B2B link building when nobody wants to write about your software

I spent nine years as a trade journalist before this job, which mostly means I know exactly why your outreach email gets deleted. It is not the subject line. It is that you sent a press release about a product update to someone whose job is to write about their readers’ problems.

Nobody wants to write about your software. They will happily write about a number you have that nobody else has.

The category problem

Consumer brands can pitch anything. A shoe company can run a survey about commuting habits and get national coverage. You sell revenue forecasting software to CFOs at mid-market manufacturers, and the pool of publications that cares is maybe forty titles, half of which are newsletters run by one person.

That is genuinely harder. It is also more tractable than it looks, because those forty titles are desperate for content and get pitched far less than TechCrunch. A relationship with the editor of a mid-tier trade publication is worth more to you than a national placement, both for links and for the fact that your buyers read it.

The mistake is applying consumer digital PR tactics to a B2B category and concluding that link building does not work. What does not work is a survey about employee wellbeing pitched to publications that cover industrial automation.

What actually gets picked up

In rough order of hit rate, from about four years of running this for B2B software companies:

Original benchmark data from your own product is the strongest asset by a distance. If you process invoices, you know average approval times. If you run applicant tracking, you know time-to-hire by sector. Publish that quarterly, segmented usefully, and trade press will cite it because there is no other source. One client’s quarterly benchmark has picked up 180 referring domains over two years and now gets cited without any outreach at all.

Surveys of your customer base come second. Weaker than product data because anyone can run a survey, stronger than nothing because you have access to a specific population. Five hundred responses from actual practitioners in a niche beats a thousand from a general panel.

Expert commentary is third and the most underrated. Journalists need a quote from someone credible, on deadline, about something in the news. If your CTO can turn around 150 usable words in two hours, you will get placements. The tools for this have a bad reputation because most responses are useless marketing copy. Answer the actual question, in plain language, with a specific opinion, and you will land a meaningful share.

Genuinely useful free tools are fourth. A calculator that does something tedious. These earn links slowly and forever.

Guest contributions are fifth and mostly a waste of time unless the publication has real editorial standards, in which case they are excellent. If they will publish anything you send with a bio link, so will they for everyone else, and the link is worth roughly nothing.

What we do not do, and why

We do not buy placements. This costs us business, because there is always someone quoting a lower price per link.

The reasoning is straightforward. A site that sells links sells them to everyone, including a large number of businesses you would not want to be adjacent to. Its outbound link profile makes it easy to identify at scale. When the link graph gets recalibrated — which happens on a cycle of roughly eighteen months to three years — those links stop counting, and sometimes they count against you.

I have watched three clients arrive with 200 to 400 paid links from a previous agency and a domain that would not rank. Two of them we cleaned up. One never fully recovered and rebuilt on a new domain.

There is a middle ground people ask about: paying a publication a “contributor fee” or a “processing charge.” I have no clean rule for this. Some legitimate trade publications charge for sponsored content and mark it as such, which is fine and also usually nofollowed. If the fee buys a dofollow link in something presented as editorial, it is a paid link with a euphemism attached.

The unglamorous half

Before any campaign work, two things reliably produce links for almost no cost.

Unlinked mentions. Companies mention you already — in customer blog posts, in roundups, in conference agendas, in job listings. A surprising share forget to link. A monthly check and a polite email converts maybe a third of these. On a client with 4,000 brand mentions, we recovered 61 links in the first quarter this way.

Broken backlinks. Links pointing at URLs on your site that no longer exist, usually from a migration nobody redirected properly. These are already earned; you just lost them. Fixing this on one client restored 140 referring domains in a fortnight.

Neither is interesting work. Both should be done before you spend money on a data study.

Anchors, velocity, and the things people worry about too much

Anchor text distribution matters, but the natural distribution for an editorially-earned link profile is heavily branded and naked-URL, with exact-match commercial anchors as a small minority. If you are earning links rather than placing them, this happens by itself and you can stop thinking about it. If you are having to manage the ratio, that tells you something about how the links were acquired.

Velocity matters less than people think. A data study landing thirty links in ten days is a normal pattern for a real story and looks nothing like a link scheme. What looks unnatural is steady, identical volume from unrelated sites every month forever.

Domain rating is a useful sorting heuristic and a terrible target. A DR 30 publication read by 8,000 people in your exact category is worth more than a DR 80 general business site whose relevant traffic to you is zero. We report both, and we push clients toward relevance when the two conflict.

Budget, honestly

For a competitive B2B software category, expect $2,000 to $8,000 a month in link acquisition budget on top of agency fees, and expect that to buy nine to fourteen editorial placements. That works out somewhere between $400 and $800 per genuinely earned link once you count research, design, and outreach labour.

If someone quotes you $150 a link, you know what you are buying.

The number that matters is not the cost per link but whether the gap between you and the sites outranking you is closing. We track referring domain counts for the client and their top three competitors every month, on one chart. If the gap is not closing, the budget is wrong or the assets are.