Should you build your own AI or use native tools?
"Should we build our own AI or use the native tool?" It's a fair question in almost every AI conversation right now.
Two questions can matter even more: How will this investment help us stand apart? And is that what we want our brand to be known for?
Native tools (e.g., from Epic or SAP) tend to bring organizations to parity, which is often what's needed, at least at first. Differentiation usually comes from what teams pair the tool with: their own data, their workflows, their people.
A helpful lens is mapping AI investments on two axes: where the tool comes from (platform-native vs. custom-built) and how much differentiation it's likely to create 👇
Platform parity: the native tool is often enough
HR, finance, IT service, everyday productivity. "Good enough" can be the right goal here, and heavy customization may add cost without adding much edge.
Applied configuration: same tools, different use
Standard platform AI, fed with the organization's own data and workflows. A useful question: is the edge in how the tool is used, and is it documented well enough to outlast team changes?
Built moat: custom build, real stakes
Usually a small number of areas where an organization truly competes. These often benefit from being funded and governed as strategic assets.
Fragmented effort: custom build, limited edge
This happens for understandable reasons: several teams solve similar problems in parallel, with good intentions. Consolidating can free up resources for the top half of the grid.
🚦 Before placing anything on the grid, a few gates come first: security and privacy, regulatory requirements, the impact of an error, and responsible AI. Ideas that don't clear one usually pause there.
⚖️ Then three lenses help weigh what remains: value (is there a named KPI?), feasibility (data, workflow fit, capacity to sustain) and durability (will the position hold as vendor roadmaps evolve?).
🏥 Durability shows up often in healthcare: A health system builds its own ambient documentation tool, and a couple of years later the EHR vendor releases a native version. A potential differentiator drifts toward duplicated effort. That isn't anyone's misstep; the market simply moved.
Another system takes the native tool, tunes it to each specialty's workflow and supports physicians as they adopt it. Same tool, but how it's applied makes a real difference.
💡 What many clients find most useful about the grid isn't where an initiative sits today, but which way it may be heading.
Context varies, but it can help leaders, especially those leading brand and experience, to ask whether each investment connects to a clear differentiation thesis.
Where would your organization's biggest AI investment sit today? Has it shifted over time? Is it driving your differentiation?
We would love to have the conversation.
Comments