Why most outreach fails

Lesson 22 of 50Module 5 · Outreach that works7 min read

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

The honest diagnosis

Most link outreach asks a stranger to do unpaid work for the benefit of someone they have never heard of, and gives them no reason to care. That is the whole diagnosis. Everything else — the subject line testing, the send-time optimization, the personalization tokens — is decoration on a request that had no foundation.

Read a typical outreach email from the recipient's chair. Someone you do not know says they enjoyed your article, which they have not read, and suggests you add a link to a page that exists to sell something. There is no benefit to you, no cost to them, and no consequence either way. The rational response is to delete it, and most people do.

Now read a message that works. Someone identifies a specific broken link on a page you maintain, tells you where it points now, and mentions they have a replacement if you want one. Or: a journalist covering a beat receives a dataset nobody else has, on the subject they wrote about last week. In both cases the recipient gets something before they are asked for anything.

The gap between those two emails is not tone. It is whether the sender did any work before writing. I have watched a great deal of outreach fail, and in almost every case the failure was decided at the list-building stage — Module 4 — and merely revealed at the sending stage.

What the inbox looks like from the other side

If you write anything visible on a subject with commercial value, you receive outreach constantly. I do, and it is instructive.

The volume is high enough that recognition happens in under two seconds, before any reading. The patterns are consistent: an opening compliment with no specific content, a paragraph about a resource that would be a great fit for your readers, a request framed as a suggestion, and a close that promises to share it with their audience. It arrives in batches, often several near-identical messages from different senders in the same week, because the same list is circulating.

Two consequences follow, and neither is obvious from the sending side.

The bar is set by the batch, not by your message. Your email is not judged against nothing. It is judged against the forty similar ones the recipient has already deleted this month, which means being slightly better than average is functionally identical to being average.

The recipient is pattern-matching, not evaluating. They are not weighing your argument. They are deciding, in a moment, which pile this belongs in. Anything that matches the template pattern goes to the delete pile before the argument is read at all — which is why a good argument in a template-shaped email still fails.

How people detect a template

Recipients are far better at this than senders assume. These are the tells, roughly in order of how quickly they give the game away.

• The unspecific compliment. I really enjoyed your article on [topic]. Great post! I loved your piece about link building. If the praise could be pasted onto any article, it identifies itself as pasted.
• Visible merge-field damage. Broken capitalization, a domain where a brand name should be, Hi there where a name failed, a first name that is actually a full email address. One of these ends the message.
• The wrong-context reference. Complimenting a post the recipient did not write, or referring to your blog when the site is a university library guide.
• Stale references. Praising an article from eight years ago as recent work. This is what a scraped list looks like from the receiving end.
• The reciprocal offer. An offer to share on our social channels, from an account with no audience, is a tell because it appears in every template and is never worth anything.
• The manufactured follow-up. Just bumping this to the top of your inbox, or Did you see my last email? — sent automatically, to someone who saw it and chose not to reply.
• Tone drift. Three sentences of casual friendliness followed by a paragraph of marketing copy about a solution. Two different people wrote that email and both were the same automation.

None of these are fixable by better templating. They are symptoms of a message written without knowing anything about the recipient, and the fix is to know something.

Volume as a substitute for relevance

The standard response to poor results is more sending. It is worth being precise about why that makes things worse rather than better.

Start with the arithmetic people have in their heads. If one in a hundred converts, then a thousand emails produce ten links, and ten thousand produce a hundred. The model assumes the conversion rate is a property of the market. It is not. It is a property of the fit between message and recipient, and that fit degrades as you scale, because scaling means loosening the list and generalizing the message. In practice, moving from a hundred well-researched emails to a thousand generic ones frequently produces fewer links in absolute terms, not merely a lower rate.

Then there are the costs that do not appear in the model at all.

• Deliverability. Spam complaints and low engagement teach mailbox providers to filter you. Once that happens the following campaign underperforms for reasons nobody in the room can see, and the usual response is to send more, which deepens it. Lesson 5.4 covers the mechanism.
• Prospect burn. A prospect who deletes your email unread is not neutral. They now associate your domain with the delete pile, and a genuinely good pitch from you in six months starts from behind.
• Reputational cost to the brand. Outreach is published constantly — screenshotted, posted, mocked. It is one of the few marketing activities where being bad at it becomes public.
• Judgment decay. A team getting no replies stops believing replies are possible, and starts optimizing the wrong variables. I have seen months spent on subject lines for a campaign whose list was the problem.

The three failure modes worth fixing

When a campaign is not working, the cause is almost always one of three things, and they need different fixes.

Wrong person. The message reached a shared inbox, a writer who left, or someone with no authority over the page. Symptom: near-total silence, no bounces, no refusals. Fix: go back to Lesson 4.4. This is the most common and the easiest to correct.

Wrong ask. The message reached the right person but requested something they cannot or will not do — add a link to a commercial page, rewrite an article, link to a homepage. Symptom: polite refusals, or replies asking what exactly you want. Fix: make the ask smaller, more specific, and actionable in two minutes.

No reason. The right person received a doable request and had no motivation to act. Symptom: opens with no replies. Fix: this is the hard one, and it is often not an email problem at all. Either you find the reason — their page is broken, their list is incomplete, their article is out of date, you have data they need — or you accept there is not one and stop.

A fourth possibility deserves naming because people mistake it for the others: the mail did not arrive. If a campaign produces literally nothing — no replies, no refusals, no bounces, no unsubscribes — suspect delivery before you rewrite anything.

What working outreach actually looks like

I am not going to give you a benchmark response rate, because the honest position is that published figures vary so wildly, and are so dependent on list quality, tactic and market, that quoting one would be inventing a number. What I can describe is the shape of a campaign that is working.

You get refusals. Real ones, with reasons — we do not link to commercial sites, this list has not been updated in years, send it again when the study is published. Refusals mean the message was read by a human who understood the ask, which is most of the battle, and refusals with reasons are how a prospect list gets better.

You get replies that are not yes or no. Questions, counter-offers, requests to be added to a media list, an invitation to send something later. These are worth more than links and almost nobody tracks them.

You get a small number of links from people who were going to be receptive anyway, and a smaller number from people who were not but were persuaded by something specific.

And you get one thing that only shows up months later: inbound. Someone you pitched in March mentions you in September because they remembered the exchange. That is what a relationship-shaped campaign produces and a volume-shaped campaign never does, which is why Lesson 5.5 is about recording relationships rather than links.