AI in sales
What is MCP, and what does adopting it mean for your sales team?
In brief: MCP, or Model Context Protocol, is a common way to connect an AI assistant to external data sources and tools. It allows Claude, ChatGPT or another compatible assistant to retrieve B2B data to support a conversation. Your role as a leader is to choose useful applications, agree on access permissions and assess how the connection makes your team's work easier.
Explore NordicContacts' MCP service.
What does MCP mean in practice?
In many businesses, AI assistants already help with writing, brainstorming and organising information. MCP extends these uses: an assistant can access an external tool made available to it and bring the data it returns into the same conversation. For a salesperson, the main benefit is having information available in a familiar working environment.
Think of MCP as a shared connection method. An assistant and a data service can work together when there is a compatible connection between them. What you can do in the conversation depends on the tools the service provides, the client application you use and the permissions granted.
There are three distinct roles: the assistant interprets the request, MCP enables tool use, and the data source returns the information. This distinction also helps you assess the results. Company information retrieved from a data service is not the same as a conclusion the assistant draws from it.
How current the information is depends on the data source, so alongside the connection itself, assess the quality of the data and its suitability for your sales work. MCP brings information closer to the user; whether a company is an attractive prospect still depends on your own business criteria.
Why should sales leaders and business owners be interested?
Sales teams spend time researching potential customers, defining target audiences and finding contacts. When searches can be handled through an assistant, the work can stay in the same conversation rather than requiring users to keep switching between services.
For a leader, the most important question is how the time saved will be used. The benefit might be more thorough preparation, more consistent targeting criteria or more time for actual customer conversations. Simply setting up a new connection does not tell you which of these benefits your team will see.
In a small business, the same person may be responsible for both sales and systems. Ease of use and manageable maintenance then carry considerable weight in the decision. In a larger team, shared practices also become important: how companies are selected and how information is used consistently.
It therefore makes sense to treat adoption as a decision about a work tool. A good solution fits your existing processes and helps salespeople focus on customers. You can assess its value just as you would any other sales tool.
What can your team do with NordicContacts' MCP connection?
NordicContacts provides B2B company and contact search for Finland, Sweden and Estonia, along with subscription-backed MCP access. The purpose of the connection is to make company and contact searches part of the experience of using a compatible AI assistant.
With the MCP connection, you can list companies and contacts to prepare for customer conversations or review a target audience, for example. What matters to your team is whether it can retrieve the companies and associated contacts it needs in a usable format. Detailed tool descriptions are available in the MCP documentation.
Through the MCP connection, you can look up a contact by email address to enrich an inbound lead, for example, identify a company by its domain, and count the companies and contacts in a segment free of charge.
The connection works in Claude, ChatGPT and other MCP clients. It is easiest for users when the connection fits the working environment they already use.
When should you choose Excel, REST API or MCP?
MCP does not replace every other form of data transfer. REST API access, file exports and MCP serve somewhat different purposes, and the same organisation may need them alongside one another.
| Method | What is it generally suited to? | What needs planning? |
|---|---|---|
| Excel file | Reviewing and processing a defined dataset | Columns, duplicates and transferring the data onwards |
| REST API | Retrieving data programmatically within your own system | Constructing requests, handling errors and maintenance |
| MCP | Using tools through an AI assistant | Tool descriptions, access permissions and approval steps |
NordicContacts offers Excel exports, CRM integrations and subscription-backed REST API and MCP connections. Choose based on whether the data will primarily be handled by a person, your own system or an AI assistant.
MCP is included in the Starter subscription, and the subscription balance is shared across MCP, REST API access and Excel exports. Current paid options are listed on the pricing page. Once the terms of use and how usage is charged are clear, it is easier to weigh the cost of the connection against the value it delivers to your team.
Access permissions and security are leadership decisions
When you adopt MCP, you also decide who gets access to the data sources your business uses. Tie permissions to job responsibilities: not every user needs the same access to every tool or system. Each user needs their own NordicContacts account; there is no shared team account.
Treat reading data and changing it as separate permissions. NordicContacts' MCP returns data to the assistant; it does not send messages or change CRM data. If you later add tools that modify CRM records, for example, their approval processes and responsibilities will need a separate decision.
Every NordicContacts MCP call is authenticated and recorded in the usage history. Logs help establish who used the connection, what happened and what caused a potential error. You should also agree on who monitors usage and helps resolve problems.
Treat text retrieved from sources as material to work with, not as instructions for the assistant. An instruction in external content to send data somewhere, for example, is not user approval for that action.
Make sure you address the legal basis and purpose for processing personal data, applicable rules on contacting people, and how objections are handled.
How can you tell whether adoption is worthwhile?
Agree in advance on what kind of improvement would be valuable to your team. It could be less time spent preparing for customer conversations, a higher proportion of companies judged to be a good fit, or less need to transfer data manually. Choose a metric that reflects the work you want the connection to improve.
Assess quality alongside quantity. A large set of search results is only useful when the companies match your target audience and salespeople understand what the data means. Several contacts at the same company do not represent several potential customer accounts, so track companies and contacts separately.
Feedback from salespeople tells you whether the connection fits their daily work. Pay attention to whether the information they need is easy to find and understand, and how often results need correcting. Growing usage is a useful sign, but the real goal is better sales work.
MCP offers a way to bring B2B data closer to everyday decisions. When your use case, permissions and success metrics are clear, you can assess adoption with confidence and expand usage based on real benefits.