Deal Clinic

Deal Clinic exists to make deal quality observable. Instead of asking a rep "how's that deal feeling?", you ask a model that has been taught what your best people look for, and you get the same lens applied to every conversation, every rep, every time.

The two parts

• Discovery Expert Model (DEM): the framework. A DEM is a structured definition of what matters in a sales conversation. You break a deal into a handful of tracks (the areas your experts always probe), and for each one you describe what a good opportunity looks like, what a weak one looks like, and the critical questions a rep should be asking. This is where you capture expertise that usually lives only in one person's head.
• Deal Diagnostic Readout: the output. When a rep runs a deal through Deal Clinic, the AI reads the conversation, evaluates it track by track against the DEM, and produces a readout: an executive summary, a per-track assessment, the gaps it found, risk signals, a plan to close, and the questions to bring to the next meeting.

Why a structured diagnostic beats gut feel

• Consistency. Every deal is judged against the same expert standard, so a strong call and a thin one are scored the same way no matter who ran them.
• Coachability. The readout names specific gaps and specific follow-up questions, which is something a manager can act on. "This feels shaky" is not.
• Honesty about what you actually know. Deal Clinic separates "this deal is weak" from "we did not ask enough to know." If a rep only skimmed a topic, the readout flags it as INSUFFICIENT instead of inventing a verdict.
• A teaching tool. Reps see the questions an expert would have asked but they did not, which raises the quality of the next conversation.

Where it lives

• Expert Models (DEMs) is in the Admin side navigation at /dem-cog/dems. This is where admins and experts build and maintain frameworks.
• Deal Clinic (the readouts list) is in the Reports side navigation at /dem-cog/reports. This is where reps and managers diagnose deals and review results.

Deal Clinic is a module that has to be enabled for your studio. If you do not see these areas, the module is off for your studio.

A Discovery Expert Model is the heart of Deal Clinic. It is how you teach the AI to reason like your strongest seller instead of a generic one. The quality of every readout depends on the quality of the DEM behind it, so this is the part worth getting right.

Open Expert Models (DEMs) under Admin at /dem-cog/dems and click Create New DEM.

Model identity

• Model Name (required). Name the framework for the motion it covers, for example "Enterprise Discovery" or "Mid-Market Renewal Risk." This is what reps pick from when they diagnose a deal.
• Description. A short note on what this model assesses and when to use it. You can dictate it with the voice input if that is faster.

Assessment Type: the lens the model applies

The Assessment Type decides how the AI interprets a positive versus negative finding on each track. Pick the one that matches the question you actually want answered.

Type	Question it answers	Positive means	Negative means
Value-Add Scope (NORM)	Where can we add value?	Clear whitespace: the prospect has gaps or needs we can address	Limited delta: the prospect already operates at a high standard
Deal Risk Assessment (RISK)	Is it safe to pursue this deal?	Low risk: clear signals, commitment, urgency, a defined process	High risk: vague timelines, single-threaded, seller-carried momentum
ICP Fit Assessment (FIT)	Is this our ideal customer?	Strong fit: right size, industry, maturity, and use case	Weak fit: likely to struggle, churn, or carry misaligned expectations

Value-Add Scope is the default and the right starting point for most discovery frameworks. Reach for RISK when you want a forecast or pipeline-hygiene lens, and FIT when qualification (whether to invest in a deal at all) is the real question.

Discovery Tracks: the areas you assess

A DEM needs at least one track, and most useful models have a handful (for example Revenue Stability, Technical Readiness, Budget Authority). Each track is one dimension your experts always probe. For every track, fill in:

• Track Name and an optional Category for grouping.
• Positive Aspects (from the seller's perspective). What does a good opportunity look like here? For a Value-Add Scope model, that is a prospect with gaps or needs where you can add real value. Be concrete; this is one half of the rubric the AI scores against.
• Negative Aspects (from the seller's perspective). The mirror image: what a weak opportunity looks like, for example a prospect already strong in this area with little room for you to add value. If you find yourself describing these backwards, the Swap pos/neg button flips the two fields.
• Tacit Expert Knowledge. This is the most valuable and most overlooked field. It is where you capture the nuance your expert would struggle to write down: mental models, what "good" really sounds like, the tells that separate a real signal from noise. Use the microphone to dictate it, because experts explain this kind of thing far better out loud than in a text box. The AI uses this context to judge the track, not just the bullet definitions above it.
• Critical Discovery Questions. One question per line. These are the questions a rep should be asking on this track. They do double duty: they sharpen the AI's analysis, and they are shown to reps as a guide while they capture a conversation.

Drag tracks (or use the up and down arrows) to order them the way you want reps to think about the deal. The form saves the moment you click Save, and you can keep editing afterward.

