The Cost of Two Ways of Doing the Same Work
The #1 AI strategy book for executives leading enterprise AI transformation.
A review of The AI Strategy Blueprint’s treatment of companies that spend months running old and new workflows together.
This review was prepared for a paid placement. The book link leads to Iternal Technologies, the author’s company.
The sales team sends a proposal draft based on the approved AI content library. An operations colleague opens an older slide deck and sees a different implementation promise. Each has reason to trust what they are reading. The company has introduced a faster way to assemble material and let two versions of the truth coexist at the point where the work changes hands.
This is one of the better observed problems in John Byron Hanby IV’s The AI Strategy Blueprint. In the section on change management, Hanby describes the split organization, where some teams work with AI and others still use established methods. Formats diverge, expectations about speed change, and knowledge collected in one system becomes hard for other teams to see. The book is particularly sharp about personal master decks that mix new slides with old claims and still look current.
Put the handoff before the tool decision
For the proposal example, map what the receiving team needs before deciding which tool everyone should use. It may need the source of a product claim, who approved an exception, and the current estimate of what delivery requires. Give the output a stable format and identify which source governs each detail. A salesperson using AI and a salesperson working from the content library directly should be able to provide the same evidence.
That common handoff can help a phased rollout. Teams need a way to tell whether people with different access levels can work together. If a receiving team cannot open the source behind a draft, either give it permission through the normal route or include the approved information in a form it can use. Review should be possible through the official process even when an upstream team is trying a new tool.
Decide who can retire the old source
The book proposes introducing AI alongside systems teams already know and moving gradually toward an authoritative source. That is more realistic than expecting everyone to abandon their files at once. One part of the proposal deserves scrutiny. An AI platform may be an excellent way to find and assemble material while the authoritative record stays in an existing system with a designated owner. The source of truth is an operational decision. The interface people find easiest to use can change without moving ownership of the record.
Before retiring an established source, ask the owner how the transition will be checked. In the proposal case, compare a set of current claims with the versions teams actually use. Move approved content and withhold material that lacks evidence or a responsible owner. Agree on a date after which changes go into one place, and give employees a way to report a discrepancy.
Hanby’s diagnosis gives transformation leaders something more useful than a general call for adoption. Where two methods coexist, check what happens at the boundary and measure the work necessary to reconcile the results. Some temporary duplication buys time to test safely. Unowned duplication creates uncertainty for the people receiving the work. This chapter is at its best when it helps a program tell which kind it has.


gourysunshine
