Not a guess. Not a gut feeling. Actual data — pulled from across your systems, organized, and summarized before the call even starts.

That is not a fantasy. It is increasingly possible today, and RevOps teams are in the best position to make it happen.

This post is about a practical approach to getting there — without a 12-month data warehouse project, without a dedicated data engineering team, and without waiting for “the perfect unified platform.”

The Problem Is Not Data. It Is That the Data Is Everywhere.

Before any important customer meeting, an Account Executive, CSM, or SDR ideally wants to know:

• What is the customer currently using, and how actively?
• Are there signs of renewal risk or churn?
• Is there a potential expansion or upsell opportunity?
• What support issues are repeating, and are they escalating?
• How is the account’s ARR trending, and what does overall account health look like?
• What is happening in their market right now that might affect them?
• What should we actually talk about in this meeting?

Every one of those questions has an answer. The problem is where those answers live.

The renewal risk signal might be in Salesforce. Product usage data lives in Amplitude or Mixpanel. Support trends are in Zendesk or Jira Service Management. Billing and ARR data is in your finance system or data warehouse. Market news is on LinkedIn, Google, or a news aggregator. Call summaries are in Gong or Chorus. Customer health scores might be in Gainsight or ChurnZero.

So your AE opens six tabs, pings two colleagues, checks the Slack channel, and still walks into the meeting with an incomplete picture. And this happens before every single customer call.

This is not a people problem. It is a systems and workflow problem.

Why “One Unified Platform” Is Hard to Achieve in Practice

The vision of a Customer 360 — one place where all customer data lives — sounds great in a board deck. In practice, it runs into some predictable walls.

First, data consolidation takes time. Migrating or syncing data from five to ten systems into a single platform is a multi-month project at minimum. It requires engineering capacity, data modeling decisions, and ongoing maintenance.

Second, different teams own different systems. Your Support team is not giving up Zendesk. Your Finance team will not move off NetSuite. Your Customer Success team has configured Gainsight to work exactly how they need it. Centralizing data often means asking teams to change their workflows, which creates resistance and slows everything down.

Third, priorities shift. The Customer 360 project kicks off, runs into a technical blocker, gets deprioritized when the company hits a rough quarter, and quietly dies. This happens more often than anyone likes to admit.

The result: teams keep working in silos, the data stays scattered, and the sales team keeps doing manual research before every meeting.

There is a more practical path.

A Practical AI-Led Approach: Let the Data Stay Where It Is

Instead of moving all data into one place, the idea is to bring AI to where the data already lives. Here is how that works in practice.

Step 1: Let every team keep using their own tools.

Do not ask your Support team to change how they log tickets. Do not ask Finance to restructure their billing data. Keep existing systems as the source of truth.

Step 2: Connect those systems using MCPs (Model Context Protocol).

MCP is an open standard that allows AI models to connect with external tools and data sources. Many platforms already have MCPs available — Salesforce, GitHub, Jira, Slack, and others. Where an MCP does not exist yet, you can build a lightweight one using the platform’s API. The technical lift is significantly smaller than a full data migration.

Step 3: Bring those MCP connections together so AI can access the right data.

Once you have MCPs set up for your core systems, you can configure an AI layer that is aware of all of them. The AI does not need to store the data — it just needs to be able to query it in real time when asked.

Step 4: Connect to the LLMs your organization already uses.

Whether your team uses Claude, ChatGPT, Gemini, or another model, the same approach applies. The AI model connects through the MCP layer to pull relevant data from the right systems on demand.

Step 5: Build predefined skills or prompts for specific use cases.

This is where it becomes genuinely useful for sales and RevOps teams. Instead of asking the AI a vague question, you define structured prompts for specific use cases:

• Account Meeting Prep — Summarize everything relevant about this account before my call today.
• Renewal Risk Summary — What are the signals that this customer might not renew?
• Expansion Opportunity Summary — Are there signals that this customer is ready for an upgrade or additional products?
• Support Trend Summary — What recurring issues has this customer raised in the last 90 days?
• Executive Briefing — Prepare a one-page summary of this account for the QBR.

These prompts can be made available to the whole team, so every AE, SDR, and CSM is working with the same level of insight — not just the ones who know where to look.

What an AI-Generated Customer Insight Summary Could Look Like

Here is a practical example of what your team could see before walking into a customer meeting:

Account: [Customer Name] | Meeting Date: [Date]

What they are using: Core product active across 4 of 6 licensed modules. Usage of the analytics module dropped 40% over the last 60 days.

Growth or slowdown signals: Headcount grew 15% last quarter based on LinkedIn data. Two new regional offices opened recently, which may indicate expansion potential.

Renewal risk signals: Contract renews in 87 days. NPS score dropped from 62 to 44 in the last survey. Two open escalations in support.

Expansion signals: The customer has not enabled the advanced reporting feature. Three users have requested it through the in-app feedback tool.

Repeating support issues: 7 of the last 11 tickets relate to SSO integration errors. Root cause is still unresolved. Escalation risk is moderate.

Recent market news: The customer’s industry is facing new compliance requirements effective Q3. This may be driving the change in product usage patterns.

Suggested talking points for this meeting:

• Address the open escalations directly and share resolution timeline.
• Ask about their expansion plans given recent headcount growth.
• Introduce the advanced reporting feature as a solution to the analytics drop-off.
• Bring up the compliance topic — this could be an opportunity to position the premium tier.

This is not a hypothetical. This is what becomes possible when AI has access to the right data at the right time.

Who Benefits Beyond Sales

While the immediate value is clear for Account Executives, SDRs, and CSMs, the same approach extends naturally to other teams.

Customer Success teams can use it to flag at-risk accounts earlier and prepare for QBRs without spending half a day gathering data.

Support teams can use it to understand customer context before jumping on a call — is this a high-value account, are they close to renewal, have they escalated before?

Finance teams can use it for revenue forecasting and to surface accounts with unusual billing or usage patterns.

Leadership and RevOps can use it to get account-level or portfolio-level views on demand, without waiting for a weekly report.

The underlying infrastructure — the MCP connections, the AI layer, the predefined prompts — is built once and used across the organization.

Where to Start

If this feels like a big initiative, it does not have to be. Start with one use case and one set of systems.

For example: connect Salesforce and your support tool via MCP, and build a simple Account Meeting Prep prompt. Give it to five AEs. See if it changes how they prepare for calls. Iterate from there.

This is not about replacing your existing tools or workflows. It is about adding an AI layer that makes everything your teams already do faster and more informed.

The data is already there. You just need to connect it.

If you are trying to solve this problem for your RevOps or Sales team, connect with me. I am currently working on implementing this kind of approach and would be happy to exchange ideas.