Ecommerce projects rarely belong to one team anymore. A replatform or new storefront pulls in ecommerce, marketing, merchandising, operations and IT, plus agencies, technology partners and a growing set of AI tools. Everyone shares the same goal, but not always the same way of working. A short set of cross-team guiding principles is one of the simplest ways to turn a group of organizations into one team.
Why projects are more multi-disciplinary than ever
Modern commerce stacks are built from specialized parts. A composable commerce build might involve a commerce platform, a CMS, search, payments, an ERP and an integration layer, each with its own vendor and experts. On the client side, the work touches brand, customer service, finance, retail and B2B sales. The more people involved, the more ways there are to misunderstand each other.
AI adds another layer. Designers prototype in code, developers work alongside AI coding assistants, marketers produce content at scale, and AI agents are starting to act on the storefront itself. Roles blur and work moves faster. That speed is a real advantage, but only when everyone is pointed in the same direction. Without shared agreements, AI simply helps each team move faster on its own assumptions.
What guiding principles are, and what they are not
Guiding principles are a handful of behavioural agreements that describe how the combined team will work, communicate and make decisions. They are not a project plan, a scope document or a list of values nobody reads. The best ones are short, specific and written together by the people doing the work.
We first tried this more than a decade ago, when a client asked us and their brand agency to write guiding principles with them at kickoff, effectively unifying three organizations into one team. It worked so well that it has shaped how we start projects ever since.
How to run a guiding principles session
- Invite everyone who will shape the work. Include client stakeholders, agency teams, technology partners and the people responsible for data and AI tools.
- Start with purpose and outcomes. Agree on why the project exists and what success looks like before talking about behaviour.
- Draft in small, mixed groups. Give each group about 20 minutes to write principles in a few words each. Keep it loose and avoid getting stuck on wording.
- Reconvene and merge. The ideas are usually very similar across groups. The hard work is merging them into a concise, unified set.
- Keep it short and actionable. Each principle should describe something people can act on and check. A useful test is whether reading them aloud at the start of every meeting would feel too long.
- Put them where the work happens. Pin them in your project tool and shared docs, and add them to the instructions of any shared AI assistants so generated work follows the same rules as everything else.
Guiding principles for the age of AI
Every team should write its own, but here is an example set that blends principles from our first session with agreements that matter more now.
- Communicate early and openly. Share work in progress, raise risks as soon as they appear and never let a question sit between teams.
- Keep the end customer in mind at all times. Every decision should make the experience better for the people who buy from us.
- Test assumptions before we scale them. Prototype, measure and learn before committing budget or rolling changes out everywhere.
- Build on each other's strengths. Let each discipline lead where it knows best, and give credit freely.
- People own the decisions. AI can draft, summarize and suggest, but a named person is accountable for every decision and every release.
- Be transparent about AI use. Say when work was generated or assisted by AI and review it with the same care as any other work. Our own AI disclosure explains how we approach this.
- Protect customer and client data. Agree which AI tools are approved and what information can and cannot go into them.
- Recognize progress. Celebrate milestones and the people behind them.
- Look after each other. Care for the wellbeing of everyone on the team, across every organization.
Questions to get you started
Use these with your own team, a partner or a client to shape your principles.
- What culture do we want to create together?
- What can we count on from each other?
- How will we make decisions, and who has the final say?
- How will we respond when things get hard or part of the project goes off track?
- Where will AI help this project, and where do we need human judgment?
- Which tools and data are approved for AI use?
- How will we measure success, and how will we celebrate it?
- What rhythms keep us accountable, such as standups, demos and retrospectives?
Revisit them as the project evolves
Guiding principles are not something you write once and forget. A few weeks into our first project, we asked everyone how the work was tracking against the principles, and the feedback was clear. Communication across three companies was stronger, ideas built on each other, and the project carried less anxiety than past attempts. Check in the same way during retrospectives, and update the principles as the team, the tools and the scope change.
This is how we start every engagement, from strategy and roadmapping to a full ecommerce replatform. To see why alignment matters even more as AI agents enter commerce, read why AI agents need a composable commerce architecture. You can also see our work or get in touch.
