A press release is a written announcement a company issues about something it has done, and press release SEO is the question of whether pushing that announcement through a wire service earns anything in search. This is for founders, growth leads, and first marketing hires who have just closed a round or shipped a launch and now have a distribution package in front of them with "SEO benefit" somewhere in the deck. The honest answer is mostly no, for reasons Google has written down and published. This guide covers that policy wording, what a 2026 study of four million AI citations found about whether answer engines quote wire copy, the one case where paid distribution is genuinely the right tool, and what to do instead when the goal is links or AI visibility.
The short version
- Google's spam policies name press releases specifically. A link in a release distributed on other sites is only compliant when it carries
rel="nofollow"orrel="sponsored", which means a compliant press release link passes no ranking credit by design.- AI answer engines almost never quote the wire copy. In BuzzStream's analysis of four million citations, direct newswire citations made up 0.21% of the entire dataset, and press releases arriving through syndicated news channels were 0.32% of news citations. Original editorial accounted for 81% of news citations.
- The same dataset points the other way for pages a company owns: internal press releases and newsroom content were 2.98% of all citations, and 18.15% of ChatGPT's citations.
- A press release is a source document, not a distribution channel. The asset is the page on your own domain, plus a handful of targeted pitches to people who cover your subject.
- There is one honest use for a wire, and it has nothing to do with marketing: regulatory and financial disclosure, where broad timestamped distribution is the actual requirement.
Do press releases help SEO?
Rarely, and almost never through the mechanism the pitch implies. The distinction the rest of this article rests on is simple: the release itself does very little, while the editorial coverage a genuinely newsworthy announcement can trigger does a great deal.
Those are two different products sold as one. Paying a wire service buys distribution of your own words onto syndication partner sites, where the text appears verbatim under a "press release" label. Earning coverage means a journalist read the news, decided it mattered to their readers, and wrote about it in their own words. The first is a publishing transaction. The second is digital PR, and it is the one that moves authority.
The confusion is understandable, because a wire release does produce things that look like results. Dozens of URLs appear. Your company name shows up on domains with recognizable logos. A pickup report lands in your inbox with a number on it. None of that is coverage, and Google's own documentation explains why none of it is a ranking signal.
What does Google's own policy say about links in press releases?
It names them directly, which makes this question settleable rather than debatable. Google's spam policies for Google web search list, among the examples of link spam:
"Advertorials or native advertising where payment is received for articles that include links that pass ranking credit, or links with optimized anchor text in articles, guest posts, or press releases distributed on other sites."
The same policy then supplies the qualification, and it is the sentence that matters most:
"It's not a violation of our policies to have such links as long as they are qualified with a
rel="nofollow"orrel="sponsored"attribute value to the<a>tag."
Read those two passages together and the press release SEO question largely settles itself. A qualified link is compliant precisely because rel="nofollow" and rel="sponsored" tell Google not to pass ranking credit through it. The link that keeps you inside the policy is the link that cannot help you rank. The link that could theoretically help you rank is the one the policy calls link spam.
Two things this does not mean. Google does not penalize companies for issuing press releases, because the policy is about unqualified paid links rather than about the format. And a qualified link is not worthless in every sense, since a human can still click it. It simply does not carry the ranking value that the phrase "SEO benefit" is meant to describe.
Do press release links count as backlinks?
Technically yes, functionally no. A link in a syndicated release exists in a crawler's view of the web and will usually appear in a backlink tool's export, which is why wire pickup reports can look impressive in a link report.
What those links lack is the thing that makes a backlink useful. Ranking credit flows through links that are not qualified with nofollow or sponsored and that someone chose to place because your page was worth citing. A paid release satisfies neither condition when it is set up correctly. Our guide to link building for SEO covers which acquisition tactics still hold up under current guidance and which quietly turn into liabilities, and paid syndication sits firmly in the second group.
The practical consequence for reporting: if your backlink count jumps by 200 the week a release goes out and then contributes nothing to rankings over the following quarter, that is the expected outcome rather than an anomaly worth investigating.
Do AI answer engines cite press releases?
Almost never, and there is now data rather than opinion on this. BuzzStream analyzed four million citations across 10 industries using 3,600 prompts, collected over one week starting January 27, 2026, on ChatGPT, Google AI Mode, Google AI Overviews and Google Gemini. Their study of how news publications appear in AI citations reports the figures below. Read the denominators carefully, because two different ones are in play.
| What was measured | Share | Share of what |
|---|---|---|
| Articles from news publications | 14.09% | All citations in the dataset |
| True news and editorial content | 81% | News citations, compared against affiliate content |
| Syndicated news pieces | 6.2% | News citations, and 0.9% of the entire dataset |
| Press releases on syndicated news channels | 0.32% | News citations, and 0.04% of the entire dataset |
| Direct newswire citations, such as PRNewswire | 0.21% | The entire analysis |
| Internal press releases and newsroom content | 2.98% | All citations, rising to 18.15% on ChatGPT |
Two readings sit inside that table, and the second is the useful part. News publications are the second-most-visible citation type overall, so the instinct that news coverage feeds AI answers is sound. The wire copy is simply not what gets quoted: original editorial takes 81% of news citations, while the press release version of the same announcement traveling through syndication partners takes 0.32%.
