Table of contents
The questions teams want answered, and what gets in the way
Every team we talk to has a running list of questions they wish they could get fast, reliable answers to. What changed in our performance last month and why. Which clients are showing the early signs of churn. Which channels are actually pulling weight and which ones are quietly burning budget. The pull toward AI for this kind of work is obvious. The answers should be a question away.
In practice, getting there is harder than it looks. Your data has to be in one place, your metrics defined the same way every time, and the math behind the answer handled correctly rather than estimated by a language model on the fly. On top of that, somebody with the judgment to know which questions matter, where to look, and what counts as a real signal has to be the one shaping the analysis in the first place.
That last part is the one most teams underestimate. The data and metric work is technical and can be set up once. The judgment of a senior analyst is what turns the output into something a team can actually act on, and it is the thing AI is least equipped to bring on its own.
That work has to happen once per question, and it is the reason most teams get one good answer out of AI and then stall. The setup cost has to be paid over and over again, and at some point it is faster to just keep doing things the old way.
A skill is that foundation, packaged
The Skills Marketplace is a library of those answers, already worked out. Each skill in the library takes one of those running-list questions and packages everything that goes into answering it well as a workflow your team can run on demand.
Each skill is built on two things. The first is the Databox foundation: your data centralized in one place, your metrics defined once and used consistently, and the math behind the answer handled by Databox rather than improvised by a model. The Databox MCP is what brings that whole foundation into the AI tool your team is working in, so the skill is running on trusted, defined data instead of whatever the model could find on its own. The second is the expert judgment of which questions to ask, where to look, and what counts as a real signal versus noise. That judgment is built into the skill itself.
Why that combination matters? It changes who can actually use it. Without a skill, getting a reliable answer to a question like “which clients are showing early signs of churn” requires somebody on your team who already knows which signals matter most, in what order, on what time horizon, with what thresholds, working from data they trust. With a skill, all of that is already in the workflow. Anyone on your team can run the analysis and get back the kind of answer a senior analyst would have produced if they had the time to dig into the question themselves.
To put a skill to work, you connect your Databox account to Claude or n8n once through the Databox MCP, which is available on paid plans and our 14-day free trial. From then on, any skill in the library is available to install and run.
Where the best AI analyses come from
The reason we built an analytics skill marketplace is that the most valuable work being done with performance data is not the work that gets designed in our product meetings. It is the work being done by the senior analysts, marketers, sales leaders, agencies, and consultants who have already figured out which questions matter and how to answer them well. And in the last year, it is also the work being done by the people who have figured out how to actually get reliable analysis out of AI itself: which prompts to use, which guardrails to put in place, where the model needs a hand, and where it can run on its own.
Most of that judgment is locked inside the few people who know how, with no clean way for it to reach the rest of the team, let alone other teams running similar businesses. A marketplace is how that work gets surfaced and shared. We are seeding it with skills built by experts on our team and by partners who work with performance data every day.
How to start using the Skills Marketplace
Browse the library and see what is there. To install and run any skill, you will need a Databox account on a paid plan and the Databox MCP connected to Claude or n8n. Every skill comes with its own setup guide and onboarding flow, which walks you through connecting the right data and adds context the skill needs to give you a good answer. Once that is in place, any skill in the marketplace is yours to put to work.
Skills also improve with use. As you run a skill, you can update it with new context, recent changes, or specifics that matter to your team, so each use gets a little more tailored to how your business actually works.
If you have already built something useful for your own team and you think other teams would benefit, submit it to the library.
