Why data quality and management are critical to properly measure the effectiveness of Marketing

Image Credit: https://en.wikipedia.org/wiki/Measurement

One of the key challenges faced by marketers is to link the outputs of the marketing activities with their company Business KPIs in order to make a solid case for their strategy and budget.

Marketing Mix Modeling (MMM) and Digital Attribution are excellent solutions to manage this challenge.

The purpose of MMM is to apply the principle of econometric modelling to marketing. It consists in  designing a statistical model which will explain the complex relationship between marketing interventions and business outputs ( Sales, SOM or Turnover..) taking into account all factors which can influence the business metrics, including external factors (weather, events…). Once the descriptive model has been designed, it can be used as a short-term predictive tool which allow marketers to simulate the business results of different marketing scenarios with a good level of confidence.

MMM requires to create a data set composed of time-series for each variable. The business variable that the model aims at explains and the variables that can influence the business metrics: mostly but not limited to the marketing metrics (spend, GRPs, clicks, impressions…). To allow for a robust statistical analysis based on linear or log-linear regression, MMM specialists will ask for a minimum of 156 data points with a weekly granularity.

The purpose of digital attribution is to attribute the online conversions (or any consumer action) to the different digital interactions with the users prior to conversion/purchase. Digital Attribution will require the access to individual data extracted from the Ad Server and the Site Centric Analytics solution.

Combining MMM& Digital Attribution in  a holistic Measurement Framework is a very reliable solution to assess for Marketing performance and inform the marketing decision process : optimal budget definition, allocation per type of marketing activity (Price, Promotion, Distribution, Paid Media, Content …) and channel and sizing and phasing of actions, continuous improvement…

This is true and we at Metrics 720 (as other Measurement specialists) can share many client cases studies to prove it.

Unfortunately, we also know from experience that the quality of the data inputs will determine the quality and relevance of the insights and learnings.

The reality is that too often the data collect phase of our Measurement missions highlights that marketers do not have access to sufficiently structured and managed data sources to measure accurately the effectiveness of their marketing strategies and plans.

We will never insist enough on the critical importance for all brands to gather, store and manage their Marketing data. This essential discipline will create the basis for a robust Effectiveness assessment and informed discussions at the C Suite table. It is no rocket science and starts with simple housekeeping rules such as:

  1. Make sure that all the existing data sources are listed and documented
  2. Store all data points in one place and ideally in a secured warehouse managed by your IT/Bi Team
  3. Demand that your teams and external marketing partners respect strictly a unified naming convention

At a time when so many articles are written and Proof of concepts are signed on Machine Learning applied to marketing, no none should ignore the critical importance of a proper management of marketing data points

Leave a Reply