Visibility

New DEMs are visible only in your studio, which is what you want for studio-specific playbooks. Global DEMs (available across every studio) can only be created by Salesably superUsers. If you see a DEM marked Global, it is a shared framework managed centrally, not by your studio.

Set a studio default

On the DEMs list, admins and supervisors can mark one active DEM as the studio Default. The default is the model Deal Clinic reaches for automatically when a readout is generated without someone hand-picking a DEM, for example on automatic diagnostics for recorded meetings. Choose the framework that best represents your standard discovery motion.

Once a DEM exists, reps use it to diagnose real deals. A diagnosis takes a conversation and returns a Deal Diagnostic Readout. The goal is not a grade for its own sake; it is a clear picture of where the deal stands and a concrete list of what to do next.

Open Deal Clinic under Reports at /dem-cog/reports and click Diagnose Deal.

Choose the framework and set the context

• Expert Model (DEM). Pick the framework to apply. The list shows each model's type (Value-Add Scope, Risk, or Fit) so you can match the lens to the question. The studio default is preselected when one is set.
• Deal Reference / Account Name. A label so you and your manager can find this readout later, for example "Acme Corp - Enterprise Deal."
• Call Stage. Where in the cycle this conversation sits, from Inbound Inquiry through to a Closing Call. The stage shapes how the AI reads the conversation, because what counts as strong discovery on a first meeting is different from what you expect at final negotiation.
• Meeting Timeline and Expected Close Date. First meeting, last meeting, and the next scheduled meeting. A deal with no next meeting on the calendar is flagged as a stall risk, which is deliberate. The expected close date drives Deal Clinic check-in reminders; deals without one are skipped from those emails. You can also turn the per-deal auto-reminder off here.

Provide the conversation

The Deal Input card lets you supply the conversation in whatever form you have it:

• Notes. Paste rough notes or a summary you typed up after the call.
• Upload. Attach a file (.txt, .md, .csv, .log, or .docx) and the text is pulled in for you.
• Voice. Dictate a spoken summary of the call and let it transcribe.
• Past. Reuse a conversation Salesably already has on file (a recorded meeting or a previous analysis), with no copy and paste.
• HubSpot or Pipedrive. If your studio has the CRM integration enabled, import a deal directly. See "Importing a Deal from Pipedrive or HubSpot" below.

The more real the input, the better the readout. A full transcript gives the sharpest analysis; rough notes still work but the AI will more often mark a track INSUFFICIENT because it genuinely cannot tell.

Use the Discovery Tracks Guide

When you select a DEM, its tracks and critical questions appear in the Discovery Tracks Guide beside the input. Reps can use this as a live checklist of what to capture, and there is a print option to take the questions into the call.

Click Diagnose Deal and the AI analyzes the input against the model, then opens the readout.

A Deal Diagnostic Readout is meant to be read top to bottom and acted on, not just glanced at for a number. It mirrors the COG idea: Capabilities (where the deal stands), Opportunities (where you can add value), and Gaps (what is missing or unknown).

How each track is scored

Every track in the DEM gets two separate judgments, and keeping them separate is the whole point.

• Capability level: POSITIVE or NEGATIVE. This is the opportunity read from the seller's perspective. On a Value-Add Scope model, POSITIVE means the prospect has gaps or needs you can address (a good opportunity), and NEGATIVE means they are already strong here (less room to add value). On a Risk model the same scale reads as low risk versus high risk, and on a Fit model as strong fit versus weak fit.
• Information Strength: STRONG or INSUFFICIENT. This is the readout's honesty check. It asks whether the rep actually discovered enough to support a verdict. A topic being mentioned is not the same as it being discovered. If a rep heard something interesting and did not dig in, the track is marked INSUFFICIENT, and an INSUFFICIENT track is always treated as a red flag (it cannot be scored as a positive). When that happens, the readout proposes the exact follow-up questions to ask next time.

This is the feature that makes the tool trustworthy. It refuses to call a deal healthy on thin evidence, and it tells the rep precisely what to do about it.

What the readout contains

• Executive Summary: the situation, the buyer's goals and drivers, current performance and capabilities, the opportunities, a possible bridge to value, and the immediate next step, plus a ready-to-send follow-up email draft.
• Component (track) analysis: per-track capability summary, gaps, gap severity, information strength, and follow-up questions.
• Discovery & Information Gaps: the tracks where you do not yet know enough.
• Risk Signals Detected: phrases from your Lexicon of Risk (or close paraphrases) that surfaced in the conversation. See "Creating a DEM" and your studio settings for how the lexicon is configured.
• Plan-to-Close: the bigger moves needed to reach a decision, broken into the decision process, the critical gaps to close, the buyer commitment events to drive, and a realistic decision window.
• Next Meeting Focus: the agenda and questions to carry into the next conversation.
• Legend: an always-visible key explaining capability levels and Information Strength, so anyone reading the report understands the scoring.

