Your recruitment CRM has an MCP. So what?

Recruitment CRM providers are making a lot of noise about their MCP integrations. AI connectivity. Natural language access to your database. Ask your CRM questions. Get answers in seconds. Great. But what are you actually going to do with it?

I've been spending time connecting CRM data to AI and building business development reporting packs. And frankly, this is one of the most interesting developments I've seen in recruitment sales technology for a long time — not because we can ask AI how many calls our consultants made last week. We've been able to report on that for years.

What's exciting is the possibility of going much deeper into the activity, relationships, conversations and commercial decisions behind the numbers. Why do some consultants win retained work while others consistently end up with contingent roles? Which clients are we spending too much time on? Which existing relationships have the potential to become significantly more valuable? And what could we learn if we compared the commercial behaviour of our highest-performing consultants with everyone else?

These are questions that have traditionally required a lot of manual analysis. With the right CRM access and a well-structured prompt, some of that work is becoming considerably easier. Here are seven reports I'd be experimenting with.

First, what does MCP actually do?

MCP stands for Model Context Protocol. In simple terms, it provides a standardised way for compatible AI applications to interact with external systems, including your CRM. Rather than exporting spreadsheets, cleaning the data and uploading everything manually, you can potentially ask an AI assistant to retrieve and analyse the information already sitting in your recruitment database — client records, opportunities, placements, commercial terms, activity history, notes and communications.

The important word is might. Different CRM connectors expose different information. Your subscription, permissions, available data and the specific MCP tools provided all affect what you can do. At the time of writing, October 2026, Recruit CRM and HubSpot are among the providers with documented MCP servers. More integrations and capabilities are emerging.

But MCP availability doesn't mean every report below will work automatically. Some analyses may require additional exports, reporting tools or access to information that your connector doesn't expose. With that understood, here's where I'd start.

1. What BD activity actually wins retained and exclusive business?

Every recruitment firm knows how many calls its consultants make. Far fewer can confidently explain what their consultants do differently when they win valuable, committed work. I'd start by examining the placements and signed agreements from the last 12 months, grouped by commercial model:

• Retained.
• Exclusive contingent.
• Non-exclusive contingent.

Then I'd ask the AI to examine the client activity leading up to those outcomes. How many conversations took place? Who was involved? What happened between the first conversation and the signed terms? How much discovery took place? Were senior stakeholders involved?

If the CRM exposes email bodies, meeting notes or call transcripts, we can go further. What were the consultants actually talking about — candidate availability and vacancies, or commercial problems, recruitment methodology, delivery risks and business outcomes? Are there recognisable differences between the conversations that led to committed work and those that didn't?

Analyse the client opportunities and placements from the last 12 months. Group them by retained, exclusive contingent and non-exclusive contingent terms. Compare the preceding BD activity, stakeholder involvement, conversation subjects and client engagement. Identify patterns associated with higher-value or more committed work. Distinguish recorded evidence from assumptions, and highlight missing information.

What I'd be looking for: The commercial behaviours associated with winning better terms — not simply the consultants doing the most activity.

Correlation isn't proof of causation. But if patterns emerge consistently, you've got something useful to investigate, test and coach.

2. Which existing client relationships should you be growing?

This could be one of the highest-value reports on the list. Most recruitment businesses spend a considerable amount of time chasing new clients while barely scratching the surface of the relationships they already have. I'd ask the CRM to identify our ten highest-revenue clients over the past 12 months, then assess the strength of each relationship. At Resonant, we use a model called REACH to examine five things:

• Rhythm: How consistently are we in contact? Is communication reciprocal, or are we always the ones chasing?
• Engagement: How meaningful are the conversations? Is the client actively discussing business challenges, or simply responding to transactional requests?
• Alignment: How many stakeholders do we know? Are we connected to the people who actually influence hiring decisions?
• Commitment: What evidence is there of exclusivity, retained arrangements, agreed terms or a willingness to invest in the relationship?
• Hero moments: Has the client recommended us, referred us internally or actively advocated for our work?

A CRM won't know everything about the relationship, particularly if important conversations aren't recorded. But it can provide useful evidence.

Identify our ten highest-revenue clients over the last 12 months. Assess the available evidence for contact rhythm, engagement quality, stakeholder coverage, commercial commitment and referrals. Highlight relationships with strong revenue but weak breadth or depth, and recommend three accounts for an expansion review. Mark any assessment where the CRM evidence is insufficient.

