AI-Native Situation Mapping
Redesign recurring work from the desired AI-native outcome backward.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 7
- Confidence
- 98%
AI-Native Situation Mapping begins with a recurring event that triggers work, documents how people respond today, and then redesigns that response from the desired future state backward. The manager assumes useful signals can arrive proactively and repeatedly asks what the AI could do next instead of stopping at notification or summarization. Each proposed step identifies its data sources, output, and human decision point. This converts a vague instruction to “use AI” into a visible operating model that employees can understand and pursue. The mechanism improves speed without surrendering judgment: AI detects, assembles, analyzes, and prototypes, while people evaluate options, make consequential choices, and remain accountable for quality and business value.
Origin
Hillary Gridley demonstrated the method on Marketing Against The Grain by redesigning a marketing team's response to a competitor copying its positioning.
Core principles
- 01Start with recurring situations rather than isolated tasks.
- 02Design backward from the desired future state.
- 03Make relevant signals arrive proactively.
- 04Ask the AI to advance beyond detection into proposed action.
- 05Reserve consequential choices and accountability for humans.
How to run it
- 1
Select a recurring situation
Choose a meaningful event or problem that repeatedly causes your team to act. Define the trigger precisely enough that a system could recognize it.
- 2
Map today's response
Write down what a person currently does after the trigger, including searches, handoffs, documents, meetings, and approvals.
- 3
Picture the future outcome
Describe what the response should look like if the team worked in a genuinely AI-native way. Work backward from that outcome rather than incrementally adding AI to the existing process.
- 4
Make signals proactive
Identify information that people currently have to notice, request, or hunt down. Design the system to surface those signals automatically with the necessary context.
- 5
Push one step further
For every proposed AI action, ask what useful next action it could also perform. Move from detection to analysis, options, prototypes, and downstream updates where appropriate.
- 6
Define the human role
State exactly what the person reviews, chooses, edits, approves, or owns. Make quality expectations and accountability explicit.
- 7
Make the vision concrete
Document the data sources, actions, interfaces, outputs, and handoffs. Walk the team through the future workflow so they have a shared destination.
In the wild
A competitor launches messaging that closely resembles the company's positioning. Instead of manually discovering and documenting it, an agent monitors competitor messaging, compares it with the company's canon, flags material overlap, proposes three differentiated angles, and produces landing-page previews with supporting estimates. A marketer applies judgment, selects and edits an option, then approves downstream page, ad, and email changes.
→ The team gives the CEO a considered prototype within hours rather than spending weeks assembling a response.
Common mistakes
Adding AI to today's tasks
Incrementally inserting AI into an unchanged workflow preserves its old assumptions and bottlenecks instead of designing for a better outcome.
Automating without strategy
A team can become highly AI-native while producing projects that customers never need or use.
Removing the human decision
End-to-end automation can eliminate the judgment that differentiates excellent work from plausible slop.
Is it for you?
Best for
It is best for managers redesigning recurring, information-heavy workflows around AI.
Not ideal for
It is not ideal for rare tasks that lack stable inputs, repeatable decisions, or accountable owners.
From the transcript
“And in general, I like to think where are we trying to get to?”
“And then take a step back and try to reimagine what that would look like if it truly had AI at the center of it.”
“What is the exact next action that the AI is doing? And what is the role of the human in all of this?”
From the episode
If Your Team Is Producing AI Slop, Here's How To Fix it