Our press page used to carry five headlines under the names of real trade publications. One of them announced that FlowSmartly had raised $32 million. Below the headlines sat a fact sheet, marked cleared for publication, giving a founding year, a headquarters, 120 staff across 22 countries, 25,000+ customers and a $48M Series B.
Every number in that paragraph was invented. So were the headlines, and so were the publications' interest in us.
What was actually on the site
The press page was the worst of it, but it was not alone.
`/company/about` carried a 2019-to-2024 company history and a band of five-figure traction statistics. The company did not do those things in those years.
Leadership showed four colleagues on the about page and two more on the press page, standing beside the one real founder. Six people who do not exist, with portraits and titles.
`/company/customers` — still the largest single concentration — carries eight statistics, three named customer companies with results attached, and three named individuals with photographs and quotes.
The history, the traction band, the invented press coverage and the invented colleagues are gone. Both leadership grids are down to the one person who has been confirmed.
Nobody sat down and decided to lie
This is the part worth saying plainly, because it is the part that generalises.
A design pass needs words in every slot. You are building a press page, so you write what a press page contains: a headline, a publication, a figure. You are building a leadership grid, so you fill the grid. The copy is scaffolding — it exists to prove the layout works, and everyone involved knows it is scaffolding at the moment they write it.
Then the layout ships, and the scaffolding ships with it, and nobody re-reads a page that already looks finished. The gap between "this is placeholder" and "this is a public claim about our company" is one deploy and about four minutes of attention.
The tell is specificity. Nobody invents "some customers" — invention produces 25,000+, 4.9 out of 5, $1.4B influenced. Placeholder copy is confident, because confident copy is what makes a layout look right.
Why a named customer is worse than a big number
An inflated statistic is dishonest and vague. A named company is dishonest and checkable.
Anyone who reads a case study and searches for the customer finds nothing. That is not a small credibility problem discovered by a diligent minority — it is the exact behaviour of the one visitor you most want, the one who is far enough down the funnel to verify a claim before bringing it to their boss.
The same goes double for a person. A stock portrait, a name, a job title and a sentence in quotation marks is a fabricated testimonial. It stays a fabricated testimonial no matter what disclaimer sits under it.
The rule we ended up with
The site already had a convention for uncertain figures, and it works: /status, /solutions/domains, /solutions/video-studio and the footer all mark their numbers illustrative. A product mockup showing a dashboard needs numbers in it, and saying so is honest and sufficient.
What we learned is where that convention stops:
1.
A mockup can be illustrative. It is a drawing of software, and the figures in it are obviously a sample.
2.
A statistic about the business cannot be. "25,000 customers (illustrative)" is not a disclosure, it is a sentence that has given up.
3.
A person can never be. A photograph with a name and a quote is a claim about a human being, and a label underneath does not unmake it.
What is still on the list
The audit is not finished. /company/customers is still standing, and the scattered testimonials on the product pages are still there.
Both are recorded, in the repository, in a file that exists so that anything left undone is a decision somebody made rather than something everybody forgot. The options are written down next to each entry: rebuild the page as an honest early-stage page, take it down until there is a real customer to feature, or label it and accept that labelling does not fix the portraits.
We would rather publish that the list is unfinished than publish a page implying it never existed. A company with no customer numbers yet is an early company. A company with invented ones is a different kind of company entirely, and the difference is worth more than the numbers were.