Choose a decision before choosing a build
Coming home with twenty promising ideas creates a selection problem. Which idea belongs in the work already waiting? Start by naming a recurring decision that better evidence could support. Keep it specific enough that someone can identify the next action and the person responsible for it.
The Work After Summit is David’s capstone synthesis of the first eight Conference Insights Engine episodes. Its Summit Adoption Model connects seven choices: one decision, process, signal, campaign, hidden object, teaching artifact and practitioner skill. These are connected parts of a small improvement effort, rather than seven new projects competing for attention.
In the recording · 08:00 ↗Week one: give one idea a working home
Gather the notes that bear on your chosen decision. Set the rest aside for sorting. Write down what the office decides, how often it decides, where the evidence lives and who acts on it. Include the timing that makes the evidence useful.
For example, a fictional team might ask which event registrants need an operational reminder before Friday. That question is narrow enough to inspect. “Improve engagement” leaves the audience, action and evidence unresolved. Keep the decision statement beside the project so later requests can be checked against its purpose.
Name the people who should participate and reserve a review date. The workbook gives the team profile a month of action and a review date because follow-through needs a place in the calendar. A promising idea without protected time is still waiting for a project.
In the recording · 13:12 ↗Week two: map the process and limit the inference
Trace the process from its starting event through incoming data, rules, staff actions and the experience it creates. End with the report or review that checks whether it worked. Keep the first map simple enough that a colleague can point out a missing handoff.
Then define one signal already captured by that process. State what it suggests, what it fails to prove, who should see it and which action it should inform. An event no-show can have several explanations. It should not silently become a conclusion about a person’s intentions.
Look for the smallest uncertainty that blocks useful action. A clearer attendance definition or an agreed audience owner may matter more than an additional dashboard. Record the uncertainty and the evidence needed to resolve it before building around an assumption.
In the recording · 08:55 ↗Week three: improve one maintainable piece
Choose a bounded change to a campaign, workflow, report or hidden object. For a campaign, document the purpose, audience, suppressions, reusable blocks, test cases and ownership. For a hidden object, record why it exists, what depends on it and what must be checked before a change.
A retirement plan can be a useful outcome when an old query has no obvious owner. First find its consumers and dependencies. An unclear purpose is a reason to investigate, not sufficient evidence that deletion is safe. The same care applies when a permission group appears broader than expected.
Adapt the conference example to local conditions. David asks builders to bring the assumptions home with the idea: required data, governance, staff capacity, failure points and teaching needs. Preserve the original session reference with the working notes so the team can distinguish a presenter’s example from its own proposed adaptation.
In the recording · 10:39 ↗Week four: teach the work and inspect the result
Create something another person can use: a source map, short walkthrough, test checklist or one-page logic note. Explain what the object does, why it exists, what can go wrong, what needs testing and who maintains it. Choose the practitioner skill that this work should develop.
Have a colleague use the artifact to explain or test the process. Record what was unclear and revise the artifact. The teach-back makes the month’s learning visible in what a colleague can explain and use.
At the month-end review, examine the decision statement, process map, signal definition, improved object, teaching artifact and skill target. Decide what should become standard practice and what should continue next month. An unresolved dependency may justify more investigation instead of expansion.
In the recording · 11:59 ↗Keep the remaining ideas findable
Sort the remaining notes into use now, plan next, study later and share with someone else. The categories help a team preserve useful ideas without giving every idea an immediate deadline. Keep source references and readiness questions attached so a later review has something concrete to revisit.
The series offers paths back into the work: reporting leads toward decision infrastructure and query literacy; communication leads toward signals and governance; onboarding leads toward teaching and practitioner development. End the month with a few decisions the team can explain and a next review it can actually hold.
In the recording · 23:34 ↗