Possibility-First Technology Training
Teach teams what technology can achieve before teaching individual procedures
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 82%
Possibility-First Technology Training shifts employee development from memorizing interface steps to understanding what a tool can accomplish. Leaders first map the technology's capabilities to meaningful business outcomes, then teach employees how to recognize relevant situations and articulate the result they need. Natural-language copilots can handle much of the changing procedural detail, while employees contribute context, judgment, and clear intent. The mechanism is capability awareness followed by use-case recognition, effective requests, and outcome evaluation. This makes training more resilient when software interfaces change and enables more employees to perform tasks once reserved for specialist administrators. Exact procedural instruction remains necessary where errors create legal, financial, or safety risks.
Origin
Extracted from Marketing Against The Grain during a discussion of the AI copilot built into Windows 11 and its implications for software training.
Core principles
- 01Understanding possible outcomes matters more than memorizing interfaces.
- 02Natural-language copilots reduce the need for procedural recall.
- 03Domain judgment determines which technological capabilities create value.
- 04Training should evolve as tools gain new capabilities.
How to run it
- 1
Map the Capability Space
Document the outcomes the technology can produce rather than listing every button or command. Group capabilities around the team's recurring business needs.
Pro tip Describe each capability with a concrete input and expected output.
Watch out Do not imply that the technology can perform tasks you have not verified.
- 2
Prioritize Valuable Situations
Identify the moments when each capability can save time, improve quality, or unlock work that was previously impractical. Use actual team workflows rather than generic demonstrations.
Pro tip Start with frequent, low-risk tasks that provide visible value.
Watch out Avoid prioritizing novelty over measurable usefulness.
- 3
Teach Outcome Articulation
Train employees to describe the desired result, relevant context, and constraints in natural language. Let the copilot translate that intent into changing software procedures.
Pro tip Provide request templates containing a goal, context, constraints, and success criteria.
Watch out A vague request can still produce a polished but unsuitable result.
- 4
Evaluate the Result
Teach employees to verify whether the output meets the business objective and to iterate when it does not. Preserve human review for consequential decisions.
Pro tip Use side-by-side examples of acceptable and unacceptable outputs.
Watch out Do not confuse successful software execution with a correct business outcome.
- 5
Refresh the Capability Map
Revisit training materials as the software gains new capabilities or limitations become apparent. Update outcome examples without rebuilding the curriculum around every interface change.
Pro tip Review the map after major product releases and recurring user failures.
Watch out Outdated capability claims can create both missed opportunities and unsafe reliance.
In the wild
A sales leader teaches new representatives which customer questions the CRM can answer, what context must be included, and how to validate the response. The representatives use the CRM copilot to find the current procedure instead of memorizing every reporting screen.
→ New representatives identify useful CRM workflows sooner while retaining responsibility for checking customer data.
A marketing team maps an analytics assistant's capabilities to campaign diagnosis, audience comparison, and anomaly investigation. Training exercises ask employees to select the right capability and define the desired output rather than reproduce a fixed sequence of clicks.
→ Employees adapt more easily when the analytics interface changes and rely less on specialist administrators.
Common mistakes
Replacing All Procedures
Possibility-first training complements rather than eliminates exact procedures. Regulated, destructive, or high-risk actions still require explicit controls and verified steps.
Teaching Features Without Situations
A catalogue of AI features does not help employees recognize when to use them. Tie every capability to a recurring problem or decision.
Ignoring Output Verification
Natural-language access makes software easier to operate, but it does not guarantee that the requested action or generated result is appropriate.
Is it for you?
Best for
Teams adopting AI-enabled software whose procedures will change faster than their underlying business objectives.
Not ideal for
Safety-critical or regulated operations where employees must still master exact, auditable procedures.
From the transcript
“it's more important, gonna be more important in the future to train your employees on what is possible with a technology than how to specifically…”
“AI is going to make it very easy for you to be like, "Oh, I know I wanna do this thing, but I'm not sure…”
From the episode
Photoshop’s New A.I. Update + 4 More A.I. Tools You Must Try (#125)