Budget Setting & Optimization Use Case

The challenge

A large global broadcaster ran one marketing effectiveness project that covered all of its products. It then wanted to use the modeling results to optimize its total budget across every product and every channel. In particular, the budget had to account for the direct impact of each campaign on its own product, and for the halo impact that campaign had on the performance of the other products. However, the question was simple to ask and hard to answer: how much to spend on every country and every channel.

One budget, many countries and channels

The broadcaster sold in the USA, the UK, China and Japan. For example, in the USA, the UK and China, each country reached customers through its own sales routes, and each route used its own media, from digital and TV to outdoor, press and pay per click. As a result, setting one budget for all of it at once meant choosing from a very large number of possible splits.

Four things the budget had to account for

• All products at once. The project covered every product, so the budget could not be set one product at a time.
• Countries and channels. In addition, spend had to be split across four countries, and across the sales routes and media inside each one.
• Direct and halo impact. Every campaign affected its own product and also the performance of the other products.
• Seasonality. Finally, the optimization had to take into account the seasonality generated period on period.

Next, the table shows the countries, sales routes and media that one budget had to cover.

The solution

MASS Analytics built a stand-alone optimizer so the broadcaster could use its modeling results to optimize across products, channels, variables and periods. The optimizer then takes seasonality into account, period after period. It also lets the client set minimum, maximum and fixed budgets for specific periods and specific channels. In short, this kind of work relies on marketing optimization models that turn a finished model into a budget decision.

One optimizer for four levels

In practice, the optimizer works on four levels together: products, channels, variables and periods. Because it sees them together, the plan can take into account the direct and halo impact of each campaign. When the client needs a guardrail, it sets a minimum, a maximum or a fixed budget for a chosen channel or period, and the optimizer works inside those limits. Overall, a media mix optimization model built this way keeps the plan tied to the numbers behind it. In addition, a team that already has a model can run this kind of work on a budget optimization platform that sits on top of it.

Two ways to ask the question

The client could then run the optimization in two ways. First, with Maximum Budget, it sets the most it can spend. Second, with Minimum Target, it sets the target it must reach. In both cases, the optimization starts from the same modeling results, and it returns a tailored optimization the client can read by channel, model or variable.

Results and Impact

With the optimizer, the broadcaster raised efficiency by 20%, a gain that came from marketing performance optimization across its whole budget. However, the figure is a modeled efficiency gain, built on optimization reports the client could update with its latest campaigns and constraints.

Efficiency rose by 20%

Ultimately, the optimization let the broadcaster increase efficiency by 20%. For instance, the client could look at the optimal solution by channel, model or variable, interpret it, and choose the budget split that reached its target. It could then update the optimization reports with the latest campaigns and constraints, and that flexibility helped the broadcaster reach the 20% gain.

Reports the client could keep updating

Meanwhile, the broadcaster built monthly and quarterly reports, viewed the solution bucketed by period, and adjusted the constraints when it needed to improve overall performance. Because of this, a team can turn model results into budget decisions it can repeat each period. The same idea has worked for other brands, for example a personal hygiene brand that optimized its media mix by moving budget away from a saturated channel.