The pattern we keep seeing
Almost every enterprise digital program starts the same way. Week one is kickoff. Week two is stakeholder mapping. By the end of week two there is a RACI chart in the project workspace — responsible, accountable, consulted, informed — laid out neatly across rows of workstreams and columns of stakeholders.
By week six, the team has stopped looking at it.
This is not a complaint about RACI as a framework. RACI is fine. It does what it is supposed to do: force a conversation about who decides what before the work starts. The problem is that the conversation happens in week two and never happens again, even though the answers it produced should be updated every time the engagement structure shifts. And in digital engineering work, the engagement structure shifts every three to four weeks.
So the RACI on the wall and the RACI the team operates from drift apart. The team adapts informally — someone on the client side becomes the de-facto decision-maker on a workstream they were originally only "consulted" on; a vendor team member gets pulled into approvals they were "informed" of. None of this is wrong. It is how good teams adapt. But it means the document on the wall is now lying.
When the next person joins the engagement and asks "who decides what," the answer they get from the document is different from the answer they get from anyone in the working group. They learn quickly to ignore the document. Six months in, when something goes sideways and leadership asks "who was supposed to own this," the document is no help because nobody trusts it anymore.
Why the document becomes a lie
This is not a stationery problem. It is not solved by better templates, RACI software, or training videos. The deeper failure is that most engagements treat the RACI as a planning artifact — something you build once at the start of the program and then file away — when it should be treated as an operating cadence.
A planning artifact is reviewed when you make it and never again. An operating cadence is reviewed on a defined frequency, by named people, with consequences for staleness. The difference is structural.
The reason most teams default to the planning-artifact pattern is that nobody owns the cadence. The PMO produces the RACI, the engagement lead approves it, the team agrees it looks right, and then there is no recurring meeting on anyone's calendar to revisit it. Without that meeting, the document drifts the moment the team has its first informal workaround. And the first informal workaround usually happens in week three.
What our team does instead
Across engagements over the past two decades, our team has converged on three rituals that turn a RACI into a working artifact. None of them are about the RACI itself. They are about the cadence of how it gets used.
Ritual one: the named owner. One person on the engagement owns keeping the RACI current. Not the engagement lead — they are too busy. Not the PMO — they are too far from the working group to know when a decision-rights drift has happened. Our team usually assigns this to a senior pod member who is in the day-to-day, has authority to call out drift when they see it, and has 30 minutes a week protected on their calendar to do exactly that.
Ritual two: the four-week refresh. Every four weeks, the named owner walks the RACI with the engagement lead and the client counterpart. The conversation is short — usually 20 minutes. The question is: has anything changed about who decides what? Most of the time the answer is no and the document stands. When the answer is yes, the change happens in that meeting, not in a follow-up. The document gets updated before anyone leaves.
Ritual three: the onboarding handoff. When a new team member joins the engagement — and on a 12-month program, that happens five or six times — the named owner does a 15-minute walkthrough of the RACI with them as part of onboarding. Not "read this document," a real walkthrough where the new person can ask "what does this row actually mean in practice." This is the ritual that catches the most drift, because new people are the ones most willing to ask "wait, is that really how decisions get made?"
None of this is revolutionary. None of it requires new tools. What it requires is the engagement lead deciding, in week one, that the RACI is going to be an operating cadence rather than a planning artifact, and then protecting the time on someone's calendar to make that real.
The deeper pattern
The RACI is a small example of a larger pattern that shows up across digital engagements. Most planning artifacts are produced under the assumption that the plan is stable and the work executes against it. In digital engineering work, this assumption is almost always wrong.
The work is exploratory by nature. New constraints appear. Stakeholders change. The org chart shifts. Vendors miss deadlines. Customers ask for things nobody anticipated. Every one of these events invalidates some part of the plan. The plan should be updated. It usually is not, because nobody owns the cadence of updating it.
The same diagnosis applies to roadmaps, success metrics, escalation paths, integration architectures, and decision-rights documents like RACI. All of them are useful when treated as operating cadences. All of them go stale when treated as planning artifacts. The difference is not in the document. The difference is in whether someone owns the cadence.
If you are looking at a RACI on the wall right now and wondering whether it is current, the question to ask is not "is the document right." The question to ask is: who owns keeping it current, and when did they last walk it with the engagement lead? If the answer is "nobody" or "I don't know," the document has already drifted. The next step is not editing the document. The next step is naming an owner.