When AI Knows Your Market Data: MCP in the OTC Market

Monday, 8:45 a.m. I have a meeting with a major mail-order pharmacy at 9 a.m. The question is predictable: Why has our product slipped in the rankings? The data is there. It’s in the dashboard, in the export, and in last month’s Excel file.

Just not where you’re working right now. This is exactly where MCP comes in. Claude retrieves the answer himself—including the current status and the trends behind it—verified and delivered in seconds.

What is MCP?

MCP stands for Model Context Protocol. It’s an open standard that connects AI assistants to data sources and software.

Think of it this way: USB-C for AI. One connector, many devices. Instead of building a separate interface for every AI and every system, they all speak the same language.

Without MCP, an AI knows only two things: what’s in its training data and what you feed it. When it comes to prices, PZN, or rankings, this often means it’s just guessing. With MCP, it checks with the source.

Three key components are all you need to understand it:

→ MCP Client: Your AI assistant, such as Claude.

→ MCP Server: The bridge to the data source. It tells the AI which tools are available, such as “Check price” or “Get price trend.”

→ Data Source: Database, CRM, PIM, or market data.

MCP is no longer a niche topic. Anthropic, the company behind Claude, released the standard in late 2024. Since late 2025, it has been managed by the Linux Foundation’s Agentic AI Foundation, with support from OpenAI, Google, Microsoft, and AWS. ChatGPT, Gemini, and Microsoft Copilot support MCP. More than 10,000 MCP servers are already running publicly. (IT Brief).

It's especially easy in Claude. There, you can integrate an MCP server as a “custom connector”—in the browser, in the desktop app, and in the mobile apps—across all Claude plans. In Claude Team and Enterprise, an owner sets up the connector once for the entire team. (Claude Help Center).

Important for users: You don't need to write any code. You integrate an MCP server just like you would an app. Enter the address, share it, and you're done.

What Problems MCP Solves in the OTC and CHC Markets

The CHC market isn't lacking data. What it lacks is quick access at the right moment—and insight into trends. Prices at online pharmacies change constantly, while reports are issued monthly. MCP bridges that gap.

Problem	Today	With MCP
Daten in Silos	Prices in the Digital Shelf Tool, master data in PIM, sales data in ERP. Every query means opening multiple systems.	The AI queries multiple connected sources in a single interaction.
Speed in Online Retail	The monthly report shows yesterday's status. Competitors' price promotions are noticed too late.	Current status plus trends over weeks and months all in seconds.
Client Meetings in Key Account	Preparation using screenshots, exports, and Excel.	A single query provides price, availability, and ranking for each client.
Ad hoc Questions from Management	“Why are sales at Shop X declining?” It takes days to analyze the causes.	Check price, availability, and ranking all at once.
Few Dashboard Experts	Only part of the team knows how to use the tools. Everyone else has to wait.	Questions in everyday language. For every role.
Copy and Paste in Emails and Reports	Numbers are typed in. Errors creep in.	The figures come directly from the source, as of the date indicated.
AI Without Facts	At best, ChatGPT and similar tools provide today's prices. They lack historical trends and fill in the gaps with guesses.	The AI uses verified market data rather than assumptions.

One thing remains important, however:

→ Plausibility: Even a well-integrated AI can make mistakes. Double-check figures for customers, especially the data status.

OTC-Cube Pulse: MCP for the CHC Market

Pulse is OTC-Cube’s MCP server. It connects Claude directly to the OTC-Cube database: prices, availability, and rankings from the German online pharmacy market. No need to open a dashboard, no need to export data, no need for Excel. Not just today’s price, but price trends—verified and available in seconds.

A Real-World Example

First question in Claude: “How do the prices for Norsan Omega-3 Vegan 100 ml compare at Shop-Apotheke and medpex?”

Claude first searches for the PZN via Pulse, then for the current prices. The answer:[MM1

Shop	Price	Available	Rating	Stand
medpex	24,99 €	yes	4,84	16.09.2026
Shop-Apotheke	26,49 €	yes	4,8	17.09.2026

Here’s the context: medpex is €1.50 cheaper than Shop-Apotheke. The MSRP for both is €29.00.

Second question: “Show me the price trend over the last 6 months as a line chart.”

Claude provides the conclusion right away: In the summer, the difference was a good €2; currently, it’s €1.50. Two questions, less than a minute. That’s exactly what a web-based AI doesn’t provide: a verified trend over six months.

Why Pulse Can Do More Than General AI

Web-based AI systems might, with a bit of luck, find today’s price. That’s not enough for decision-making in the CHC market.

→ History: Verified time series instead of a snapshot. You see the trend, not just the current status.

→ Data quality: Validated, cleaned data. The AI works with facts instead of assumptions.

→ Completeness and transparency: Clearly defined coverage: 30 stores in Germany, 7,000 PZN codes, and historical data dating back to [MM2] 2022. Data status indicated for each value. You know what you’re getting—and what you’re not.

→ Structure and Speed: Accurate PZN mapping across all pharmacies and ready-to-use query tools. No lengthy web searches, no waiting.

→ Comparability: Consistent methodology across all pharmacies and time periods: package sizes, base prices, rankings.

→ Reproducibility: The same query yields the same numbers because they come from the same database. Important for reports and client meetings.

What can be achieved with Pulse MCP:

→ Prices and availability: Find products by PZN, brand, or name. Current prices per pharmacy, price trends, historical low and high prices, out-of-stock frequency.

→ Visibility in the online store: Category and keyword rankings, current or historical. Which keywords perform best for each online store.

→ Customer view: Watchlist, keywords, and competitors per customer. Your own prices and rankings displayed directly next to those of the competition.

→ Data quality: Verify product image URLs, check data status, and identify anomalies such as price spikes or gaps.

And many more use cases.

Pulse provides exclusively market data: prices, availability, rankings, reviews, and product information.

Pulse nutzt dieselben Daten wie das OTC-Cube Digital Shelf.

Fazit

Die Frage ist nicht mehr, ob KI im CHC-Vertrieb ankommt. Die Frage ist, ob sie mit Ihren Daten arbeitet oder ohne. Mit Pulse und Claude arbeitet sie mit.

Für die meisten Teams ist schon der erste Schritt ein großer: wissen, wo man steht, und Zusammenhänge erkennen. Der nächste Schritt zeichnet sich bereits ab: Pulse mit eigenen Daten kombinieren, etwa mit Sell-out-Zahlen, in einem Gespräch.