How we work

It starts with a conversation. You describe the project, the question behind it, the decision it has to support, and where things feel stuck. We ask questions, mostly about context and constraints, and say honestly whether we think we can help.

If we are not the right fit, we will say so, and where we can, suggest what kind of help would suit you better. There is no obligation to continue after the call.

If it makes sense to work together, you receive a written proposal. It is short and specific, and it covers:

• The question or problem, in our own words, so you can check we understood it.
• What we will do, and what is explicitly outside the work.
• What you will receive, in what form.
• Milestones, check-in points and the overall timeline.
• The terms of the engagement, including fees, confidentiality and how changes are handled.

Nothing starts until you have agreed to the proposal. If the scope needs to change later, we write the change down and agree it before doing the work.

Larger pieces of work are split into milestones, each with something concrete to review: a design memo, a search strategy, a first set of results, a revised draft. At each milestone we meet briefly to look at what exists, settle open questions, and confirm the next step.

Between milestones you can expect short written updates at an agreed rhythm. If something unexpected turns up, such as a data problem or a finding that changes the question, we tell you promptly rather than at the end.

Final deliverables arrive in the formats agreed in the proposal, together with what you need to use them without us: documentation, code where relevant, a summary of the main decisions and their reasons, and notes on limitations.

We then walk you or your team through the work, answer questions, and make sure every file is where it should be. Handover is part of the job, not an extra.

Research continues after delivery. Reviewers respond, data arrives, plans change. You can arrange follow-up support, such as answering reviewer comments, rerunning an analysis on new data, or updating a review, either as a small new project or as part of a retainer.

• Plain language, in writing and in conversation.
• A clear written scope, and no work outside it without your agreement.
• Honest assessments, including when the answer is not what you hoped.
• Prompt notice of problems, delays or anything that affects the work.
• Deliverables you can use, understand and maintain on your own.
• Respect for your ownership of the ideas, the data and the final decisions.

• A named contact who can answer questions and make decisions.
• Access to the documents, data or systems the work requires, shared in the agreed way.
• Feedback at milestones within the agreed time, so the schedule holds.
• Early notice of changes to deadlines, priorities or the question itself.
• Responsibility for decisions that belong to you, your institution or your review board.

• Confidential by default. Ideas, drafts and data you share are treated as confidential, and we are happy to sign a reasonable confidentiality agreement.
• Least data necessary. We ask only for the data the work requires, and prefer de-identified data wherever possible.
• Agreed transfer and storage. How data is shared, where it is stored and who can access it are agreed in writing before anything moves.
• No AI tools without agreement. Your material is never put into AI tools unless you have agreed to it for a specific purpose.
• Deletion at the end. At the end of an engagement we return or delete your data as agreed, and confirm when it is done.
• Discretion about who we work with. We do not name clients or describe their projects publicly without written permission.