What the audit produces
A ranked list of tasks, by time spent, by role. The list is the backlog for any AI rollout, automation project, or reorg conversation. Without it, teams buy platforms hoping the platform will pick the right job. Almost always wrong.
The 30 minute workflow
Set the time window
The last 10 working days. Two weeks of recency, before memory fades.
Distribute the template
A simple Google Sheet. Columns. Date, time bucket of 30 minutes, task name, tool used, output type. Each person fills theirs in 15 to 20 minutes.
Roll up the totals
Pivot by task name. Sort by total hours. The top three per role are the automation candidates.
Tag automation potential
For each top task, mark High, Medium, or Low. High means template plus AI can do 80 percent. Medium means AI assists but human stays in loop. Low means judgement call, do not automate.
The three patterns that always show up
- Format and reformat. Briefs into decks, decks into emails, emails into briefs. AI handles this cleanly with a template per direction.
- First draft of repetitive copy.Captions, subject lines, ad variants. AI drafts, human edits, output ships twice as fast.
- Data wrangling. Spreadsheet joins, CRM hygiene, cross tool reporting. n8n and AI middleware kill this category almost entirely.
What the audit catches that nothing else does
- The senior who is doing the junior's formatting work in secret.
- The tool nobody uses but everybody pays for.
- The meeting that consumes 15 percent of the team's week with no decisions.
- The handoff between two roles that always loses 30 minutes.
What to do with the output
Top three high potential tasks per role become the automation backlog. Build the first three prompt templates or n8n workflows in the next two weeks. Re run the audit 30 days later. Reclaimed hours become the metric for the rollout, not adoption rate.
Quick answers
- Will the team push back on logging hours?
- Usually yes. Reframe it as input to an AI rollout that frees their time. Make the rollup transparent, share the reclaimed hours number publicly, and the second audit becomes much easier to run.
- How accurate does the logging need to be?
- Good enough. 30 minute buckets, not 5 minute precision. The pattern matters more than the exact numbers.
- Should managers also do the audit?
- Yes. Manager time is often the most fragmented and the most automatable. Their audit usually reveals the cleanest wins.
- How often should I rerun the audit?
- Every six months in stable teams. Every quarter in fast scaling teams. After any major reorg.
- What if the audit shows no automation candidates?
- Rare but possible. The team is either small enough that AI tooling is not cost effective, or the work is genuinely not template shaped. Either way, the audit saved you from buying tools that would not pay back.