The Vanity Goal Trap
Goals like “deploy AI across the company” or “achieve AI maturity” are vanity. They optimize for the appearance of progress, not progress itself. The CFO sees through them; the board eventually sees through them; the team feels their emptiness.
What Good AI Goals Look Like
Good AI goals share four properties:
- Tied to a business metric (revenue, margin, CSAT, retention, time).
- Measurable against a documented baseline.
- Achievable in 90–180 days.
- Owned by a named person.
Example: “Reduce first-response time on support tickets from 4 hours to under 30 minutes by Q3, measured against the Q1 baseline, owned by the Head of Support.”
Goals By Function
Support
- First-response time reduction.
- Tickets resolved without human within SLA.
- CSAT maintained or improved.
Sales
- Time from inbound lead to first meaningful response.
- Reps' admin time per week.
- Win rate at stage-2 (post-discovery).
Marketing
- Content output volume at consistent quality.
- Personalized email click-through lift.
- Time from brief to published asset.
Operations
- Hours per week recovered through automation.
- Error rate on routine processes.
- Time to onboard a new hire.
- • Can you trace the goal to a P&L line?
- • Is the baseline documented and defensible?
- • Is the target ambitious but believable?
- • Is there a fallback plan if missed?
Establishing Baselines
The hardest part of good AI goals: getting a credible baseline before you deploy. Most teams skip this and have nothing to measure against later. Spend two weeks gathering the baseline before the deployment starts. Without it, you can't prove ROI to your CFO or your board.
Review Cadence
Monthly review on tactical AI goals; quarterly on strategic. Same discipline as any other strategic capability. Don't exempt AI from accountability because it's “new.”
Warning Signs
- Goals expressed as activities (“deploy X”) rather than outcomes.
- Baselines that can't be defended.
- Goals owned by “the team” rather than a person.
- No review cadence.
A CFO who can't see the AI investment in next quarter's P&L will quietly defund AI. Make the AI investment visible in numbers the CFO already tracks.
FAQ
What about exploratory AI projects? Different framing — learning goals, not outcome goals. Don't mix.
Quarterly vs annual targets? Quarterly for tactical, annual for strategic. Both, not either-or.
Should AI be a separate budget line? Yes, for visibility. But the outcomes should sit inside business function goals.