Mine post-sale data for cross-sell opportunities

Step 1: Define what a cross-sell signal actually looks likeExcel

Look back at your last several successful expansions and write down the usage pattern that preceded each one as a specific, numeric rule.

Hitting a usage cap 3+ times in a month, 90%+ seat utilization, adoption of a feature that commonly pairs with an add-on — rather than a vague sense of "accounts that seem engaged."

Step 2: Analyze the usage exportCopilotClaude

Load the usage data into Excel and ask Copilot to surface accounts matching those defined signals, rather than a generic "who looks like they might buy more."

Prompt idea:

Using this usage export, flag any account that has hit their usage cap 3+ times in the past month or has more than 90% seat utilization. For each, state which signal triggered the flag and the specific numbers behind it.

Step 3: Prioritize by signal strength and account valueExcel

Rank the flagged accounts by how strong the signal is and the account's existing contract value, so the account team works the highest-value, clearest opportunities first.

Step 4: Hand the list to account owners with the evidence attachedExcel

Give each account owner the specific usage data behind the flag, not just a name on a list — that's what lets them have a grounded conversation instead of a generic upsell pitch.

Check the work

• Verify the usage numbers behind a sample of flagged accounts against the source export directly.
• Confirm the account isn't already in a churn-risk or support-escalation state — a cross-sell pitch to a frustrated customer backfires.
• Check that flagged signals reflect recent, sustained usage, not a one-time spike.

Source: Microsoft Copilot Scenario Library — Sales (2026)