November 14, 2025
On the Guru Convention 2025, Mailjet’s Principal Product Manger, Natalie Lynch, introduced a ahead‑wanting session titled Smarter than the Inbox: AI for Actual E-mail Outcomes, exploring how Mannequin Context Protocol (MCP) is reshaping the way in which electronic mail entrepreneurs and senders work. Moderately than merely automating electronic mail sends or personalization, MCP permits entrepreneurs to shift from static dashboards towards conversational, query‑pushed entry to knowledge and marketing campaign efficiency.
The outcome: quicker insights, extra agility, and campaigns that really feel personalised – not simply “bespoke for segments” however “written for me”. The session targeted on what MCP connectors are, how they work in follow for electronic mail advertising and marketing, what pitfalls to look at, and methods to start with out over‑engineering.
Under you’ll discover the important thing learnings from the session, adopted by the complete recording and a FAQ that will help you and your staff discover MCP in your electronic mail‑advertising and marketing stack.
Smarter than the Inbox: full session recording
Catch the complete recording of Natalie’s session beneath:

Key takeaways from the session
- What’s an MCP connector?
Natalie described an MCP connector as AI that speaks your language – constructed for entrepreneurs, not machines. It bridges your electronic mail system, analytics, and AI assistant so you may ask plain‑language questions (e.g., “Which reactivation electronic mail carried out greatest this quarter?”) and get solutions instantly, as a substitute of digging by means of dashboards. - From dashboards to dialogue
Conventional electronic mail advertising and marketing workflows depend on charts, CSV exports and handbook evaluation of opens/clicks/unsubscribes. However with MCP, you start by asking a query and the connector fetches related knowledge, highlights efficiency, and surfaces the why behind what occurred. The analogy she used was “As a substitute of flipping a Rolodex you simply say “Name Mother”. - MCP on one actual workflow
Moderately than overhaul the whole lot, Natalie recommends beginning with one workflow (e.g., your month-to-month report deck) the place you exchange handbook steps with “ask the AI assistant by way of MCP” and let knowledge lead. That turns into your proof‑of‑idea. - Information high quality issues
In case your CRM/ESP knowledge is messy, the MCP will shine a lightweight on inconsistencies (and make them apparent). Entities like segments, tags, send-times have to be aligned and clear – in any other case the solutions shall be noisier. - Pitfalls and the place AI wants human oversight
MCP is a software, not a silver bullet. AI can confidently give flawed solutions; entrepreneurs should interrogate the outcomes and guarantee good questioning. Additionally, over‑personalization can backfire. Bear in mind the distinction between “We thought you’d like this since you purchased sneakers” (good) and “We noticed you stroll previous our retailer at 3:12 yesterday” (a tad creepy). - Velocity‑to‑perception turns into aggressive benefit
The worth of MCP connectors isn’t in radically new metrics, however in making perception technology a lot quicker – so you may iterate, experiment and optimize on the tempo of curiosity moderately than the tempo of reporting cycles.
You may see how it works in Mailjet beneath:


FAQ: What electronic mail entrepreneurs ask about MCP
1. What’s a easy technique to get began with MCP?
The best technique to begin is to attach a conversational AI software (like Claude or ChatGPT) to an MCP-compatible server – such because the open-source one Mailjet now supplies. From there, you may ask plain-language questions and retrieve reside knowledge out of your Mailjet account. You could find extra data on our just lately revealed article “MCP: A brand new period for electronic mail advertising and marketing knowledge” the place you get a deeper perception into distinction use circumstances and methods to set it up. If you’re much more curious then try our Mailjet documentation on implementing MCP.
A great first step? Change a single reporting process – like constructing your month-to-month efficiency deck – with a easy immediate:
“Evaluate the open and click on charges of our March and April campaigns. What drove the distinction?”
That’s your proof of idea.
2. Is there a price related to utilizing an MCP Connector?
Mailjet’s MCP Server is open supply and free to make use of. The price comes right down to your AI assistant of selection and the way you deploy it. For those who’re already utilizing instruments like Claude or OpenAI’s ChatGPT Professional, integration could also be easy and low-cost – particularly for read-only use circumstances. For groups scaling throughout departments or use circumstances, you could need to put money into customized internet hosting, API rate-limit changes, or premium AI instruments for broader entry and stability.
3. How is proprietary knowledge stored protected when utilizing MCP?
Safety is a high precedence. Mailjet’s MCP Server provides read-only entry by default, which means your AI assistant can retrieve knowledge however not modify it. That’s intentional so entrepreneurs can discover insights freely with out risking adjustments to manufacturing knowledge.
For builders, the server is absolutely extendable. You may outline precisely which Mailjet API endpoints are uncovered and implement strict entry controls, monitoring, and encryption. It’s a managed, clear setup designed with security and compliance in thoughts.
Unlock your marketing campaign insights with AI now
The Mailjet MCP Server redefines electronic mail analytics by connecting your Mailjet platform to the ability of conversational AI. It’s designed for each advertising and marketing and growth groups who need to unlock deep, actionable insights from their electronic mail knowledge just by asking questions.
Prepared to start out the dialog along with your knowledge?
Hold me posted! Get nice assets in your inbox each week.
Ship me the Sinch publication.

