Pengo asks for one paragraph about who you sell to, and then judges every stand at every fair against it. It is the highest-leverage text box in the product, and the most common way to get thin results is to fill it in the way you would fill in a form.
What the paragraph is used for
Two things. Scoring reads it to decide whether a company on a floor is worth walking to, and quotes the sentence it judged on. Discovery reads it to decide which fairs to go looking for at all. Nothing else sees it: the passes that read a directory, profile a company or list the people are deliberately never shown your description, because what a company sells is the same fact for every customer and a profile written to flatter one buyer is that seller's opinion in a shared table.
What a good one names
- What you sell, in the words your customers use rather than the words on your homepage. "Vibration sensors for rotating machinery" beats "industrial IoT solutions".
- Who buys it, by role and by company shape. Maintenance managers at plants over 200 staff is a filter. Enterprises is not.
- What has to be true of a company for it to be worth your time. Owns its own production line. Runs a fleet. Ships regulated goods. This is the sentence scoring leans on hardest.
- Who you cannot sell to, when it is a real exclusion. Consultancies, resellers and government stands are the three that clog a floor most often.
Two versions of the same company
- Thin
- "We provide AI-powered solutions that help manufacturing companies improve efficiency and reduce costs." Every stand at an industrial fair matches this, so the scores come back clustered in the sixties and the ranking tells you nothing.
- Specific
- "We sell predictive-maintenance software to plant and maintenance managers at manufacturers running their own production lines, typically 200 to 5,000 staff, in DACH. Best fit is a plant with unplanned downtime it already measures. Not a fit: consultancies, system integrators, and companies that outsource production." Now a score has something to disagree with.
What to leave out
Adjectives about yourself do no work. Leading, innovative, trusted and world-class describe your marketing rather than your buyer, and scoring has nothing to check them against. Mission statements are the same problem at greater length.
Leave out your product's feature list too, unless a feature is what qualifies a buyer. Scoring is matching a company to a need, not a spec sheet to a spec sheet.
The region is a separate field, and it matters
Region and vertical are their own settings rather than sentences in the paragraph, because discovery uses them to bound where it searches. Naming a region in prose and leaving the field empty is the most common reason a discovery pass comes back with fairs on the wrong continent.
Let Pengo draft it, then edit it
Settings will read your website and draft the paragraph for you, at a price stated before you press it. Two things about that draft are worth knowing. It states how confident it is, and a low-confidence draft arrives with a sentence naming what your site did not say, which is usually who the buyer is. And it never saves: the paragraph lands in the editor and nothing is written until you press Save.
That is deliberate. One thin read of a landing page must not be able to replace a description you tuned by hand, because you would find out a month of research later.
Editing it makes your scores stale, on purpose
Every score is stamped with the description it was an answer to. Rewrite the paragraph and the old scores are answering a question nobody is asking any more, so the app marks them stale and offers a rescore rather than leaving them looking current. Scores written before the stamp existed are never called stale: they were correct when they were made.
So the order that works is: write the description properly, then buy scoring. Doing it the other way round means paying for the same pass twice, and the second one is the only one you can use.
What each pass costs is on pricing, and the limits scoring works under are in what Pengo will not guess.