Link building tools, by category

By Bill Hartzer · Hartzer Consulting · 25+ years in search, domains and expert witness work

Categories, not a leaderboard

Almost every article about link building tools is a ranked list, and almost every ranked list is out of date within a year and shaped by affiliate arrangements you cannot see. This page does something different: it describes what each category of tool is structurally for, and what it is structurally unable to tell you. Those properties do not go stale.

Five categories cover the work.

• Backlink indexes — independent crawls of the web that record which pages link to which.
• Prospecting tools — finding candidate sites and the people at them.
• Outreach and CRM tools — sending, tracking and remembering conversations.
• Monitoring tools — noticing when links appear, disappear or change.
• Free options — including the one dataset nobody else has.

The order matters. A tool cannot decide which sites you should want a link from, cannot write a reason for someone to say yes, and cannot tell you whether a link was worth having. Those are judgment tasks, and buying a subscription before you have made those judgments is the most common way to spend money on link building without doing any.

Backlink indexes: Majestic, Ahrefs, Semrush, Moz

A backlink index is a crawler and a database. The vendor crawls the web independently of Google, records every link it finds, and exposes that record through an interface and an API. The four widely used ones are Majestic, Ahrefs, Semrush and Moz. I have been a Majestic Brand Ambassador for over fifteen years, which is a disclosure rather than a recommendation — the reasoning below applies to all four.

What an index is for:

• Seeing who links to a site you do not own — your competitors, a prospect, an acquisition target.
• Reading the shape of a profile: how many referring domains, how concentrated, what anchor text distribution, what tier structure.
• Comparing profiles against each other, which is the basis of link gap analysis.
• Historical view: what linked in the past and no longer does.

Each vendor layers its own metrics on top of the raw graph. Majestic supplies Trust Flow (a measure of how trustworthy the sites linking to a page are), Citation Flow (a measure of link volume irrespective of quality), and Topical Trust Flow (the subject categories those trusted links come from). Ahrefs supplies Domain Rating and URL Rating. Moz supplies Domain Authority and Page Authority. Semrush supplies Authority Score. All are vendor models, none is used by any search engine, and none should be treated as a rank prediction.

What an index cannot tell you: whether Google has seen the link, whether Google counts it, whether the linking page has any readers, whether the link drove a single visit, or whether the link will still be there next quarter. It tells you the link exists in that vendor's crawl. That is a narrower claim than most people read it as.

Why the indexes disagree with each other

Run the same domain through all four indexes and you will get four different backlink counts, often differing by multiples. This is not a sign that three of them are broken. It is the expected result of four independent crawls with different rules.

The disagreement has specific causes.

• Different crawl budgets and priorities. No crawler visits the whole web. Each vendor decides where to spend, and those decisions differ, so each index has blind spots the others do not.
• Different recrawl cadence. A link found last week may be in one index and not yet in another. A link removed last week may still be in one index and already gone from another.
• Different definitions of a live link. Some indexes are quicker to drop a link once the source page stops resolving or the link disappears from the markup.
• Fresh versus historic views. Majestic in particular separates a recent index from a long-horizon historic one. Comparing another tool's default view against a historic index is comparing two different questions.
• Different handling of duplicates and redirects. Whether ten links from ten near-identical template pages count as ten links or one is a policy decision, and the policies differ.

The practical rule: pick one index as your system of record for a given site, and stay in it. Switching mid-analysis produces changes in your numbers that have nothing to do with anything that happened on the web. If you must compare across tools, compare referring domains rather than link counts — domain counts are far more stable across vendors than raw link totals.

Prospecting tools

Prospecting tools answer the question who might link to this. In practice this category is less a set of products than a set of techniques, some of which have software wrapped around them.

• Link intersect features inside the major index tools: find domains linking to several competitors but not to you. This is the highest-yield prospecting method available and it lives inside a tool you probably already pay for.
• Search operators. Free, and still the most flexible prospecting instrument there is. Module 4 covers working operator patterns for resource pages, roundups and directories.
• Contact discovery tools that attempt to find and verify an email address for a named person at a domain. These vary enormously in accuracy and none should be trusted without verification; sending to a guessed address that bounces damages your sending reputation.
• SERP scrapers and list builders that turn a query into a spreadsheet.

What prospecting tools cannot tell you is whether a site would ever link to anyone. A domain can look perfect on every metric and have an editorial policy of never linking out. Qualification is manual, and the ratio of raw prospects to qualified ones is brutal — expect most of a scraped list to fail a two-minute human check.

Outreach and CRM tools

This category sends the email and remembers the conversation. At the low end it is a mail-merge with sequencing. At the high end it is a customer relationship manager adapted for publishers instead of buyers.

What they are for: personalized sending at moderate volume, follow-up scheduling, reply detection so a sequence stops when someone answers, team visibility so two people do not pitch the same site, and a record of every prior interaction with a domain.

What they cannot do: make a bad pitch work. And they introduce a specific hazard — the ease of adding another step to a sequence. A tool that will cheerfully send seven messages to a person who never opened the first one is a tool for burning prospects.

A genuine observation from running programs at various sizes: a spreadsheet is adequate for a long time. The threshold where dedicated software starts to pay is not a volume threshold, it is a headcount threshold — the point where more than one person is contacting sites and the risk of collision or duplication becomes real. Below that, the tool mostly adds overhead you will stop maintaining. Module 7 covers the minimum workable tool set.

Monitoring, and the free options that matter

Monitoring is the category most programs skip, and the one with the clearest return. Links are not permanent. On a real twenty-year profile, 49.6% of all links ever recorded were eventually lost, and the median referring domain stopped linking after 1,080 days. If nothing watches for that, a program can run flat for a year while quietly replacing losses.

Monitoring covers three jobs:

• New link alerts from your index of choice, so coverage you did not know about gets noticed.
• Lost link alerts. Worth acting on, because most losses are recoverable — in an analysis of 645,202 lost links, 97.4% disappeared while the source page was still perfectly reachable. The page lived; the link was edited out. Only 4 of those 645,202 were lost to a 404.
• Brand mention alerts for unlinked citations, which matter more than they used to.

On the free side, the most important dataset in link building is free and most people underuse it. Google Search Console reports the links Google itself has recorded for a site you verify ownership of. That is a fundamentally different claim from a third-party crawl: it is the search engine's own view. It is sampled rather than exhaustive, it is slow to update, it gives you no metrics on the linking domains, and it only works for sites you control. Within those limits, nothing else substitutes for it. Bing Webmaster Tools offers a comparable view. Most index vendors also expose a limited free tier, and site: and link-pattern queries in a search engine cost nothing at all.

My honest summary of the whole category: one backlink index, Search Console, a spreadsheet, and an email client will run a competent link building program. Everything beyond that buys speed, not capability.