Not every question deserves a program. Not every decision survives on a single snapshot. This guide shows how to make that call from the half-life of your question, not from gut feeling — and when a program starts to pay for itself.
Why this matters
One-shot studies are like photos: clear, expensive per image, out of date as soon as the situation moves. Programs are like sensors: continuous stream, cheaper per observation, but only valuable if leadership actually uses the cadence. Confuse the two and you burn either budget or decision space.
Seven steps to the right choice
- Map leadership's decision cadence. How often is the question actually raised? Once a year, quarterly, continuously?
- Classify questions by half-life. Does the answer move in weeks (customers, frontline, operations) or in years (positioning, brand core)?
- Cost model: setup vs run. Programs have higher setup costs but a lower per-observation cost. One-shot studies are cheap to set up, but every new occasion restarts from zero.
- Define cadence. A program without fixed rhythms turns into an unplanned backlog. Monthly, quarterly, or half-yearly — set it and defend it.
- Ad-hoc triggers. Programs still need occasions that justify extra studies (crisis, acquisition, strategy shift). Rule: no more than two per year, otherwise the program unravels.
- Ownership. A program needs a single owner with mandate. Distributed responsibility becomes distributed irresponsibility.
- Exit criteria. A program that hasn't influenced a decision after two cycles is either badly framed or unnecessary — either way, someone must notice.
Decision matrix
| Situation | Recommendation |
|---|---|
| One-off strategic question with a clear deadline | One-shot study |
| Customer experience in a shifting market | Program, quarterly |
| Transformation over 12+ months | Program, semi-annual (Reality Scan) |
| Brand positioning | One-shot, repeated every 2–3 years |
| Root cause for an acute KPI move | Ad-hoc within an existing program |
Pitfalls
- Program without ownership. Fizzles out after two cycles.
- One-shot with a repetition habit. Running the same study three times in a row is a program — without the benefits.
- Coupling cadence to availability. A cadence you renegotiate every time isn't a cadence.
- No exit criteria. Programs without an end condition turn into furniture.
Frequently asked questions
From what size does a program pay off?
Organisation size isn't the deciding factor — question frequency is. If a question comes up at least quarterly, a program is usually worth it.
Can you combine both?
Yes. The common setup: one running program for core questions plus one or two one-shot studies per year for strategic add-ons.
What if a program turns into a compliance exercise?
Apply the exit criteria. A program that no longer influences a decision should be paused or restructured — not continued.