The direct answer
When every employee invents their own AI process, you get duplicated work and inconsistent results. Same company, eleven different voices, eleven different ways to do the same task. The cost doesn't show up on an invoice. It shows up in quality, and in the hours spent redoing work that was almost right.
The fact: Microsoft found 78% of AI users bring their own AI to work. When people bring their own tools, the company has no quality control over the output.
SOURCE: Microsoft & LinkedIn, 2024 Work Trend Index — Read the report
The fact: Research on brand consistency found that 81% of companies still deal with off-brand content, while consistent branding can increase revenue by up to 33%.
SOURCE: Lucidpress, State of Brand Consistency (2019) — Read the study
Standardize what should repeat
You don't standardize judgment. You standardize the pattern. The repeated tasks, the ones with an input, a process, and an output, should run the same way every time, in the company's voice. That is what a play is: a workflow with the pattern built in.
Then leave room for the person's judgment on top. The draft is standard. The decision is theirs.
What the inconsistency actually costs
Two things. Time: every person solving the same problem again, from scratch. And trust: a client who gets two different voices from the same company starts to wonder who is actually paying attention.
Where to start
Find the three tasks with the most duplication across your team. Build one play for each. Run them for two weeks and compare the drafts to what the team was producing before. The numbers will make the case for you.