The AI Scorecard
Dashboards show the value Matador delivers. The AI Performance Scorecard shows the value your store is leaving on the table — it reads every conversation, every user, every intent, then tells you in plain language what's leaking and what to fix.
Read Everything.
Name the Leaks. Hand Over the Plan.
It reads everything
5,112 conversations · every user · every intent
AI Reply · AI Follow-Up · reviews · handoffs · monthly
A GM can't read 5,000 conversations to find out why out-the-door pricing deals die. The scorecard reads all of them — every month.
It names the leaks
"Your team has sent zero videos." "OTD conversations convert at 7%." The uncomfortable truths a dashboard is too polite to surface, in plain language.
It writes your next 30 days
one broadcast · two trainings · one playbook
specific · actionable · tied to real conversations
Every scorecard ends with a concrete action plan — the performance manager's next meeting agenda, pre-written and prioritized by what it's worth.
Seven Sections.
Zero Sugarcoating.
The numbers below come from an actual 30-day scorecard — a franchise store on the full Matador suite, names anonymized. Every store gets the same spine: what happened, what's leaking, and exactly what to do about it.
AI handled 54% of all replies at a consistent 15-minute median while the team averaged 3 minutes — 457 AI-influenced wins across 701 active conversations, with 6 appointments booked and a 2% conv-to-appt rate leaving clear room to grow.
Human vs. AI reply time
| Who replied | Replies | Median | 90th pct |
|---|---|---|---|
| Your team | 125 | 3 min | 7.8 hrs |
| AI Reply | 147 | 15 min | 15 min |
The team is fast when it answers — the 90th percentile shows leads waiting hours. The AI's floor never moves. AI Follow-Up added 429 proactive messages to 195 cold leads, and 38% replied.
What customers are asking about
| Customer intent | Convos | Booked | Rate |
|---|---|---|---|
| Buying new | 76 | 2 | 2.6% |
| Buying used | 31 | 0 | 0.0% |
| Models & configurations | 11 | 0 | 0.0% |
143 conversations carried a buying signal. The scorecard also caught 4 mandatory AI instructions — price, down payment, credit, trade-in — still sitting blank, and said so.
Hand-offs to your team, by name
Completion looks perfect until you read the reply rates: the top rep answered 75% of her hand-offs, three others answered 0%. Seventy customers left waiting — each one tied to a name.
Which outreach actually gets answered
| Sequence | Leads | Reply rate |
|---|---|---|
| Inbound / chat | 21 | 71.4% |
| After hours | 85 | 50.6% |
| Third-party leads | 181 | 36.5% |
| Trade-in | 18 | 11.1% |
The verdict was specific: rewrite the trade-in opener around a concrete appraisal offer, and copy the inbound/chat playbook to the sequences that lag.
The biggest unused lever
| Group | Convos | Appts | Rate |
|---|---|---|---|
| Received a video | 0 | 0 | — |
| No video | 322 | 6 | 1.9% |
Zero videos sent across 322 new conversations. The scorecard flagged it in orange and put "record one 60-second walkaround" in the plan — with a named owner.
An idle channel, called out
No hedging: "this channel is completely idle while sequences work hundreds of leads." One targeted send to the unbooked pool made the next-30-days list.
Proof the AI can carry a deal — receipts included
Every scorecard closes with real conversations where the AI booked the appointment start to finish. This month, all four featured wins were booked autonomously:
Your Next 30 Days,
Ranked by Conversion Impact
No "consider exploring." Six actions from the same scorecard — each with an owner, an effort estimate, and what it's worth.
| # | Action | Owner | Effort | Est. lift |
|---|---|---|---|---|
| 1 | Fill the 4 blank mandatory AI instructions | GM / BDC Manager | 1–2 hrs | +appointments |
| 2 | Launch a walkaround video in the 71%-reply sequence | BDC Rep | 2–3 hrs | +conversion |
| 3 | Coach the three 0% task responders | Sales Manager | 1 hr | +task conversion |
| 4 | Rebuild the 11%-reply trade-in sequence | BDC Manager | 2 hrs | +reply rate |
| 5 | Send one re-engagement broadcast to unbooked leads | BDC Manager | 1–2 hrs | +pipeline |
| 6 | Add a follow-up prompt for buying-used leads | BDC Manager | 1 hr | 0 → booked |
Accountability,
Down to the Rep.
BDC performance metrics tied to real conversations, not gut feel — who's keeping up with the AI, who's leaving deals on the floor, and exactly what to change first.
User-level accountability
The rep completing 65% of 78 tasks gets highlighted as the model — "have Vanessa share her approach in a team huddle." The reps at 0% get a conversation with the manager. Handoff and task-completion rates, named per user.
Human vs. AI reply time
See exactly how long customers wait for your team versus your AI — and what that gap costs in deals. The number that ends the "we're fast enough" debate.
Coaching from what already works
Intent-level recommendations and lead-source coaching — copy the instructions and frequencies that win to the sources that lag. Plus the greatest hits: a curated list of the best AI conversations, proof of what good looks like.
Accountability
Has Names on It
The scorecard doesn’t stop at the store level. Every rep’s follow-through on AI handoffs is measured — because the AI can set the table, but somebody still has to show up.
| Rep | Handoffs answered | Avg response | Tasks closed | Trend |
|---|---|---|---|---|
| Dana R. | 96% | 4 min | 41 of 43 | Improving |
| Alex M. | 88% | 11 min | 33 of 39 | Improving |
| Sam K. | 74% | 19 min | 26 of 38 | Flat |
| Jordan P. | 61% | 38 min | 12 of 35 | Needs coaching |
In one audit, engaged users completed 65% of AI-created tasks — disengaged users completed0%. The gap is coaching, not talent. The leaderboard shows you exactly where to spend the coaching.
Get the Diagnosis
and the Prescription
On the demo, we'll walk a real scorecard end to end — the leaks, the per-rep accountability, and the "Your Next 30 Days" plan your managers will actually run.