When we audited Stackline’s internal stack, the spreadsheet had 47 line items. Not 47 users — 47 products. Some were $29/month. Some were $29,000. Finance wanted to cut spend. I wanted to cut confusion.
Every tool was someone’s favorite. Notion for docs, Confluence for “official” docs, Google Docs for drafts, Coda for one team’s wiki. Three chat tools because sales adopted one before IT approved another.
We got to 12 in four months. Here’s how.
Step one: follow the login, not the invoice
Finance sees contracts. Ops sees reality. We ran Okta reports for 90 days and tagged anything with under 40% MAU as “zombie.” Zombies totaled $312K annual spend.
Step two: one owner per job
We defined 12 jobs — not tools:
- Source control
- CI/CD
- CRM
- Support desk
- Internal docs
- Design
- Analytics
- HRIS
- Expense
- Security
- Video
- Project tracking
Each job got one owner and one product. Exceptions needed VP approval and a sunset date.
Step three: migration windows, not big bangs
We gave teams 60 days to export and move. Confluence → Notion was the bloodiest. We hired a contractor for two weeks to script migrations. Worth it.
What we kept that surprised people
Salesforce stayed. Everyone hates it. It also runs $14M in pipeline and integrates with everything. The spreadsheet fantasy died fast.
We dropped a trendy AI writing tool — 12 seats, 9 never logged in, $18K/year.
Tool audit scorecard
- Monthly active users / paid seats
- Annual cost
- Overlap with another tool (Y/N)
- Integration criticality (1–5)
- Migration effort (hours)
Four months out
Fewer tools, clearer defaults, faster onboarding for new hires. New employee IT setup went from 2 days to 4 hours.
Cutting tools isn’t austerity if you cut the right ones. Start with logins, not line items.