do damage?

On their own, usually not; Google ignores most of it. The tilt does damage: if directories make up nearly the whole profile, that is visible and it means there are no natural links.

Should a new site disavow anything?

Almost certainly not. With no purchase history and no manual action, it is work on an imaginary problem.

Which pays back faster, new links or recovered ones?

Recovery. The link already exists, the site already agreed, and most often what is needed is a redirect rather than a negotiation.

Is the webmaster console enough on its own?

For a fifty-page site, yes. What it lacks is link history and competitor data, which is exactly what lost links and the gap are made of.