AI adoption
Our enthusiasm is not a good gauge of business value
We're all enthusiastic about what we are achieving with AI, but our enthusiasm is far from a good gauge of business value.
We’re all enthusiastic about what we are achieving with AI, but our enthusiasm is far from a good gauge of business value.
Most of us can empathise with the successes we’re personally experiencing with AI and its impact on our productivity. But that’s not translating into organisational processes and productivity in the same way.
Gains that do not scale
Superficially, that seems confusing. Why aren’t we seeing those individual benefits mirrored across our organisations?
For me, it comes down to one key consideration: experimental pilots don’t naturally convert into successful business processes. There’s a disconnect between what we’re doing individually and what we need to do as organisations.
Pilots are cheap now
Creating pilots is now a low-effort bar. Building and writing code used to be where the friction was. That challenge has moved upstreamI made this case in agencies sell judgment, not delivery: selecting the right thing to build, selecting the right thing to scale. That takes judgment. We must all move the conversation from ‘could we’ to ‘should we’.
Building the thing right is now becoming easier with AI, so we need to focus on building the right thing.
Why pilots fail
There are a few reasons we see pilots fail. They are often treated like side projects, lacking the required organisational support, hours and funding.
There’s often a pre-baked assumption that they will scale. When they don’t, the whole idea is rejected rather than narrowing in on the scaling challenge.
What’s got us here, in terms of our individual AI productivity, isn’t going to get us to successful organisation-wide AI adoption.
There is a chasm between individual AI productivity and organisational adoption. You can’t jump that chasm. You need to build a bridge, one compounding section at a time.