Every planning cycle begins with evidence.
Which channels performed? Which audiences responded? Where did reach become expensive? What drove attention, engagement or sales?
These are sensible questions. The problem is that they all look backwards.
By the time the results have been analysed, presented and incorporated into the next plan, the audience has moved. Behaviour has changed. New cultural signals have emerged. A channel that was efficient six months ago may now be approaching saturation.
Yet its previous performance makes it the safest place to put the next pound.
And so last quarter keeps winning next quarter’s budget.
Familiar channels have a structural advantage
Established channels are not only familiar. They are measurable, scalable and easy to buy.
They have mature reporting, established benchmarks and years of historical performance. Agencies understand them. Finance teams recognise them. Marketing leaders know how to defend them.
This creates a powerful gravitational pull.
Budget continues to flow towards the channels with the clearest evidence of past performance, rather than necessarily towards those offering the greatest opportunity now.
The issue is not that historical data is wrong. It is that the decision is incomplete.
It tells us what worked under a previous set of conditions. It cannot, by itself, reveal where attention and demand are moving next.
Measurement can reinforce concentration
The channels receiving the most investment generate the most data.
That data creates greater confidence. Greater confidence attracts more investment. More investment produces more data.
It is a self-reinforcing cycle.
Meanwhile, emerging opportunities can struggle to compete because they have less history, fewer established benchmarks and more uncertainty. They are judged against mature channels using evidence they have not yet had the opportunity to generate.
This can leave brands with a strange imbalance: sophisticated optimisation inside individual platforms, but limited independent evidence about whether the overall allocation between those platforms is still right.
Every channel may be getting better at spending its share of the budget.
Far fewer systems are asking whether it should still have that share.
Optimisation and planning are different problems
Real-time optimisation is extremely valuable.
Agencies and execution teams can adjust audiences, creative, placements and bids as performance changes. That helps each channel work harder once the investment decision has been made.
But optimisation usually begins after the budget has entered the channel.
Media planning asks a different question:
Where should the budget go in the first place?
Answering that requires a wider view. Not only platform performance, but signals from search, social behaviour, entertainment, news, weather, events, commerce and the physical world.
Individually, these signals may appear weak or disconnected. Together, they can reveal where attention is accumulating, how quickly it is spreading and whether it is likely to become meaningful demand.
This does not replace historical performance. It adds a forward-looking evidence layer to it.
Attention moves before budgets do
Audience behaviour rarely changes neatly at the beginning of a planning cycle.
A theme can emerge in one community, appear in search, cross into entertainment and accelerate through news or events. Its relevance may vary dramatically by audience, category and market.
By the time it appears in established brand research or campaign reporting, the most valuable moment to act may already have passed.
The opportunity is to identify that movement earlier and ask:
- Where is attention beginning to build?
- Is it sustained momentum or temporary noise?
- Which audiences and markets are leading the change?
- How significant could it become?
- How long is it likely to remain relevant?
- Which channels are best placed to capture it?
- Where is existing investment showing diminishing opportunity?
These are not simply insight questions. They are allocation questions.
Making every channel earn its place
A more dynamic approach to planning does not mean chasing every trend or moving budget whenever a graph changes direction.
It means introducing a consistent process for challenging inherited allocations.
For every significant investment, the organisation should be able to explain:
- Why this channel deserves its current share
- What evidence suggests that opportunity is growing or declining
- What alternative opportunities were considered
- What would need to change for investment to move
- How confident the organisation is in that decision
The outcome might be to increase investment. It might be to hold it. It might be to reduce or redirect it.
The objective is not constant movement. It is justified movement.
A better planning question
Instead of asking only:
What performed best last quarter?
Brands could also ask:
Where is attention moving, and what would have to be true for investment to follow?
That shift matters.
It moves planning from repeating yesterday’s winners towards comparing tomorrow’s opportunities. It allows emerging channels, behaviours and markets to be considered before they have accumulated years of performance data.
It also gives marketing teams a stronger basis for diversification. Not intuition, novelty or pressure to try something new, but evidence, timing and an explicit level of confidence.
Historical performance will always matter. It should inform the next decision.
It should not make the decision by default.
The next media pound should not go somewhere simply because the last one did.
Every channel should earn its place in the plan.
How we are approaching the problem
Absurd is a product and service design agency that specialises in complex, data-led problems.
We help organisations turn fragmented data, emerging technology and difficult operational challenges into products and services that people can understand, trust and use.
That matters because this opportunity is not simply about building a machine-learning model. It requires the design of a complete decision-making system: how signals are collected, structured and interpreted; how predictions are tested; how uncertainty is communicated; and how recommendations fit into existing media-planning processes.
Our approach combines signals from multiple sources, identifies patterns that may be difficult to see in isolation and measures how quickly attention is building, spreading or declining. Machine learning can then help estimate the potential scale, timing and longevity of an opportunity, supported by a clear confidence score.
The technology is only part of the answer.
Reliable predictive products depend on well-structured data, carefully designed signals, transparent models, robust validation and meaningful human oversight. They must communicate uncertainty honestly and translate complex outputs into recommendations that people can understand, challenge and use.
We have spent years developing the knowledge and practical experience required to build and operate this kind of intelligence in live environments. We are now applying that experience to a new media-planning question:
Can independent attention signals help brands make better decisions about where investment should increase, hold or reduce?