To score a discovery call before it happens, rate the booker's answers from 0 to 10 on three or four criteria, weight each criterion by how much it predicts a deal for you, and take the weighted average. A booking that scores 7 or more gets your full preparation, one between 5 and 7 gets a hard qualifying question early in the call, and one under 5 gets a shorter slot or a different person. Then check the scores against what actually happened, because a rubric nobody tests is a guess with a decimal point.
Most writing on discovery call scoring is about grading the call afterwards: how well the rep asked questions, whether next steps were agreed. That is coaching. This is triage, done the evening before: deciding which of tomorrow's four calls deserves an hour of preparation and which deserves ten minutes.
Pick three or four criteria, not ten
Each criterion should be something the booker can tell you on a form, and something that has predicted deals in your own history. For most B2B services and software, four cover it.
- Fit. Whether they are the kind of company you serve well: size, industry, the systems they run. Asked as a band or a pick-one list.
- Budget. Whether they can pay for what you sell. Asked as a band with the lowest option below your floor, never as an empty box.
- Timing. The reason they are looking now. A trigger with a date attached scores high, and just exploring scores low.
- Authority. Whether the person booking signs, or is gathering options for somebody who does.
A fifth criterion is tempting and usually a mistake. Every criterion you add dilutes the ones that matter, and every question you add to the form loses some of the people answering it. The three questions a booking form should ask covers the wording that gets honest answers.
Turn each answer into a number
Give every answer on a pick-one list its own value from 0 to 10, decided once and in advance. A timing question might be written like this, one option per line with its value after the equals sign:
Something broke and we need help now = 10
Starting a project this quarter = 7
Planning for next year = 3
Just exploring = 1The values are the rubric. Writing them down before any booking arrives stops you scoring somebody higher because their email was friendly.
Free text is harder. An answer to what do you need help with? tells you a lot, but only a person or a language model can read it. If nobody will read it before the call, score what you can check: whether it was answered at all, and whether it contains specifics such as a figure or a date.
Weight the criteria by what predicts a deal
Weights say how much each criterion counts. Write them as relative numbers rather than percentages: 3, 2, 2 and 1 means budget counts three times as much as the last criterion, and nothing has to add up to 100. The score is the weighted average: multiply each value by its weight, add the results, and divide by the total of the weights.
Set the weights from your own lost deals rather than from a framework. If most of last quarter's losses died on timing, timing deserves more weight than authority, whatever the acronym on the sales wall says.
A worked example: three bookings, one rubric
Booking A picked the second-highest budget band (6), has a project starting this quarter (7), signs the contracts (10) and wrote three specific sentences about the problem (8). Weighted: 6 × 3 + 7 × 2 + 10 × 2 + 8 × 1 = 18 + 14 + 20 + 8 = 60. Divided by 8, that is 7.5.
Booking B picked the top budget band (9), is just exploring (1), is collecting options for a director (4) and wrote four words about the problem (1). 27 + 2 + 8 + 1 = 38. Divided by 8: 4.8, rounded from 4.75.
Booking C picked the lowest band above the floor (3), needs help now (10), signs (10) and left the problem question blank (0). 9 + 20 + 20 + 0 = 49. Divided by 8: 6.1, rounded from 6.125.
The biggest budget scored lowest. The rubric says a large budget with no timeline and no authority is a research call, and that is the point of writing weights down: they tell you things your instinct would talk you out of on the morning of the call.
What to do with each score
A score is only useful if it changes what you do. Setupp sorts scores into five bands, and the line under each is a sensible default for any rubric:
- 8.5 and above, Exceptional. Prioritise it. Prepare properly and bring your best.
- 7 to 8.4, Strong. A solid fit, worth the full slot.
- 5 to 6.9, Moderate. Qualify further early in the call before committing to scope.
- 3 to 4.9, Weak. Consider a shorter call, or handing it to somebody else.
- Under 3, Poor. A low fit. Offer something asynchronous, or decline.
So in the example, A is Strong and gets the full hour of preparation. C is Moderate: open with the budget question before you commit to scope. B, the biggest budget of the three, is Weak, and gets a fifteen-minute call or a colleague.
Some answers deserve a penalty rather than a criterion of their own: a contract with a competitor that runs another year, a channel you do not work in. Setupp calls these flags. Each flag takes 1.5 points off, and at most three count, so the largest possible penalty is 4.5. Booking A with one flag would fall from 7.5 to 6.0, from Strong to Moderate: still a call worth taking, with one question you now know to ask first.
Check that the score predicts anything
A rubric is a prediction, and a prediction can be checked. After each call, record what happened: did they turn up, and did the call produce a next step you would have written down anyway. Then compare the bookings that scored 7 or above with the rest.
Wait for enough outcomes before reading the result. With three calls, one no-show moves a rate by a third, which is noise. Setupp's calibration panel shows no headline attendance rate until eight outcomes are recorded, and leaves any band blank until it holds four. Those thresholds are not statistical significance. They are the point where one surprise stops dominating the number.
If high scores and low scores turn up and convert at about the same rate, your criteria describe something that does not predict anything for you. Change the weights or the criteria and check again in a month. Auditing last month's calendar applies the same test to your own instinct, which deserves it just as much.
Score a booking to decide how to prepare for it, never to decide whether to turn up.
Four mistakes that make a score useless
- Showing it to the booker. A score is an internal triage note. Nobody should find out they are a 4.8.
- Weights that say everything matters. If every criterion is weighted 3, nothing is weighted at all.
- Too many criteria. Ten criteria at equal weight pull nearly every booking towards the middle of the scale, where the score stops telling calls apart.
- Never revisiting it. Your offer changes and so does your market. A rubric written in March describes a different business by the autumn.
Where Setupp fits
Setupp scores every booking from 0 to 10 on every plan, Free included, and writes the reason in plain words next to the number. When a booking scores under 7, the reason names the single biggest risk to raise early in the call. From Pro, at $19 a month for the workspace, you write the criteria and weights yourself, one per line, with a description after the bar:
Budget fit = 3 | Spending $15,000 or more a month
Timing = 2 | Wants to start within 30 days
Authority = 2 | Signs off the budget
Problem clarity = 1 | Describes the problem with specificsEach criterion is matched to the question that answers it. Without a language model configured, the values come from the weights on your form's options. With one, the model reads the answers against your criteria and writes the reasoning, and if it is unavailable the rules produce the score instead and the booking shows which one ran. Flags take off 1.5 each, at most three. The booker never sees any of it: the score goes to your dashboard and the host's email, and from the Team plan into a meeting document for each booking.
Once eight outcomes are recorded, the calibration panel runs the check described above. The Setupp Score page shows a scored booking, pricing lists what each plan includes, and the Acuity comparison is the closer read if your current form collects answers that nothing acts on. If you have not decided yet who should reach the calendar at all, what a qualification gate is comes first.