The last row is the one worth pinning to a wall. Content on a company's own newsroom outperformed the syndicated copies of comparable material by a wide margin, and the gap is platform-specific: 18.15% of citations on ChatGPT, against roughly 3% across Google's AI surfaces. Search Engine Journal's coverage of the same research reports the same pattern.
Why the original survives and the copies do not is worth being careful about. BuzzStream's report measures the outcome and does not test the cause, so what follows is reasoning rather than a finding.
A retrieval system looking at 300 near-identical documents has little reason to surface more than one, and the one it surfaces tends to be the version with the strongest signals attached: the original, on the domain that owns the claim. If that holds, flooding the web with identical copies of your announcement is not a neutral act, because it creates competitors for your own text. Our guide to how brands earn AI citations covers the wider set of conditions a page has to clear before an answer engine will quote it.
What is a press release actually good for?
For creating one authoritative, timestamped, quotable version of a fact. That is a real job, and it is worth doing well. The mistake is assuming the wire is how that job gets done.
Based on our research, four situations account for most releases. Here is what the wire buys in each, against what actually produces the outcome you wanted.
| Why you are issuing it | What the wire buys you | What actually buys the outcome |
|---|---|---|
| Funding announcement | Near zero for links. Some entity corroboration if a real outlet picks it up | A pitched exclusive with one reporter who covers your sector, plus a permanent page on your own newsroom |
| Product launch | Near zero. Launch copy is among the least quotable material a wire carries | Editorial pickup, plus a product page built to rank for the problem the product solves |
| Regulatory or financial disclosure | This is the real use case: broad, timestamped, defensible distribution on the record | The wire, used correctly and judged on compliance rather than on links |
| Hiring, awards, milestones | Near zero, and the pickup report will still show a number | Usually nothing. Consider not issuing one, and spending the budget on the asset instead |
Three of those four rows point away from paid distribution. The fourth points straight at it, for a reason that has nothing to do with search.
Where should the release live, your own site or a wire?
Your own domain, on a newsroom page you control, and it is not a close call. The announcement then has a permanent home you can link to, update, add schema to, and cite in a pitch. It also lives on the domain you want an AI answer engine to associate with the claim, which is what the 18.15% ChatGPT figure above describes.
The duplicate content worry that usually follows deserves a plain answer. Syndicated copies of your release are not a penalty. What you have is a consolidation problem, where several near-identical URLs compete to be the version search engines treat as primary. That is a solvable mechanical issue rather than a risk, and our guide to republishing without losing your search traffic covers the canonical handling in full rather than repeating it here.
The sequence that keeps ownership on your side: publish the release on your newsroom first, let it be crawled, then distribute if you have a reason to. Reversing that order hands the first-published version of your own words to somebody else's domain.
Does paid distribution ever make sense?
Yes, in two narrow cases. Both are worth naming, because a flat "never" is as unhelpful as the sales pitch.
- Regulatory and financial disclosure: when a company has a legal or contractual obligation to make information broadly and simultaneously available, a wire is the correct instrument. The timestamp, the breadth, and the audit trail are the product. Links are irrelevant to whether it worked.
- A trade wire your industry's journalists genuinely read: some sectors have a specific service that reporters on that beat actually monitor. That is a media-access purchase rather than an SEO purchase, and it should be judged on whether a named journalist saw it.
One question settles most of these decisions: would you still pay for this if it produced zero backlinks and zero AI citations? If the answer is yes, because of disclosure obligations or because three reporters on your beat read that wire, buy it with clear eyes. If the answer is no, you were buying SEO value, and the policy wording plus the citation data both say it is not on sale.
Wire pricing and packaging vary widely, so treat any figure you read, including one in a vendor's own comparison table, as something to verify before it goes in a budget.
What should you do instead when the goal is links or AI citations?
Give a journalist something worth writing about, then make sure the citable version lives on your domain. Our recommendation is to treat the newsroom page as the asset, targeted pitches as the distribution, and the wire as a compliance tool used only when compliance is the reason.
In practice that means four things, roughly in order of how much they change outcomes.
- Original data: a number nobody else has is the most reliable reason for a reporter to cite you, because a statistic is quotable and the citation has to point somewhere. Even a small internal dataset, honestly described, outperforms an announcement.
- A claim worth quoting: a clear, specific, defensible sentence about your market that a writer can lift intact. Vague positioning language is unquotable by construction.
- A named human available for comment: a real person with a title and a beat, who replies within hours. Journalist request platforms exist precisely because reporters are looking for this and cannot find it fast enough.
- Coverage you can compound: once a mention exists it carries value even without a link, which is a relatively recent change in how this works. Our guide to unlinked brand mentions and why they now matter more explains what to do with the ones you already have.
The full campaign version of this, including how to build an asset that survives a skeptical editor, sits in our digital PR guide. If the work needs a team rather than a process document, our digital PR work is built around earned coverage and measurable authority rather than distribution volume.
