Amazon Tracks AI Adoption, Finance Should Too
Amazon monitors every engineer's AI usage tied to productivity goals. Here's what finance leaders can steal from that playbook before their competitors do.
Amazon Is Measuring AI Adoption by Employee — Your Firm Should Be Doing the Same
Business Insider reports that Amazon tracks engineers' AI tool usage in granular detail and ties adoption rates directly to productivity benchmarks — even as some employees push back. This isn't a soft culture initiative; it's operational measurement with teeth. For accounting firms, the implication is direct: if you're not tracking which staff are actually using AI-assisted workflows versus who's still doing manual reconciliations and journal entry prep, you have no baseline to improve from — and no way to defend your pricing when clients start asking why your fees haven't moved.
You can't manage what you don't measure. If you don't know your team's AI adoption rate by workflow, you're flying blind on your own efficiency story.The Accountants Who Resist AI Tracking Are the Ones You Should Be Most Worried About
Amazon's internal pushback on AI monitoring mirrors exactly what firm owners are hearing from senior staff who built careers on manual expertise. The resistance isn't laziness — it's identity threat. But here's the hard truth: a staff accountant who completes a month-end close 40% faster with AI assistance is more valuable than one who takes full time and resents the tool. Start with one workflow — AP coding, bank rec, or expense classification — measure time before and after, and make the numbers visible to the team.
Pick one workflow this week, time it manually, then time it with AI. Show the team the delta. That number does more than any training session.AI Usage Dashboards Are Arriving in Your ERP — Here's What to Actually Do With Them
NetSuite, Sage Intacct, and several mid-market ERP vendors have quietly added AI usage telemetry into their 2025-2026 releases — logging which automated suggestions staff accept, override, or ignore during close workflows. Most firms have this data sitting unused in admin panels right now. The business problem it solves is concrete: you can see exactly where your team trusts the AI (vendor matching, duplicate detection) and where they don't (accrual estimates, intercompany eliminations), which tells you where additional training or model tuning will have the highest ROI.
Log into your ERP admin panel this week and find the AI activity or suggestion audit log. If it exists and you've never looked at it, that's your starting point.The Billable Hour Isn't Dying From AI — It's Dying Because Firms Won't Measure AI
Everyone says AI will kill the billable hour. The real killer is that firms are automating work without tracking the time savings, so they keep billing the old rate for a task that now takes a third of the time. That's a margin story until a client figures it out — and then it's a trust story. Amazon's model is the blueprint: measure adoption, tie it to output, and reprice accordingly before someone else does it for you.
If you automated something and didn't change the price, you didn't capture the value — you just delayed the reckoning.