Investment Reviews
Part of Media investment reviews
Deciding whether to reallocate, retest or stop a media channel
Use delivery, outcomes, uncertainty and commitments to decide whether to move spend, retest or stop a channel.
Reallocate when credible evidence and a feasible alternative support moving flexible spend. Retest when the channel may still serve a useful role but a specific, repairable uncertainty blocks the decision. Stop future spend when the channel no longer fits the brief, cannot deliver acceptably or cannot answer a worthwhile question at feasible cost. Record the conditions behind the choice.
Check whether the decision is ready
Start with the campaign review: confirm the channel's intended job, actual placement and dates, spend that can still move, the business outcome definition and material reporting limits. Repair a broken destination or missing export before treating the resulting rate as a fair test of the channel. Check booking and cancellation terms before assuming committed money is flexible.
Ask whether past evidence applies to the proposed next spend. An observed average may reflect a different audience, offer, price or flight. An attributed conversion rate describes credit under a reporting rule; it does not show by itself that another allocation will cause extra orders. A controlled lift estimate can address a tested change, but its relevance depends on the tested population, dates, spend difference and uncertainty.
Choose the action
| Action | Conditions that can support it | Record before approval |
|---|---|---|
| Reallocate | Credible evidence suggests a weaker use of flexible spend and a receiving use is feasible | Amount, destination, intended job, commitments and review date |
| Retest | The channel has a plausible role and a bounded test can resolve a decision-changing uncertainty | Question, correction, comparison, spend limit and stop rule |
| Stop | The placement or audience is unsuitable, acceptable delivery is unavailable, or further evidence would not justify its cost | Reason, affected bookings, customer impact and alternative plan |
These are decision criteria, not a scoring system. A bounded amount can also be maintained while the agreed outcome window closes; label that choice provisional.
Make a retest answerable
Name the uncertainty and the change that addresses it. If the first flight missed the intended market, correct the placement or settings and verify delivery. If accepted enquiries were not recorded, fix the outcome path before spending again.
If the question is whether the channel adds business beyond what would happen anyway, specify a feasible comparison and the media change it will test. Repeating a flight with the same missing records does not resolve the question.
A marketing mix model may inform a broader budget change, but an effect estimate from observational data depends on assumptions. Check those assumptions and whether the proposed spend is within a range the model can assess credibly. A suitable controlled comparison can provide evidence for the particular change tested. In either case, confirm that the channel can buy the required inventory.
Bound a reallocation or stop
Before moving spend, check the receiving channel's inventory, creative, response capacity and time left to produce useful evidence. Keep the total payable amount and GST basis consistent. Record a stated amount and set an early delivery check and a later outcome review.
If stopping, distinguish stopping new commitments from cancelling existing ones. Record whether the reason was poor observed value, an unsuitable placement or evidence that could not economically be improved. A failed setup does not establish that a channel can never work; a well-measured result may still justify declining a similar offer next time.
Close the decision with an owner, effective date, amount at stake, supporting evidence and the observation that would prompt reconsideration.
