With data so prevalent, data science has become a buzzword for sophistication. However, so much time is wasted on bad analysis of mass data simply because of a failure to ask the right questions and apply the right techniques from the outset.

Central to our ethos is offering fast and very low cost analysis upfront. This takes the form of a feasibility study and allows the client to break the project if there are data gathering barriers. This feasibility study provides an overview of the data we have collected, any gaps in the data or issues we foresee. It also involves early correlations and allows us to discuss these findings with the client.

A core part of our offering involves three overarching pieces of analysis:

  1. Econometric modelling, ROI and Diminishing Returns (via a strategic workshop)
  2. Paid, Earned, Owned and Social interaction (via a strategic workshop)
  3. Media optimisation (via a simulator)

As separate analyses (or as part of a project) we can also offer:

  1. Total media budget setting (using Jones analysis)
  2. Brand equity modelling and long term impact (pending available data)

Why does anyone need econometrics?

Quite simply econometrics is the best technique available to understand the impact of marketing, media and communications on the bottom line of a business. It gives clients proof that their efforts return to the business, and allows them to argue for more robust budgets in areas they know to be effective. It:

  1. Is founded with strict laws and tests that need to be passed in order to provide a valid model
  2. Is based on cumulative and not individual data, therefore navigating GDPR problems associated with other techniques
  3. Looks at the full impact of advertising, and not just the immediate impact that attribution and last click models work with. This means the results are holistic (i.e. if someone sees a TV ad and then searches for the product, then this sale is driven as much, or more, by TV than paid search)

For more information on how econometric models are built click here.

What are the outputs?

Every project starts with identifying a list of key questions, these are critical to the focus and the output of a project. There can be many questions, but the typical ones found in every project are:

  1. What are the key factors that drive sales?
  2. What are the returns on investment by media?
  3. For each media when do diminishing returns to investment set in?
  4. What are the decay rates of the media?
  5. What is the optimal media mix?
  6. What is the recommended online/offline ratio?
  7. When are the optimal seasonal lay-down?

However findings are not limited to these questions, if we discover interesting insights in other areas (for example price promotions or distribution changes) we will report on them.

Model fit and quality

ROI by media

Sales decomposition

Diminishing returns to investment

In the era of modern marketing, the worlds of paid, owned, earned and social media are increasingly blurred. Econometric models should be built to reflect the reality of this world.

Intrinsically we know that Paid media drives people to other forms of media, and that media itself interacts and synergises, and that the ultimate goal of all this is to drive an increase in sales.

Traditional techniques tend to focus entirely on how one of these channels drive sales (or often a weaker metric higher up in the funnel).

It is however also possible to investigate the impact of how each of these interact with each other and how they all in turn drive revenue.