What I'd be looking for: Valuable relationships with obvious opportunities to deepen engagement, expand stakeholder access or develop additional business.

It's a much more useful starting point than telling every consultant to contact their top ten clients more often.

3. Where are you giving away fees?

One of the first things I'd want to understand is the gap between the fees we could be charging and the fees we're actually agreeing. I'd examine fee percentages and actual placement fees by:

• Client and client segment.
• Market and role type.
• Consultant.
• Retained, exclusive and contingent terms.
• Service or recruitment product.

Are particular consultants consistently discounting? Do certain client segments produce better fees and stronger commitment? Are we winning lots of low-value work in markets where our offer doesn't command much differentiation? And are we confusing a fee-negotiation problem with an audience-selection problem?

Sometimes the issue isn't that consultants can't defend their fees. It's that they're selling the wrong service to the wrong clients.

Analyse agreed fee percentages and placement values over the last 12 months. Segment the results by consultant, client type, sector, role, service and commercial terms. Identify where fees are consistently below our agreed benchmarks, and show whether the differences are associated with particular clients, segments, services or consultants.

What I'd be looking for: Where the firm can improve pricing discipline, client selection or the way its services are positioned and packaged.

Of course, you need to supply the right fee benchmarks. AI can't reliably tell you what your firm should charge without knowing the value you deliver and the market you serve.

4. Where are your opportunities getting stuck?

Most CRM dashboards show the value of the sales pipeline. I'd be more interested in understanding why opportunities aren't progressing. Take your recorded BD opportunities and examine the journey from initial engagement to discovery, proposal, agreed terms, job registration and ultimately a successful placement.

Where do deals stall? How long do opportunities sit between stages? Are some consultants brilliant at booking meetings but poor at obtaining commitment? Are proposals being sent without meaningful discovery? Are clients agreeing to discussions but never moving towards a commercial decision? And are we losing opportunities because of poor follow-up, weak qualification or a sales process that doesn't create enough buyer commitment?

Analyse our BD opportunity pipeline over the last six months. Calculate stage-to-stage conversion rates, average time in each stage and the most common points of drop-off. Compare patterns across consultants, client segments and commercial terms. Identify where the greatest potential improvement in conversion appears to be.

What I'd be looking for: The specific stage of the sales process that needs attention.

The answer might be more discovery training, better qualification or a more compelling proposition — but at least we'd have evidence to start with.

This works best when the CRM records meaningful sales stages and their history. If opportunities are simply marked open or closed, the analysis will be more limited.

5. What should you actually be coaching each consultant on?

This is where things start getting particularly interesting. Most recruitment managers have a reasonable sense of who is performing well and who isn't. But diagnosing why is considerably harder.

I'd want to compare each consultant's commercial activity with the outcomes they're generating — not simply calls versus placements. I'd look at the quality of their targeting, the opportunities they create, their conversion between stages, fee performance, client commitment and account development. If recorded conversations are available, I'd also look at what they're doing in discovery and how they're presenting their services.

Imagine being able to distinguish between a consultant who's speaking to plenty of the right people but struggling to progress opportunities, another who's excellent at conversion but targeting clients unlikely to buy valuable services, and another who's consistently winning placements but leaving significant fees on the table. Those are three very different coaching problems.

Review each consultant's recorded BD activity and commercial outcomes for the last 90 days. Identify patterns in targeting, opportunity creation, stage conversion, client commitment, fees and account development. Recommend one or two coaching priorities per consultant, supported by the available evidence. Do not infer individual skill deficits where the data is insufficient.

What I'd be looking for: A genuinely individual coaching plan.

Not another blanket instruction to make more calls. For me, this could be one of the most valuable applications of CRM-connected AI: turning the records of what consultants actually do into practical evidence for improving how they sell.

6. How much of your BD activity is actually worth doing?

Here's a report that could make a few sales managers uncomfortable. We measure BD activity obsessively. But how much of it is commercially worthwhile?

I'd want to examine the activity completed in the last week and classify it against a clear set of criteria. Was the client within our target market? Were we engaging an appropriate stakeholder? Did we have a credible commercial reason for the approach? Was the message relevant to the client's likely priorities? Was there a sensible follow-up plan? And did the activity create any meaningful engagement?

