MCP Use Cases and Best Practices
Scout or the MCP? Scout lives inside MedScout and answers quick questions in the moment. The MCP brings MedScout data into Claude, where you can combine it with your own files, your own templates, and anything else you're already working on.
Not yet on the Claude connector? Connect the MedScout MCP →
Where to start
When | Ask MedScout to |
|---|---|
Before a call | Pull a provider's volume, referral patterns, payer mix, and reimbursement so you know why they matter |
Building your pipeline | Generate a target list by geography, specialty, or code set. No filters needed |
Planning days in the field | Build a route around the highest-volume stops and the days providers are actually in the office |
Making the ROI case | Combine MedScout data with your own numbers into a customer-ready model |
Prepping a QBR | Fill your existing deck with account-specific market data |
Considering a new market or territory | Show volume trends, site of care shifts, payer mix, and reimbursement across any geography |
Determining market share | Cross-reference your CRM, ERP, or install base against the market to find what's missing |
How to effectively use MedScout's AI
Each of these includes a prompt you can copy and paste. Fill in the bracketed parts and go.
Before a call: pre-meeting brief
Who they are, what they do, how they get paid, and how to approach them. Claude frames it in the context of your commercial strategies rather than handing back a profile.
Prompt
I have a call with a provider and I want to walk in knowing why they matter.
Pull who they are and where they practice, the procedures they perform and at what volume, their patient population, their site affiliations and health system membership, their payer mix, and their reimbursement on the codes I care about. Include trend direction where the data supports it.
Frame it around what I should lead with and what I should ask. Tell me what's unusual about this provider compared to others with a similar profile, and say where the data is thin rather than filling the gap.
Provider: [Name and location] Target Procedure or Diagnoses: [Names or Codes] Strategy: [Strategy name in MedScout]
Other ways to ask
- "Why should I care about Dr. Johnson?"
- "I have a call with Dr. Patel tomorrow. What should I know?"
- "Who practices at [hospital name], and which of them are doing [procedure]?"
Building your pipeline: discovery and targeting
Describe who you're looking for in plain language and get a ranked list back, scoped to a territory, region, or nationally. Simple volume-based searches come back almost immediately. Multi-criteria targeting takes a few minutes while a strategy runs in the background.
Prompt
Find the providers who match, ranked by fit, scoped to the geography I give you. Show volume, patient counts, and site of care so I can see why each one landed where it did.
If my criteria are going to return either almost nothing or thousands of matches, tell me before you run it and suggest what to tighten or loosen. When the list looks right, save it as a provider list I can work from.
Who I'm looking for: [Description] Geography: [Territory, state, region, or nationally]
Other ways to ask
- "Find urologists with strong robotic surgery volume and at least 25% commercial payer mix."
- "I want to find providers in Los Angeles County who recently started performing knee replacements for the first time."
- Search for physicians performing [procedure] in Florida who have no open payments from [competitor]."
- "Show me everyone in my territory doing [code] who isn't already on one of my lists."
Planning days in the field: routes and office days
Build routes from search results, a strategy, or an existing list, with addresses resolved for you. Paired with day-of-week data, routes get built around when providers are actually in the office rather than just drive time.
Prompt
I have a limited number of days in a city and I want the route to be worth the trip.
Build a route from the providers I give you, ordered to keep drive time reasonable but weighted toward the highest-volume stops and the days those providers are typically in the office.
If a high-value provider's office days don't line up with my trip, tell me rather than dropping them. Save the result I land on as a named route.
Location and dates: [Where and when] Providers: [List name, strategy, or results from the last search]
Other ways to ask
- "What days is Dr. Knopp in Brownsville typically in each of his practicing locations?"
- "I have two days in Phoenix. Build a route that hits the highest-volume providers for [procedure, code set, or strategy] and the referring providers feeding them."
- Aft"Create a route called 'Monday AM, Downtown' and add the top 10 providers from these results."
