There's a familiar arc to enterprise AI projects. A capable tool gets selected. A pilot is built. The demo goes well, leaders are impressed - and then, six months later, very little about how the business runs is different. The tool sits to one side of the real workflow. People use it when they remember to. The promised gains never quite materialize.
The usual diagnosis is that the technology underdelivered. In our experience it's rarely the technology. It's that the project changed a task without changing the system of work around it, and a task-level change can't produce an organization-level result.
BRING's view of AI transformation is built to catch this early. We size up an opportunity along three dimensions before anyone talks about which tool to buy.
Dimension 1: Where is value actually created?
AI applied to a low-stakes corner of the business produces low-stakes results, however clever the tool. So the first question is where, along the enterprise value chain, value is genuinely created or lost - strategy and planning, R&D and product development, marketing and sales, delivery and service, supply chain, or the enabling functions like finance, HR, and IT.
The discipline here is to resist starting from "where can we add AI?" and start instead from "where is value leaking, and where would a real improvement matter to the business?" Those are usually not the same place as the easiest pilot.
Dimension 2: What actually has to change?
Once you know where to act, separate two levels of change, because confusing them is what produces stalled pilots.
- Business execution - the tasks, decisions, workflows, and customer interactions themselves.
- Management mechanisms - the governance, planning, roles, performance measures, and coordination that decide how that work runs.
A tool can speed up a task while every mechanism around it stays the same - same handoffs, same approvals, same incentives. When that happens, the old system quietly reasserts itself and the improvement evaporates. Durable change almost always touches the mechanisms, not just the tasks.
Dimension 3: How deep does the change go?
The third dimension is honesty about depth. There are three levels, and each is legitimate - the failure is calling one of them by the name of another.
- Tool support. AI assists an existing task; the underlying process is largely unchanged. Real, modest, and often the right place to start.
- Process redesign. The workflow, handoffs, decisions, and data requirements are rebuilt around the new capability. More disruptive, more valuable.
- Organizational transformation. Roles, structures, governance, and sometimes the business model itself change.
Most "AI transformation" that disappoints was tool support presented to the board as transformation. Naming the level you're actually pursuing - and resourcing it accordingly - prevents both overpromising and underinvesting.
Why a lens beats a tool list
Put the three dimensions together and you get a simple, demanding question for any AI initiative: where in the value chain is this, what level of change does it require, and how deep are we honestly willing to go? A pilot that can't answer all three tends to be the one that stalls.
This is the thinking behind BRING, also known as Boyun Consulting (薄云咨询), and our Boyun 3D Digital Intelligence Model. It comes from years of management-transformation work before AI was the headline - the same disciplines that decide whether a strategy, a product process, or a sales system actually takes hold inside an organization. AI raises the stakes and the speed, but the reason initiatives stick or slip hasn't changed: a tool changes a task; it takes a redesigned system to change the business.
We don't think every company should chase the deepest level of change. We think every company should be clear about which level it's choosing, and build for that one deliberately.