Short answer: a wire pickup number counts the sites that automatically republished your release because they subscribe to the distribution feed. It is a measure of how far the network reaches, not of whether anyone chose to cover you. Earned coverage means a journalist read the story and wrote something. One line in a report is a distribution statistic; the other is an outcome. Most reports print them as the same number.
If you have ever received a report saying a release achieved "412 pickups" and wondered what that entitles you to, this is the answer. It entitles you to very little, and the number is not dishonest — it is just measuring something other than what the reader assumes.
What a pickup actually is
Distribution networks have subscriber sites: aggregators, regional news portals, industry feeds, financial data sites. Many of them republish incoming releases automatically, in whole or in part, with no human involvement.
A pickup is one of those republications. It means the network reached that site and the site's system accepted the feed. It does not mean an editor saw it, that a reader saw it, or that anyone made a judgement about whether the story mattered.
This is why pickup counts are reliably large and reliably flat. They are a property of the network you bought, not of the news you had. Two releases from the same account — one genuinely newsworthy, one a routine staff announcement — will produce roughly the same pickup count. If a metric cannot distinguish between your best story and your worst, it is not measuring the story.
The three tiers hiding in one number
Almost every distribution report merges three very different things.
| Tier | What happened | Human involved? | Worth reporting as |
|---|---|---|---|
| Syndicated mirror | The feed republished your release verbatim, often on a subdomain or an archive page | No | Reach, at most |
| Republished with edit | A site took the release and trimmed, rewrote the headline, or added context | Minimally | Coverage, with a caveat |
| Earned editorial | A journalist wrote their own piece, quoted you, or interviewed someone | Yes | The outcome |
The third tier is what a client is paying for and is usually the smallest number in the report — sometimes zero — while the first tier dominates the headline figure.
Splitting these is not a technical challenge. It is a reporting decision that most agencies avoid because the honest version of the number is much smaller and needs explaining.
How to tell them apart in five minutes
You do not need tooling for this. Take the pickup list and sample fifteen of them.
Is the headline identical to your release? If every result carries your exact headline, you are looking at syndication. A journalist almost always rewrites the headline, because their publication has a house style and their editor wants their own framing.
Is there a byline? A named reporter is the single clearest signal. "Staff", "Newsdesk", "Press Release" or no byline at all points to automated republication.
Is your boilerplate still attached? Nobody writing their own article keeps your About block at the bottom.
Does the page carry advertising and an audience? Many mirror domains are archives with no readership. A quick look at whether the site has current, original content on its front page tells you most of what you need.
Is the link followed? Syndicated release links are usually nofollow or sponsored, which is consistent with Google's spam policies treating optimised links in "press releases distributed on other sites" as link spam. Earned editorial links generally are not marked that way, because they were an editorial choice.
Fifteen samples will tell you the approximate ratio across the whole list. In our experience the earned tier is usually in the low single digits as a percentage, and frequently zero.
Why this distinction is getting sharper, not softer
Two things are pulling in the same direction.
Duplicate content is a retrieval problem. Four hundred near-identical copies of the same text is a deduplication task for any system that has to choose a source. It does not read as four hundred endorsements. If anything, a heavily syndicated release makes it harder for a retrieval system to identify which copy is canonical — which is why the version on your own domain matters more than the mirrors.
Answer engines select sources, and selection favours the distinguishable. A verbatim mirror adds nothing an engine cannot already get from the original. An independently written article, with a named journalist and original framing, is a genuinely separate source that corroborates the story. One of those helps you become the answer. The other is noise you paid for.
It is worth noting how ordinary the requirements are here. Google states that to appear in AI Overviews or AI Mode a page simply needs to be "indexed and eligible to be shown in Google Search with a snippet", and that there are "no additional requirements... nor other special optimizations necessary" (Google Search Central). Nothing about that favours a syndicated duplicate. Meanwhile the controlled research in GEO: Generative Engine Optimization (Aggarwal et al., KDD 2024) found that what improves a source's visibility is added statistics, cited sources and quotations — all properties of an original article, and none of them created by copying one four hundred times.
The gap between the tiers therefore widens over time. Reports that merge them will get less accurate, not more.
Three report patterns worth questioning
Some presentations of pickup data are not just imprecise, they actively mislead. These three come up often enough to name.
"Potential audience" or "estimated reach"
A figure in the tens or hundreds of millions, produced by summing the total monthly audience of every site that carried the release. It treats a person who visited a site's homepage once as having been reached by your announcement on an archive page they never opened. The number is arithmetic, not measurement, and it should never appear next to real metrics where a reader might average them together.
"Featured on" logo walls
A row of recognisable news logos implying editorial endorsement, assembled from syndicated mirrors on those networks' feed subdomains. If the logos come from automated republication rather than from a journalist choosing to cover you, presenting them as coverage is a claim your client may eventually have to defend to someone who checks.
Screenshots without URLs
A screenshot proves a page existed at a moment. A URL proves it exists, lets the reader see the byline, and lets them check whether the page is still live. Reports that show screenshots but withhold links are usually hiding tier one.
The test for all three is the same: can the reader click through and verify it themselves in thirty seconds? If not, ask why.
How to report it honestly
Three lines, in this order, and never blended:
- Earned editorial coverage. Named publication, named journalist, link, date. If there were three, say three. A short list of real coverage is more persuasive to a serious client than a four-figure pickup count, because they can click it.
- Republished with edit. A count, with a note on what the tier means.
- Syndicated reach. The pickup number, labelled as network reach rather than coverage.
Then, separately, whatever you can establish about audience and outcome: referral sessions from those articles, any inbound enquiries that referenced the story, and — increasingly — whether the story shows up when you ask an assistant about the topic. That last one needs its own method, which we cover in measuring AI citations from PR.
The conversation this creates
Reporting it this way is harder in the short term. A client who has been shown "412 pickups" every quarter will notice when the number becomes "3 earned articles, 11 edited republications, 398 syndicated mirrors."
That is the right conversation to have, and it usually goes better than people expect, for one reason: the honest number is actionable and the vanity number is not. Nobody can improve a pickup count except by buying a bigger network, which is a procurement decision rather than a communications one. Everyone can improve earned coverage, by choosing better stories, building a media list for the beat, and pitching the right people at the right time.
It also protects you. A pickup-count report sets an expectation that quietly inflates every quarter until someone senior asks what any of it produced, and at that point the whole programme is on trial. Reporting the real number from the start means you never have to walk one back.
For the wider question of whether releases still earn anything at all, see do press releases still help SEO in 2026.
We help PR agencies, communications teams and distribution platforms rebuild reporting so it survives scrutiny, usually alongside search and answer-engine work and PR and brand building. If your coverage reports lead with a pickup count, get in touch.
