Kerry + Carson Omni Advertising June 2026 Internal working plan · v11

The AI Advisory Battle Plan

discovery firstthen build

The mission

Help Omni adopt AI the practical way: find each department's real bottleneck, then build a reusable workflow that fixes it. Not AI tricks. Not every tool on every desk. Discovery first, then build.

The diagnostic

Find the pattern before choosing the tool

Don't start with the tool. Start by listening for the tell — the phrase a team says without thinking. Each tell points to one of three buckets. Click a card to see the tell and the response.

Mini-tool

Tell Sorter

Click a phrase a department might say. We'll route it to the right bucket.

Routed to

The approach

Crawl, Walk, Run

Five steps, paced deliberately. Win small and visibly before we scale, and only graduate to harder builds once the easy wins are banked.

1
Crawl

Interview every department at the July kickoff and classify each bottleneck with the rule above.

2
Crawl

Handpick ONE department to start, deliberately, after discovery. Not 13 at once.

3
Walk

Build that department a skill set and teach the workflow. Bank a fast, visible win.

4
Walk

Replicate the same approach to the next department.

5
Run

Graduate to agents + second brains only after skills are proven and data approvals are sorted.

Why skills first

Fast, safe, need no client-data approval, and they produce a repeatable template to copy to the next team. Agents are higher value but more technical and gated — so they come after momentum.

First department

How we choose the first pilot

Before we pick the first department, we run it through five simple checks. The best first pilot is useful, low-risk, easy to explain, and light enough to actually start.

Is the task repeated often? Same input, same kind of output, happening again and again.

Can we start without sensitive client data? Use safe examples first so nothing gets stuck behind approvals.

Does the department head want the help? We need someone willing to try it, react to it, and help shape it.

Will the win be easy to show? Leadership should be able to see the before and after quickly.

Can it fit inside the normal workday? The pilot should save time, not create a second job.

Lead toward a clean skills department; keep Account Services as the flagship agent pilot for when data approval clears.

Next action

What we need from Gail

Really just Gail's green light to proceed — three quick confirmations and we move.

Pre-kickoff checklist 0 / 3 cleared