[ AI productivity · Full playbook ]
// 06 / 20The buyer's playbook.
The working reference behind AI Productivity Consulting Buyer Guide 2026: what to ask, what to watch and how to run the first 90 days.
← Back to the guide[ RFP questions ]// 01
Questions that reveal real capability.
Ask for a relevant example, the decision made, the result or learning, and the person who would own the work.
- 01Which tasks should remain human-only?
- 02How will workflows be selected by value, feasibility and risk?
- 03What data may be entered, retained or used for training?
- 04How will output quality and hallucination risk be evaluated?
- 05What training changes behaviour after launch?
- 06How will time saved become capacity or better outcomes?
- 07What adoption telemetry is privacy-respecting?
- 08How can tools and policies change without disruption?
[ Risk ]// 02
Make failure visible before signature.
| Risk | Early warning | Control |
|---|---|---|
| Licence shelfware | Seats without changed work | Pilot role-specific uses |
| Unrealized time | More low-value output | Redesign workload |
| Confidential data leak | Policy vague | Approved tools and controls |
| Quality decline | Fluent output bypasses review | Rubrics and human gates |
| Shadow AI | Official path unusable | Safe practical alternatives |
[ Value case ]// 03
Measure outcomes—not activity.
01Cycle timeBaseline ________
90-day target ________
90-day target ________
02Output qualityBaseline ________
90-day target ________
90-day target ________
03Adoption and repeat useBaseline ________
90-day target ________
90-day target ________
04Time shifted to higher-value workBaseline ________
90-day target ________
90-day target ________
05Error and escalation rateBaseline ________
90-day target ________
90-day target ________
06Cost per completed workflowBaseline ________
90-day target ________
90-day target ________
[ First 90 days ]// 04
Move from evidence to operating rhythm.
01
Days 1–30
Workflow inventory, risk boundaries and shortlist
Exit evidenceApproved baseline, owners, scope and risk log
02
Days 31–60
Pilot, training, support and quality measurement
Exit evidenceWorking outputs, QA evidence and decision record
03
Days 61–90
Scale decision, governance and operating routines
Exit evidenceMeasured result, learning backlog and operating owner
[ Reference desk ]// 05
Speak the same language.
| Term | Plain-language meaning |
|---|---|
| Copilot | AI assistance embedded in human work. |
| Prompt pattern | A reusable instruction structure. |
| Grounding | Trusted context supplied to an answer. |
| Evaluation | Repeatable quality and risk measurement. |
| Human-in-the-loop | A required human decision point. |
| Adoption rate | Meaningful use by intended people. |
[ Playbook in hand? ]