Making the ROI case
Turn back of the napkin math into a data-backed conversation. You bring the revenue or cost-savings assumptions, the MCP supplies the market inputs.
Prompt
You're building the economic case for a site that I will present to them.
Pull volume for every provider at the site for the codes I name, apply the payer mix and the then the average reimbursement for that site and code. Then apply my assumptions to get the total opportunity.
Keep it to something I could put in front of a site admin or physician. Flag any input you had to estimate rather than pull.
Site: [Name] Codes: [Code set or procedures] Template: [Attach a template for it to work the data into, if you have one] My assumptions: [Revenue or cost savings per procedure, adoption rate, anything else]
Other ways to ask
- "What's the addressable volume for [procedure] within 50 miles of [site]?"
- "Using this site's payer mix and reimbursement, estimate the revenue increase they'd see for [procedure] if they increased throughput by 25%."
- "Turn this into a one-page ROI summary I can leave behind."
Prepping a QBR or site conversation: fill your own deck
You already have a deck you run in your sales cycle. Hand it over and ask for the account-specific market data to be filled in.
Prompt
Here's the deck I already use. Fill in the market data for a specific account.
Keep my structure, my slide order, and my language. Replace only the data placeholders with real numbers for this account: provider volume, site of care mix, payer mix, reimbursement, and trend direction.
Don't add slides or reorganize. If a slide needs data that isn't available, leave it and tell me which one so I can decide what to do.
Account or site: [Name] Deck: [Attach it]
Other ways to ask
- "Build the QBR section on market opportunity for [account] using 2026 data for [procedure]."
Reading a new market or territory
Anything available in Analytics is available here: volume trends, payer mix, place of service, average reimbursement, market size. Custom time frames work too, so July 2024 through July 2025 is as easy as a calendar year.
Prompt
I'm trying to understand a market I don't know well yet.
Show me the size and shape of it for the procedures I name: total volume and how it's trending, where those procedures are being performed, who the highest-volume providers are, and what payers are reimbursing and what the average amount is.
Tell me what stands out, whether that's a site of care shift over the last few years, a provider growing unusually fast, or a payer paying well above or below the rest. Keep the commentary short and grounded in what the data actually shows.
Market: [Procedures or code set] Geography: [Territory, state, region, or nationally] Time frame: [Year] to [Year]
Other ways to ask
- "How is [condition] volume trending in my territory year over year?"
- "Where are [procedure] being performed in my territory: office, ASC, or hospital?"
- "Compare all territories for CABG volume in 2025 and so far in 2026."
- "Which of my territories is most underserved relative to market opportunity for [code set]?"
Working with your own data
MedScout data plus your own sales, CRM, or ERP data plus whatever third-party sources you care about, cross-referenced in one place. Teams that already know their market use this to find the intersection they've never been able to see.
Prompt
I'm giving you an export from our ERP. Cross-reference it against MedScout and spot gaps.
Match my records to providers or sites in MedScout. Tell me the match rate and what didn't match before you analyze anything.
Then show me the highest-volume providers in my territory who aren't in my file at all, and flag anything in my file that looks low against what the claims data shows. Don't guess at ambiguous matches; put those in a separate list for me to review.
My file: [Attach it] What I'm trying to find: [Coverage gaps, stale accounts, competitive exposure]
Other ways to ask
- "Here's our billing data. Which of my target accounts haven't had a billing event yet?"
- "Cross-reference this account list against open payments and flag anyone taking money from [competitor]."
- "Here's our CRM export. Show me high-volume providers in my territory who aren't in it."
Getting better answers
- Be specific. Naming a territory, code set, or provider gets you a sharper answer than a broad question.
- You don't always need to know field names or filters. Describe what you want to the best of your ability.
- Feed it your own materials: collateral, branding, competitor's one-pager, your CRM export. The more context, the better the output.
- Save and share out what works with the rest of your team. If you keep asking the same kind of question, save the prompt so you don't have to back into it later.