The boilerplate is the part of the press release nobody thinks about, because it sits at the bottom, and because every company treats it as a set-it-and-forget-it block. That inattention is exactly why it’s the highest-leverage section most companies neglect.
A good boilerplate travels with every release. (For the full writing framework, see how to write a good press release.) When a wire service syndicates the release, the boilerplate is copied. When a reporter uses the release as source material for a story, they often lift the boilerplate verbatim as their company description. When an AI model is trained on news data, the boilerplate phrases become the phrases the model associates with your brand. One paragraph, repeated thousands of times across the web, shapes how you’re described everywhere.
This post is how to write a boilerplate that actually earns that repetition.
What a boilerplate is
The boilerplate is the “About [Company]” block at the bottom of every press release. It’s a standardized company description, usually 60 to 100 words, that explains who you are for readers who don’t already know. It ends with a link to the company website and sometimes social handles.
The structure is basically always the same:
- Company name, year founded, what it does (one sentence).
- Customer count, product traction, or market position (one sentence).
- Investors, notable hires, or credibility markers (one optional sentence).
- Call to action, website link, contact.
Every major company has a boilerplate. Most are terrible. Not because they’re written badly — because they’re written once, years ago, and never updated to reflect what the company actually is anymore.
Why the boilerplate gets reused
Here’s the thing nobody tells you about press releases: the boilerplate gets copied far more often than the headline does.
When a wire service distributes your release, the full text including the boilerplate is syndicated to hundreds of sites. Each of those sites carries a verbatim copy. When a small trade publication writes a short news item about your release, the writer often lifts the boilerplate wholesale as their company description — it saves them two minutes of rewriting and it’s legally safe because you published it yourself.
When a reporter at a larger publication writes a real story, they still often start with your boilerplate as a reference for how to describe the company, and their final description ends up as a rewritten version of your phrasing. The bones of what you wrote in the boilerplate shape what gets published.
Multiply that effect across 20 or 30 releases over a year and you have hundreds of near-identical company descriptions floating around the web, all descending from whatever you put in the boilerplate six months ago. That corpus is what Google’s knowledge graph and AI training data sees when it builds a model of your brand.
The entity effect
The boilerplate is the single most direct lever you have on how entity recognition systems describe your company. (This feeds directly into how AEO works.)
Google’s knowledge graph tries to build a canonical understanding of every entity on the web — for a company, that means what industry it’s in, where it’s based, who founded it, how big it is, what it makes. The graph builds this understanding from repeated mentions across authoritative sources. When 100 different news sites all describe your company with nearly identical phrasing, the graph treats that phrasing as the canonical description.
Language models work similarly. They learn during training that certain phrases reliably co-occur with your brand name, and those phrases become the ones the model uses when generating answers that mention your brand. If your boilerplate consistently describes you as “a payroll platform for small businesses,” the model will describe you as a payroll platform for small businesses. If your boilerplate rambles through three product lines and two market segments, the model will be confused about what you actually do.
The boilerplate is entity engineering. Pick the description you want the knowledge graph and the language models to use for your company, write it into the boilerplate, and stay consistent.
The common mistakes
Three mistakes I see in 90 percent of boilerplates.
Wrong length. Too-long boilerplates (150+ words) don’t get reused because reporters have to trim them. Too-short ones (under 40 words) don’t give the reporter enough material. The sweet spot is 60 to 90 words.
Vague positioning. “Acme is a technology company that helps businesses grow.” That’s not positioning, that’s wallpaper. It won’t get reused because it doesn’t say anything, and when it does get reused, it doesn’t move entity recognition in any particular direction. Write a description that would only fit your company.
Outdated traction markers. “Founded in 2019, Acme serves over 500 customers.” If it’s now 2026 and you have 5,000 customers, that boilerplate is actively lying for you. Update quarterly. At minimum, update whenever the customer count, funding stage, or product description materially changes.
What a good boilerplate looks like
Let me walk through a real example of the structure I recommend.
About Acme Payroll: Founded in 2019, Acme Payroll is a payroll platform built for small businesses with under 50 employees. The company serves over 12,000 small businesses across the United States and has processed over $2 billion in payroll transactions since launch. Acme is backed by investors including Bessemer Venture Partners and Tiger Global. Learn more at acmepayroll.com.
That’s 58 words. Let me break down what each sentence is doing.
Sentence one: establishes the category (“payroll platform”) and the segment (“small businesses with under 50 employees”). These two phrases are doing the entity-engineering work. Every reuse of this boilerplate reinforces “Acme Payroll = payroll platform for small businesses.”
Sentence two: gives the reporter a usable traction number. 12,000 customers is specific and verifiable. $2 billion is impressive but tied to a timeframe that makes it credible. These numbers are what reporters quote when they want to establish the company’s size.
Sentence three: credibility markers. Real investor names add weight without being the focus.
Sentence four: the link, plainly.
Every sentence earns its place. Nothing is filler. The total package tells a reader in 60 seconds what the company is, how big it is, and why it matters.
The update cadence
Your boilerplate should change at least once a quarter. Ideally once a month if you’re growing fast.
What to update:
- Customer count, transaction volume, or usage metric (whichever is most impressive at the time).
- Funding stage if it changed.
- Product description if you launched something meaningful.
- Geographic footprint if you expanded.
What not to update (usually):
- Founding year. It’s founding year. It doesn’t change.
- Core positioning. Consistency matters more than constant refinement. Pick a description and stick with it for at least 12 months.
- Investor list. Only update when you close a new round.
A simple practice: every time you publish a new release, open last release’s boilerplate and check whether any of the numbers need to change. If yes, update them before sending. If no, use as-is.
The consistency trap
One caution. The reason consistency matters is entity engineering — you want the same description repeating across the web so that Google and language models recognize it as canonical. That means you should not aggressively rewrite the boilerplate between releases just because you feel like changing the phrasing.
Small updates to numbers are fine. Full rewrites of the core positioning sentence should happen rarely — once a year, maybe — and only when there’s a real reason. If the position of the company actually changed (new product category, new market focus), the boilerplate should reflect it. If you’re just getting bored of your own phrasing, leave it alone.
The companies with the strongest entity recognition in Google and in AI answers are the ones that picked a description years ago and stuck with it. The ones that rewrite their boilerplate every month are confusing their own signal.
The check before publishing
Before sending any press release, read the boilerplate out loud. Ask yourself three questions.
Is the category clear? If someone who had never heard of you read just this paragraph, would they understand what your company does?
Are the numbers current? Check the customer count, the funding stage, the traction metric. Update if any are stale.
Does it match the other boilerplates on your recent releases? If you’re using radically different phrasing, pick one and align the rest. Consistency is the point.
If the answer to all three is yes, the boilerplate is doing its job. If not, spend ten minutes fixing it before the release goes out. That ten minutes is the highest-leverage editing work on the entire release.
The boilerplate is the part of the press release that keeps working after everyone forgot about the release itself. Treat it that way.