Field notes from the implementation leg.

Most rollouts are decided before go-live. Ten things a group should have established before the first location goes live, almost none of which have anything to do with the software.

Notes from the Overjet AI Summit in Scottsdale, where the question shifted from what AI can do to how DSOs get it working in their offices. Dental AI does not have a product problem anymore.

Plenty of leaders commit to a platform this size on a conference pitch, a Facebook group recommendation, or one ninety-minute demo. Two years later the rollout is still unfinished and every other initiative has stopped.

Many dental technology vendors are not ready for enterprise implementation, and it is not just early stage companies. Organizations with hundreds of customers still treat a 40-location rollout as the single-practice process, run 40 times.

A thread of DSO operators recommending software to each other, with no discussion of what else was running in the organization, what outcome they were chasing, or what workflow the tool was supposed to change.

Vendors overcharging by 15 to 25 percent through contract drift. Groups using 40 to 60 percent of the features they pay for. Duplicate functionality across five to seven products. The clutter is real, and so is the opportunity.

A DSO signs the contract, the vendor runs onboarding, go-live happens on schedule — and six months later nobody's using it. It isn't a vendor problem or a DSO problem. It's the gap between them.

Vendors win the sale and lose the customer at deployment. After hundreds of rollouts, the failure points cluster in five predictable places — and none of them are about the software.

Your champion isn't your buyer. By the time the rest of the committee shows up, the deal you thought you had is one you're negotiating from scratch.

When you acquire your way to scale, you inherit your sellers' technology decisions. Standardizing the stack post-close is harder, slower, and more expensive than the deal model assumed.