Solution Selling Drills: Score the Vision

What does the 9-box vision processing model actually score?

It scores sequence and ownership: who said what, and in what order. The model crosses three question intents — open, control, confirm — with three subjects: the reasons behind the problem, its impact on other roles, and the capabilities the buyer says they need. Open means the buyer speaks first without being fed the answer. Control means the rep offers a hypothesis when the buyer stalls, as in Is it because the handoff notes arrive after the call ends? Confirm means the buyer restates the point in their own language before the conversation advances.

Two behaviors fall straight out of that grid, and both are verifiable from a transcript. Order: no capability is named until an impact has been confirmed. Ownership: the admitted reason and the restatement come from the buyer, not the rep.

It is easy to score the vocabulary and miss the sequence. Read the transcript backwards from the first capability sentence the rep speaks. Before that line you should find an admitted cause and a confirmed consequence in the buyer's voice. A vision the seller narrated is not a vision; only the buyer's own restatement counts as evidence. If the rep supplied both halves, the box is empty no matter how polished the language sounded.

Rubric rows for your Solution Selling scorecard

Write rows as behaviors, not qualities. "Strong discovery" is not a row. "Buyer names a cause of the problem before any capability is mentioned" is a row, because two managers reading the same transcript will agree on it. A rubric row that cannot be checked against a transcript is an opinion with a number next to it.

Keep the rows tied to your call stages. If your discovery stage already requires a verified consequence and an agreed next step, the vision model does not replace those exit criteria — it tells you what good sounds like inside them. Our recommendation is to score four or five rows per session and leave the rest of the scorecard alone, so the debrief can isolate one behavior.

The table below is a starting set. Edit the pass bars to match your stage definitions, then calibrate them on one recorded session with your other managers before you grade anyone for real.

Drill 1: diagnose before you prescribe

Run it as a consultative discovery drill built from a live deal. Give the buyer persona a symptom only — tickets reopening, renewals slipping, a report nobody trusts — and a mild willingness to be helped. Tell the rep the constraint out loud before the timer starts: no capability, no product noun, no reference story until the buyer has confirmed an impact on a role other than their own. We recommend a ten-minute rep inside a fifteen-minute drill block.

Pass bar: the transcript shows an admitted cause in the buyer's words and a confirmed consequence, both before the rep's first capability sentence, with no more than one hypothesis offered per subject. Anything else is a fail and gets re-run this week with the same persona and a different symptom.

A failing attempt sounds like this: the rep asks what is going wrong, hears one complaint, and answers that's exactly the problem we solve. Fluent, friendly, and empty. Diagnosis is the drill; capability language is the reward the buyer earns by admitting a cost.

Debrief in ten minutes on one moment. Ask the rep to read the line where they first named a capability, then ask what the buyer had admitted before it. Let the self-diagnosis land before you render a verdict, and book the re-run in the same conversation. Our one behavior, then a re-run debrief format holds the structure.

Drill 2: pain chain practice across roles

The pain chain is the part reps skip because it takes nerve: showing how the frontline symptom becomes someone else's number. Set the persona as a manager two levels below the economic buyer, with a colleague whose quarter the problem quietly damages. The rep's job is to get the buyer to name that second role and its consequence, then to ask for access.

Pass bar: the buyer names the other role and states the consequence themselves, and the rep leaves with either an agreed introduction — person, date, purpose — or an explicit refusal with a stated reason. A polite maybe is a fail.

A failing attempt sounds like the rep supplying the chain: I'm sure your CFO cares about the write-offs. The buyer agrees to be agreeable and nothing is verified. Assumed pain in another department is a guess wearing the costume of qualification.

Score the ask separately from the diagnosis, because reps fail these two rows for different reasons. Weak diagnosis is a questioning problem. A missing ask is a courage problem, and courage responds to reps, not to advice. Re-run the drill with a persona who says no the first time. If the rep needs more work on getting to the second stakeholder at all, drill it with multithreading exercises before you raise the difficulty here.

Drill 3: capability language under demo pressure

Capability language survives discovery and then dies in the demo, where feature questions arrive in the buyer's own vocabulary. Set the drill there. The persona asks direct product questions — does it integrate, can it do custom fields, what about approvals — and the rep must answer each one with what the buyer would be able to do, then confirm whether that matters.

Pass bar: every answer names a buyer action and ends with a confirming question, and the buyer restates at least one capability before the rep proposes a next step. Product nouns in place of buyer actions are a fail even when the answer is technically correct.

A failing attempt has two signatures. The first is the spec recital: it supports SSO, SCIM, and role-based permissions, with no reference to anything the buyer said. The second is the reference story as filler — a customer name dropped to fill silence, unconnected to the admitted pain. Capability language describes what the buyer will be able to do, never what the product contains.

When the honest answer is that the capability is missing, the rep still has to hold the line without bluffing. That is a separate rep with its own pass bar; run it against our feature gap objection drills rather than folding it into this one.

Where is Solution Selling weak?

In two places, and you should say so before your team finds out on a live call. First, buyers who arrive with a requirement already written. The vision sequence assumes the buyer's picture of the solution is still forming; when procurement has published a spec, opening reasons questions reads as stalling. Train the pivot: acknowledge the written requirement, then diagnose how each line item got there and which lines are actually negotiable. That is a different drill with a different pass bar.

Second, reference stories. The method's insider stories are meant to prompt admission of pain the buyer has not voiced. Used as filler they cost credibility, and they are hard to score because they always feel like value to the rep telling them. Grade a story only when the buyer responds with an admission; otherwise mark it as talk. Our reference call drills treat proof separately for exactly that reason.

There is a third limit worth naming. Solution Selling structures a conversation; it does not qualify a deal, and it is not a substitute for qualification. A checklist like MEDDIC covers economic buyer, decision criteria, and decision process - and MEDDPICC adds paper process — none of which the vision model claims to score. Keep them in separate rubrics. And if your motion is high-volume transactional inbound, the full sequence is overhead: drill the opening and the next-step rows only, and be honest that the rest is not your fit.

How do you keep the scoring honest week to week?

Calibrate first, then protect the time. Have two managers score the same recorded session independently and compare row by row; where you disagree, the pass bar is written badly, not the rep. Re-run that check when you add a row or change a stage definition. Then put the drill block on the calendar as its own recurring commitment, separate from pipeline review, because deals are urgent and skills are only important.

The format we recommend costs a manager thirty minutes to run — five to set up, fifteen to drill, ten to debrief — plus roughly ten minutes beforehand to pick the scenario from a live deal and choose the rubric row. Add a fifteen-minute 1:1 built around one flagged moment, in addition to the pipeline 1:1 rather than instead of it.

Read the results per rep per scenario. Blended scores hide the scenario where a rep actually breaks, so read trends per rep per scenario. A rep can pass diagnosis against a friendly operations manager and collapse against a buyer holding a written spec; one average hides both facts.

If you run the drills in software, the useful part is that the scoring is graded against your loaded call stages and rubrics and that each flag links back to the exact transcript moment, which is how XL Roleplay reports work. Treat talk/listen ratio as an input to investigate, never as a pass bar. Where SDRs own the first conversation, drill the opening reasons questions inside prospecting call practice instead of the discovery format.