How to Measure Press Release Results Beyond Counting Links

Measure a press release on four layers: search visibility, business outcomes, earned authority and AI visibility. The number of syndicated copies tells you where the release ran, not what it achieved. Record a baseline before you publish, check the same figures at fixed points afterwards, and keep it in one simple sheet.

Why is the link count the wrong measure?

The old way of judging a release was to count the places it appeared. That number mostly measures the size of the distribution network. Syndicated copies are the same text on many sites, their links are nofollow, and a list of 400 copies can sit next to no change at all in enquiries or searches.

The count still has a use. It is the record of where the release ran, and the proof report gives it to you as verified links with dates. But it is the starting line, not the result. One relevant industry article, or one useful source page that people keep finding, can be worth more than hundreds of copies.

What should you measure?

Search visibility

Use Google Search Console, and Bing Webmaster Tools if you have it. Track:

• impressions and clicks for your site;
• average position for the queries you care about;
• branded queries (people searching your company or product name) against non-branded ones;
• whether the source page the release points to is indexed;
• which pages gain visibility after the release; and
• queries connected with the topic of the announcement.

Watch branded search closely: people who come across the news and want to know more usually look you up by name.

Business outcomes

This is the layer that pays for the release. Track what your business already counts:

• enquiries and qualified leads;
• sales and sign-ups;
• calls and bookings;
• demo requests; and
• conversions where the release or its source page was one step along the way.

Add a "How did you hear about us?" field to your contact form if you do not have one. It catches the people who read the news, then typed your name into a search box days later, which analytics alone will miss.

Earned authority

These are the things other people chose to do because of your news:

• independent articles about the announcement;
• relevant links from those articles;
• citations of your data or report;
• podcast or interview invitations;
• conference references and partner mentions; and
• journalist enquiries, even ones that did not become a story.

Keep these separate from the syndicated copies. The syndication and earned coverage post explains why they carry different weight.

AI visibility

Where you can measure it, track:

• referral traffic from AI assistants such as ChatGPT, as shown in your analytics;
• AI citation counts and the pages cited, which Bing Webmaster Tools reports through its AI performance report;
• the grounding-query themes that report shows;
• recurring mentions of your brand in AI answers; and
• which customer questions cause your pages to surface.

AI citations are not the same as rankings, and a citation does not automatically mean traffic. It is another visibility layer, and the measurement tools for it are still limited. The post on how AI assistants find and cite companies covers what makes a page likely to be retrieved at all.

When should you measure?

Pick fixed points and stick to them, so each reading is comparable with the last.

Search and AI systems take time to crawl, index and use new pages, so a flat first reading is normal. Look at the three-month figures before deciding a release did nothing.

What does a simple tracking sheet look like?

Copy this into a spreadsheet. One sheet per release is enough.

The last row is there on purpose, at the bottom. It shows the release ran. The rows above it show whether it did anything.

To read the sheet, compare each release against its own baseline, not against another company's figures. Suppose Parcel & Pine, a fictional furniture maker in Leeds, publishes a report on how long its customers wait for delivery. If branded impressions and "heard about us" mentions rise in the fortnight afterwards while nothing else about the business changed, the release is a reasonable explanation. If a trade magazine then quotes the report, that goes in the earned row, and it is likely the most valuable line on the sheet.

A release is one input among many, so be careful with cause and effect. A seasonal peak, a price change or a sales campaign in the same month can move the same numbers.

Frequently asked questions

How long does it take to see results from a press release?

Referral traffic and journalist enquiries, if they come, usually come in the first days. Search and AI effects take longer because pages have to be crawled and indexed first, which is why the sheet runs to three months.

Can I measure AI citations directly?

Partly. Bing Webmaster Tools reports citation activity across the AI experiences it supports, and your analytics may show referrals from assistants. Most other AI products do not report citations to site owners, so treat any figure as incomplete.

What if nothing moved?

Check the basics first: is the source page indexed, does the release lead with real news, and did anyone who covers the subject hear about it directly? Then look at the first-party source page post, because a release that points at a thin page has little for search or AI systems to use.