Scoring an exhibitor list means giving every company in a trade show directory a number between 0 and 100 for how well it fits the buyer you sell to, and a sentence saying why. The input is what each company sells in its own words. The output is a walk route, ordered by fit, that a seller can act on in three days on a floor.
The list itself is free and published by the fair. Scoring is the work that makes it worth having. This page is the method, including the parts you can run in a spreadsheet.
What you need before you can score anything
One thing, and almost every failed attempt at this skips it. You cannot score a company name. Siemens, Krones, Festo and four hundred others tell you nothing about whether they buy what you sell, and a score assigned from a name alone is a score assigned from brand recognition.
So the first pass is not scoring. It is finding out what each company on the list actually sells, in the company's own words, from its own pages. That is roughly two minutes a company by hand. For a 200-stand fair it is six or seven hours, and it is the step that decides whether everything after it means anything.
The five signals worth scoring on
Weighted for a seller walking a floor with three days and about forty walkable stands. The weights are a starting point and the first two carry most of the decision.
| Signal | What you are reading | Weight |
|---|---|---|
| What they sell | The company's own description of its product, from its own site. Does your product sit in the same sentence as theirs? | 35 |
| Who they sell it to | Their customers, named or described. You are looking for your buyer appearing as their buyer, or as them. | 25 |
| Size and shape | Employee band, and whether they are a manufacturer, a distributor, an integrator or a consultancy. A ten-person reseller and a four-thousand-person plant are different sales. | 15 |
| Why they are at this fair | A stand at a packaging fair says something a company website does not. Exhibiting is spend, and spend is intent. | 15 |
| Reachability on the day | Hall, stand number and whether the stand is staffed by people who can talk about buying. A perfect fit in a hall you will not reach is a call, not a walk. | 10 |
The two signals that look useful and are not
- Stand size. It is tempting because it is on the floor plan and it is a number. It measures marketing budget, not fit. The largest stands at most fairs belong to companies that sell to the same market you do, which makes them competitors rather than buyers.
- Brand recognition. A ranking that puts the names you know at the top has told you what you already knew. The value of a directory is the eleven companies in it you have never heard of, and a recognition-weighted score buries exactly those.
Where to draw the cuts
A 0 to 100 score is useless without stated cuts, because the reader has to know what 61 means before they can act on it. Four bands is enough, and three would be too few: a stretch and a non-buyer need to be separable or the bottom of the list is unreadable.
- 75 and above
- This is the buyer you described. What they sell, who they sell it to and their size all point the same way. These are the stands the trip is for, and there are usually fewer than you expect.
- 50 to 74
- A plausible buyer with one real gap. Right industry, wrong size. Right size, adjacent product. Worth a walk if the hall is on your route, worth an email either way.
- 25 to 49
- A stretch. Something connects and the rest does not. This band is where most of a directory lands, and it is the band a good rubric is judged on: if half your fair is scoring 70, the rubric is flattering you.
- Below 25
- Not a buyer. A supplier to your market, a competitor, or something your exclusions rule out. Score it low and say which, because a competitor at a fair is worth knowing about for a different reason.
What a score has to show before you trust it
A bare number is not a finding. Three things travel with it or the score is unusable, whether a person or a model produced it.
- 1The reason, in a sentence, naming the specific thing that decided it. Not "strong fit" or "highly relevant", which restate the number and tell you nothing. A rationale you can disagree with is a rationale you can overrule in four seconds.
- 2The page it was read on. A claim about what a company sells is checkable or it is a guess, and the difference matters most for the rows you are about to spend a morning on.
- 3An admission when there was nothing to go on. A company whose site says nothing useful must come back unscored rather than scored 50. A placeholder in the middle of the range is the one output that is worse than a gap, because it is invisible.
That third one is where automated scoring usually goes wrong, and it is measurable. In one pass over fifty companies here, the model returned a judgment on 38 and omitted the other 12 because their pages did not support one. A system that had scored all fifty would have looked more complete and been less true.
Three ways to run this
| By hand | Spreadsheet plus an enrichment tool | Pengo | |
|---|---|---|---|
| Finding what each company sells | Two minutes a company, so six or seven hours | Firmographics arrive quickly. What they sell, in their own words, usually does not | A profile for every company on the list, from the company's own pages |
| Applying the rubric | A judgment call per row, and your judgment drifts by row 120 | A formula over the columns you managed to fill | One pass over every booth, same rubric on every row |
| The reason behind a score | In your head, and gone by the following week | Whatever the formula was, which is not a reason | A sentence per stand, with the page it was read on |
| Rows with nothing to go on | You guess, or you leave a gap you later forget was a gap | They come back with blank columns and score as zero | Returned unscored and counted, so the gap is visible |
| What it costs | Most of a working day | A seat, monthly | 250 credits a fair, about $1 |
The Pengo column is one press: every booth judged against a description of who you sell to, with the sentence it was decided on. It needs the exhibitor list and the company profiles underneath it, which is the fair unlock, priced by the size of the fair and quoted before you press.
Writing the description the score is judged against
Whether you score by hand or buy it, the quality of the output is decided almost entirely by how well you described your buyer. "We sell software to manufacturers" produces a ranking in which two thirds of an industrial fair scores 60. The descriptions that work name the product, the buyer's job title, the size of company, and at least one kind of company you explicitly do not sell to.
Exclusions do more work than people expect. A rule saying you never sell to systems integrators removes forty rows from a 200-stand fair and removes them for a stated reason, which is worth more than nudging them down a band.
There is a page on writing one: how to write an ICP description that scores well.
Questions
How do you score an exhibitor directory for sales?
Find out what each company sells from its own pages, then score five signals: what they sell (35), who they sell to (25), their size and shape (15), why they are at this fair (15) and how reachable the stand is (10). Cut at 75, 50 and 25, and require a reason with every score.
What should I not score exhibitors on?
Stand size and brand recognition. Stand size measures marketing budget rather than fit, and the biggest stands often belong to competitors. Recognition-weighted scoring ranks the names you already knew at the top and buries the companies a directory exists to show you.
Can I score an exhibitor list with just company names?
No. A company name carries no information about what the company sells, so any score built on names alone is a ranking of brand familiarity. You need each company's own description of its product first, which is about two minutes of work per company by hand.
How long does it take to score a 200-stand exhibitor list by hand?
About six or seven hours, almost all of it spent finding out what each company sells rather than applying the rubric. The scoring itself is fast; the research underneath it is not.
What does a good scoring rationale look like?
One sentence naming the specific thing that decided the score, plus the page it was read on. "Strong fit" and "highly relevant" restate the number and cannot be disagreed with. "Sells filling lines to beverage plants, which is the buyer described" can be checked and overruled.
Should companies with no usable information be scored?
No. Return them unscored and count them. A placeholder score in the middle of the range is indistinguishable from a real judgment, which makes it worse than a visible gap.
What does it cost to have Pengo score a fair?
250 credits, about $1, for every booth at the fair judged with its rationale. It needs the exhibitor list and company profiles underneath it, which is priced by the size of the fair: 1,500 credits for up to 100 stands and 2,500 up to 250.