We build bespoke optimisation and allocation tools for our clients, as every business is different. We build the tool around the challenge, not force fitting the problem into the tool. Generally this stage is built using information from market mix modelling, but can also be built making estimations using simpler (annualised) data. Some examples of response curves are visualised below.
Response curves are built using modelled data and allow the prediction of sales from different levels of media budget by channel. An example is shown below. These are usually built using market mix modelling, with the econometric models delivering the measure of marketing ROI at current spend levels. If a medium hasn’t been deployed, or it’s a new brand, then these can be estimated using marketing ROI benchmarks and consensus workshops.
These are built across brands, products and markets to optimise global media investment across all touchpoints.
Once the response curves are constructed via the market mix models (above), it is then possible to trade them off against each other in order to get the most sales out of total media budgets. This is done by playing off investments across different touchpoints, and then spending where the return is greatest, or optimal. This way the marketing ROI is maximised by delivering the highest return (e.g. sales) for the given investment.
Below is an example of response curves at work. To view the chart at its best, click on the option ‘same colour’ and select ‘unique colours’ for a bit more colour, then click on all the channels you want to track, then click play in the bottom left.
The animation shows the allocation in progress – firstly to twitter, then the spend becomes more efficient in other channels and so is spent elsewhere in order to maximise marketing ROI.
You can also watch the animation via bars to visually see where the budget is allocated in terms of spend – click on the ‘bars’ tab.