Starcount has a specific four-step approach to delivering data-driven business transformation. It’s called DIAL: Data, Insight, Action, Loyalty/Learning. In this new series we explain more about the process and how it can help improve customer understanding and transform business performance.
The first stage in Starcount’s D.I.A.L model is a crucial one: cleaning up our clients’ data to make it fit for purpose.
Frequently we find clients’ data is in disparate systems (operational, CRM, loyalty) or even kept in spreadsheets, and is brought together for analytics in an ad hoc, messy way. More often than not, clients attempt to implement ambitious data science before having a data set that is complete and fit for purpose, rather like trying to run a car with crude oil, rather than refined petrol.
That’s where Starcount comes in. Retailers’ data comes from a mix of digital and physical sources, and as new services are added they are often “bolted-on” rather than integrated with existing systems. While this is totally understandable (bolting on is quicker than integrating), this usually leads to silos of data with little coherency in the way it’s stored, processed and, importantly, consumed as a complete set of data. Siloes also lead to repetition and inconsistency; and this inconsistency, coupled with the disparate nature, means the data set isn’t easily consumable for insight.
Operational data’s central purpose is to allow the business to run like normal but, just because the data keeps the business operational, it doesn’t mean the data can be used to analyse these operations in a meaningful way. To take one example: the way a promotion is described and categorised when set up may not make any difference to the operational execution of that promotion at the till, but inconsistencies in set up will have a huge impact on how the effectiveness of groups or types of promotions can be analysed. Small errors can soon mount up, creating complete confusion and a lack of effective insight, potentially costing a huge amount in the bottom line.
As part of Starcount’s first stage of work for our Enterprise Consultancy clients, we combine data from operational systems to create a data set fit for analytics. We have years of experience across the team in setting up the right systems to interrogate and interpret the data.
Changing mindsets on data
Beyond the practical realities, a key challenge for many companies is that the final purpose of the data wasn’t considered when it was first generated. In order to create a smoother, more agile data analytics process, every employee should understand the eventual value of the customer insight (and how their role fits into the overall process). A core part of our approach is embedding this data-focused mindset in our clients’ businesses.
Checks and balances
The maturity of our clients’ mindset around data and how it’s used in their business will affect how difficult the clean-up process is. For all clients, no matter how mature, we undertake fundamental checks of the data to make sure it has good integrity and makes sense. Our process is rigorous and based on a tried and tested method.
If a company has a more advanced set-up around data, we’ll do checks that blend more into data science; these are less about fundamentals and more about showing clients how our expertise can enhance their current processes.
With your data fit for purpose, we can really focus on making the changes that count – embedding a data culture in your business and transforming your business performance.
Ed Blake leads Starcount’s Data and Systems team. Ed is a data and technology leader with over 15 years’ experience in big data and analytic technologies.
Find out how Starcount can help you get control of your data and transform your performance with an expert consultation.