Manager's Notes and re-running

Admins and supervisors can add Manager's Notes to any readout. These are visible to everyone viewing the report and, importantly, are fed back into the AI as context if the report is re-run, so coaching guidance actually shapes the next analysis. Use them for questions to the rep, strategic steers, or correcting context the AI missed.

Any readout can be re-run with Re-run Analysis, optionally against a different DEM. Re-run after you have updated the source material, added manager notes, or want to see the same deal through a different lens (for example switching from Value-Add Scope to Risk).

Ask Studio coach

Each readout has an Ask Studio coach. It already knows this report's contents, so reps can ask it to plan next moves, draft messaging, or pressure-test the plan to close without re-explaining the deal.

Most of your richest source material is the conversations Salesably has already recorded and transcribed. Rather than copying a transcript into the Diagnose Deal form, you can launch a diagnosis straight from a transcript result, so a recorded meeting becomes a Deal Diagnostic Readout in a couple of clicks.

How to run it

1. Open a transcript under Results and click Run Diagnostic in the action bar.
2. Pick the Expert Model (DEM) to apply. The studio default (or the first active model) is preselected.
3. Confirm the Deal Reference, which is pre-filled from the transcript title, and the Meeting Date.
4. Choose the Call Stage that matches the conversation.
5. Click Run Diagnostic to generate the readout from the transcript text.

The saved readout is linked back to its source conversation, so you can move between the transcript and the diagnosis later. This is the fastest path from a real meeting to a coached next step, with no manual data entry.

If your studio has the Pipedrive or HubSpot integration enabled, reps can diagnose a deal straight from the CRM instead of pasting notes. Deal Clinic reads specific fields from the CRM record and assembles them into the input the AI analyzes. This keeps the diagnosis anchored to the same deal record your team already works from.

The importer appears as a tab in the Deal Input card on the Diagnose Deal screen, and reps connect their account from there the first time.

Finding the deal

• Browse recent shows the most recently updated open deals. Won and lost deals are excluded.
• Search finds deals matching what you type.

What gets pulled in

From the deal record: the deal name, status, value and currency, pipeline stage, the linked organization, the primary contact, the expected close date, and a link back to the deal in the CRM. From Notes and Activities: the most recent notes and engagements (calls, meetings, tasks) attached to the deal, with their timestamps.

Where the deal narrative has to live

This is the part that trips people up. For the AI to read your call narrative (discovery details, customer pain, next steps), that narrative must be in the deal's Notes or in an Activity or engagement description. Deal Clinic does not read emails, attached files, custom fields, or call recordings stored in the CRM.

Tip: Paste a call summary or transcript into a Note on the deal before importing. You can also edit the imported text in the form before running the diagnosis.

Deal Clinic can run on its own. When the module is enabled and a meeting is recorded with the Salesably meeting bot (Zoom, Google Meet, Teams, and similar), a Deal Diagnostic Readout is generated automatically from the transcript, so coaching insight shows up without anyone remembering to ask for it.

How it works

1. A meeting is recorded with the Salesably meeting bot.
2. The transcript is analyzed and the rep receives the standard transcript analysis email.
3. If Deal Clinic is enabled, a readout is generated. It uses the studio default DEM when one is set, and otherwise the first available Expert Model.
4. The rep gets a separate email with the diagnostic results: the score, key insights, and a link to the full readout.

Call stage on automatic readouts

The AI infers the call stage from the conversation itself rather than assuming one. If it genuinely cannot tell which meeting in the cycle this was, it falls back to First Full Meeting. You can always re-run the readout with a specific call stage if the inferred one is off.

To control which framework automatic readouts use, set a studio default DEM on the Expert Models page (see "Creating a DEM").

As a team generates readouts, the Deal Clinic list at /dem-cog/reports becomes a running record of deal health across your pipeline. Filters help managers focus that record on the rep, deal stage, or coaching question in front of them.

Available filters

• Filter by user. View readouts for a specific team member, or all of them. This is how a manager reviews one rep's deals before a one-on-one. Regular members see only their own readouts; this filter is for admins and supervisors.
• Filter by call stage. Narrow to a single point in the cycle, for example only First Full Meetings or only Final Negotiations, to compare how deals look at the same stage.

Reading the list at a glance

Each row shows the deal reference and rep, the next scheduled meeting (with a clear flag when none is booked), the call stage and DEM used, a quick positive-track count, and an Information Strength indicator. A note icon marks readouts that already carry Manager's Notes. Use the filter controls at the top of the page to combine user and call stage and zero in on what you need.