I'd also look at the opposite problem. Which consultants are spending disproportionate amounts of time following up with prospects who repeatedly show no evidence of interest? Where are we neglecting good opportunities while continuing to chase poor ones? You could build a simple weekly BD quality report — not to reward or punish consultants according to an AI-generated score, but to identify where activity is being invested well and where it isn't.

Review our recorded BD activity from the last seven days. Using our defined target-client criteria, assess which activities were directed towards relevant accounts and stakeholders, had a clear commercial purpose, and were followed up appropriately. Flag low-confidence classifications and missing records. Identify both neglected high-potential opportunities and areas of apparent overinvestment in poorly qualified prospects.

What I'd be looking for: Better allocation of commercial effort.

Because the answer to poor business development performance isn't always more activity. Often it's doing less of the wrong activity.

The important requirement here is to define what good looks like before asking the AI to assess it. Otherwise you're simply asking a language model to invent your standards.

7. Build your own commercial heads-up display

This is probably the application I'm most excited about. Once you've built a few useful reports, there's no reason their results need to disappear into another lengthy AI conversation. You could create a simple, visual commercial dashboard around the questions your business actually wants answered — and you don't necessarily need a permanent integration between your CRM and the dashboard. Here's the approach I'd experiment with:

• Step 1: Connect an AI assistant to your CRM through its MCP.
• Step 2: Run a repeatable set of reporting prompts covering the commercial measures you care about.
• Step 3: Export the results into a structured file — perhaps a Markdown report containing the commentary, alongside JSON or CSV for the underlying measures.
• Step 4: Upload that report into a simple dashboard you've built using a vibe-coding tool.

The dashboard then turns those results into something visual. Imagine opening a reporting page and immediately seeing:

• How much of this week's BD activity was directed towards priority clients.
• Which existing accounts warrant an expansion conversation.
• Where fees and commercial commitment are weakening.
• Which opportunities have stalled.
• Which consultants have the clearest coaching priorities.

Run the analysis again next week, upload the new results, and you've got an updated management view — no need to keep asking the same questions manually or commission a complicated, permanently connected reporting platform just to test the concept.

Using the agreed commercial reporting framework, produce a reporting pack for the last seven days. Include consistent metric definitions, reporting dates, source-record counts, a summary of material changes, key exceptions and recommended actions. Export a structured data table suitable for dashboard visualisation, together with a short narrative explanation of the findings.

What I'd be looking for: A repeatable, practical way of making commercial intelligence part of how the business is managed.

A reporting tool built around commercial decisions, rather than whichever charts happen to be available in the CRM.

One caution: once CRM information is exported, the CRM's access controls don't automatically protect the exported file. Any dashboard or AI tool receiving client or candidate data needs appropriate permissions, data minimisation and security controls.

Where I'd start

I wouldn't try to build all seven reports at once. I'd pick one commercial question that genuinely matters to the business. If your biggest challenge is winning committed work, start with report one. If you've got a valuable client base you're not developing properly, start with report two. If consultants are busy but commercial performance isn't improving, start with report six. And if you're managing a team, I'd be looking particularly closely at the coaching opportunities in report five.

Start with a manageable period of data. Check the evidence, challenge the assumptions and see whether the findings are actually useful. Then refine the prompt and run it again.

The first output probably won't be perfect. Your CRM data might be incomplete. Some fields might be unavailable through MCP. And you may discover that the way your team records activity needs improving. That's fine — understanding what the business doesn't know can be useful in itself.

The interesting thing isn't MCP

It's what MCP makes easier. For years, recruitment firms have collected huge amounts of information about commercial activity without necessarily learning very much from it. Calls recorded. Emails logged. Meetings booked. Opportunities updated. But collecting activity and understanding commercial performance are two very different things.

What excites me about MCP isn't the prospect of generating another dashboard. It's the opportunity to examine how recruitment firms actually win business, develop relationships, negotiate fees and convert opportunities — then use that evidence to improve the way their consultants work. Better systems. Better commercial decisions. Better skills. And ultimately, better business.

If you've started using your CRM's MCP, I'd love to know: what's the most useful report you've managed to produce? And what are you seeing now that you couldn't see before?

Ben Browning — Resonant