How do you measure whether a press release worked?
Not by pickup count, which is the one number the wire reports by default. Pickup measures how many partner sites republished your text automatically, and automatic republication is a contractual outcome rather than an editorial judgment about your news.
| Metric | What it actually tells you |
|---|---|
| Named editorial placements | Whether a journalist chose to write about this. The only signal that a story landed |
| New referring domains from those placements | Whether the coverage produced authority, counted at the domain level rather than the link level |
| Branded search movement | Whether anyone who saw the news went looking for you afterward, visible in Search Console over weeks rather than days |
| Brand mentions, linked and unlinked | How widely the claim now travels, which is the input AI answer engines learn from |
| Wire pickup count | How many partner sites are contractually obliged to republish. Not a result |
Say the limits out loud in the report, because attribution is genuinely weak here. A branded search lift after a funding announcement cannot be cleanly separated from the funding itself being newsworthy. Across the sources we reviewed, nobody has a clean method for isolating a wire's contribution from the news the wire was carrying, and reports that claim one are usually crediting the story to the channel.
Which mistakes make a press release actively worse?
Four, and each turns a neutral spend into a negative one.
- Optimized anchor text in a paid release: this is the exact behavior Google's policy names as link spam. A commercial keyword as anchor text, in an article you paid to place, without
rel="sponsored", is the textbook example rather than a grey area. - Keyword-stuffed boilerplate: the company description at the bottom of a release is where keyword stuffing usually hides. It makes the copy unquotable to a journalist and adds nothing on the search side.
- Publishing only on the wire: if the announcement has no page on your own domain, every citation of your news points somewhere else, and you have donated the canonical version of your own claim.
- Counting syndication pickups as coverage: reporting 180 pickups to a board or an investor as media coverage misrepresents the result, and it sets the expectation that next quarter's number should be higher.
A fifth pattern is worth watching in 2026: generating release copy at volume with an AI writing tool because distribution is cheap. Volume was never the constraint. The constraint is whether a specific person finds the news worth repeating.
The habit worth building is separating the announcement from the channel. Every time news comes up internally, ask two questions in order: where will the permanent, citable version of this live on your own site, and which named journalist would find it interesting. The wire, if it appears at all, comes third, and it answers a compliance question rather than a growth one.
This week, find the last announcement your company made and check whether it has a page on your own domain. If the only version lives on a wire service, publishing it on your newsroom, with a real date and a person's name attached, is a one-hour job that puts the citable copy back where it belongs.
Want to Build Authority Online?
Our Digital PR strategies help businesses earn trusted backlinks, increase brand visibility, and strengthen their online reputation.
Strengthen My Brand AuthorityFrequently Asked Questions
Are press releases good for SEO?
Not directly. Google's spam policies specifically name links in press releases distributed on other sites, and say such links are only compliant when qualified with rel="nofollow" or rel="sponsored", which means they pass no ranking credit by design. What can help is the editorial coverage a genuinely newsworthy announcement earns, because those links and mentions sit on pages a journalist chose to write.
What is a press release in SEO?
In an SEO context, a press release is usually treated as a distribution product: you write an announcement, pay a wire service, and it appears verbatim on syndication partner sites. The more useful way to think about it is as a source document that journalists and analysts can quote, rather than as a channel that earns rankings on its own. The page on your own domain is the asset; the wire copy is a set of duplicates.
Should links in a press release be nofollow?
Yes, when the placement is paid or syndicated. Google's guidance is explicit that advertorial and press-release links are not a policy violation as long as they carry rel="nofollow" or rel="sponsored". Leaving them as plain follow links in paid distribution is the version that risks a link-spam problem, and adding optimized commercial anchor text on top of that is the clearest example Google gives.
Do press releases get cited by ChatGPT or Google AI Overviews?
Rarely. In BuzzStream's analysis of four million citations across 3,600 prompts on ChatGPT, Google AI Mode, AI Overviews and Gemini, direct newswire citations accounted for 0.21% of the whole dataset and press releases arriving through syndicated news channels were 0.32% of news citations, while original editorial made up 81% of news citations. Content on a company's own newsroom did far better, at 18.15% of ChatGPT citations.
Is paid press release distribution worth it for a startup?
For links or AI visibility, generally no. For regulatory and financial disclosure, where broad timestamped distribution is the actual requirement, a wire is the right tool. A trade wire that reporters on your beat genuinely monitor can also justify the spend as media access. A useful test before buying: would you still pay for this if it produced zero backlinks and zero AI citations.
Where should I publish a press release, my own site or a wire service?
Your own domain first, on a newsroom page you control, so the citable version lives with you and the announcement is something you can link to, update and pitch. Wire distribution, if you use it at all, is a secondary copy. Those copies are a consolidation question rather than a penalty risk, so handle them with canonical signals instead of avoiding them out of fear.
GrowthHasten Team
Editorial Team, GrowthHasten
Articles from the GrowthHasten editorial team, grounded in primary research, hands-on client work, and testing across SaaS, AI, and B2B technology, and fact-checked in-house.
View profile
