The diio MCP connects Claude to the intelligence in your team's sales conversations. The difference between a useful answer and a generic one almost always comes down to how you frame the question.
This guide explains how to ask diio for what you need in order to decide: where the risk is, where the opportunity is, what changed, and what you should do today.
Before you start: you need the diio MCP connected in your Claude account. If it isn't yet, start with How to integrate diio with Claude using the MCP.
Ask business questions, not searches
diio knows where to look across hundreds of conversations. Your job isn't to tell it what to search for, but what you need to decide.
A good question defines five things:
What you need to decide. The outcome you're after, not the data point.
Which period matters. Concrete dates, not vague expressions.
Where to look. A deal, a client, a team, a playbook, a segment.
Which signals matter. Risk, interest, objections, performance.
What you need to get back. Priority, explanation, comparison, or action.
An example of the difference
Weak question
"Analyze my sales conversations."
It returns everything and doesn't tell you what to do.
Strong question
"Which 5 opportunities need attention today? Review the open deals and their recent conversations. Prioritize by risk, missing next steps, relevant objections, and drops in interest. For each one, tell me what's happening, the evidence, and the recommended action."
The numbers tell you where to look; the conversations tell you why
Most useful questions combine three moments:
Signal. 7 opportunities lowered their success odds over the last two weeks.
Context. In 5 of them a budget objection came up, and 4 were left without a next step.
Action. Prioritize those deals and decide who should step in to regain momentum.
Questions to find out where to look:
"Which playbooks are closing best this quarter?"
"Where did sentiment drop compared to last month?"
Questions to understand why it's happening:
"Which objections are slowing down the most deals this month?"
"What keeps coming up in the conversations that aren't moving forward?"
How to narrow the search
You can scope any question using these criteria:
Date, client, executive, group.
Playbook, tracker, conversation type.
Sentiment, performance, success odds.
Spotting risk and opportunity in time
Some signals don't show up when you review conversations one at a time. These are the ones worth asking diio to track.
Before you lose a deal: falling interest, unresolved objections, competitors, budget, postponement, missing next steps, absent decision-maker, dropping sentiment, and low probability of closing.
Before you let an opportunity slip: new needs, other teams mentioned, expansion, urgency, interest in specific features, concrete next steps, and buying signals.
Example questions:
"Which opportunities need intervention this week? Prioritize by risk, explain which signals appeared, and tell me what the executive should do in each case."
"Where is there expansion we're not working on yet? Prioritize by how clear the signal is and recommend the next step for each account."
Prepare the next conversation, not the report on the last one
diio cross-references CRM deals with what was actually said in each conversation. That's useful both for the person selling and the person leading.
If you sell, walk into every meeting knowing what to do:
"I have a meeting with [client]. What do I need to know before I go in? Review the deal and its recent conversations. Tell me what changed since the last meeting, which topics are still open, what risks there are, and what I should aim to get today."
If you lead a team, spot where to step in before the next 1:1:
"Who is struggling most with discovery this month? Identify the recurring patterns, show the evidence, and compare them against the practices of the top performers."
It's especially useful before a renewal, a QBR, a forecast review, a 1:1, or a client meeting.
Seven questions to start with today
Pipeline. "Which deals need my attention today?"
Meeting. "I have a meeting with [client]. What do I need to know before I go in?"
Risk. "Which clients are showing risk signals, and why?"
Expansion. "Where is there expansion we're not working on yet?"
Coaching. "Who needs coaching this month, and on what specifically?"
Patterns. "What's holding back our winnable deals?"
Priority. "If I could only review five opportunities today, which should they be?"
For each one, define the period, where to look, and which signals matter. Ask for what's happening, the evidence, and the recommended action in every case.
The golden rule
The conversation already happened. What matters is what you do next. Don't ask what they said: ask what it means for the business.
Where is the risk?
Where is the opportunity?
What changed?
Who needs attention?
Where should I step in?
What should I do now?
diio turns conversations into signals. You turn them into action.
Keep in mind
The MCP works with read-only access: it cannot write or modify information in diio.
Each user can only access the information their permissions already allow them to see in diio.