Thinking

How to get out of the data rut, now

Written by Milan Patel | September 29 2020

Living in a digital age, data is an abundant resource, yet remains undervalued and underinvested in. The result of poor management, maintenance and governance of data is severalfold: employees reduce system interaction due to a lack of trust in its data, reduced operational and transactional efficiency, a poor customer experience (CX), lack of insights and misinformed decision-making are just a few examples. Bad data has a negative ripple effect seeping through all layers of the business, yet this effect can be reversed into a virtuous spiral when data is put in the middle of the conversation, trusted and legitimised. Improving data leads to an increased dependence on it for the business activities, driving further demand to improve data, thus spreading its powerful enabling effect. Success ensues…

So how can you take immediate action to get out of your data rut and leverage its power without being all consumed by its vastness?

Start with cleaning historically bad data that is business critical; several of our customers have proven to us that over 70% of their account and contact records hadn’t been touched for a year. A retrospective data clean is key because it acts as a reference point for current and future data comparatives, trends, and impacts several business activities.

What exactly is “bad” data? 

“Bad” data can be measured against 5 key dimensions:

  1. Accuracy – does the data conform to business rules for the data type? E.g. “Account Name” must always be a legal defined entity name made up of a string of letters & numbers
  2. Completeness – how populated is the field across records? E.g. 100% of Account records must have an account name
  3. Consistency – is the data type consistent between tables, systems, records etc? E.g. is a field named “Account” always used to represent a Legal Entity Name and is the data the same across systems?
  4. Reliability – how much can the data be trusted throughout its lifecycle? E.g. it conforms to defined data standards
  5. Existence – does the data refer to something that exists? Old data that refers to non-existent or irrelevant real-world entities is a huge compliance risk

But where to start with bad data? 

The best place to start is with the data that matters the most to the business and that can prove to deliver value quickly. The reasoning here is twofold: Firstly, quick-wins will generate the buy-in needed in the organisation to legitimise investment in data. How? Report success regularly and widely, through clear dashboards that scores the most important data, and is tracked and trended. Secondly, select a single area of focus then iterate this focus to other areas. This ensures progress is continuous and value-adding, instead of a single effort to “boil the (data) ocean” which can drain time and workforce motivation.

We believe good business starts with a great customer experience, and improving customer data is a basic fundamental building block to providing an in-depth, sophisticated customer experience. For example, accurate and complete basic customer attributes like name, location and contact detail, is a necessary input to CX initiatives.

A charity asked us to help uncover a single customer view to identify their marketing and sales channels. However, we couldn’t answer this without first addressing the data that was poorly captured, mapped, and disseminated which blocked us from answering their exam question “what do our customers really care about?”. We were only able to get a single view of the customer by first deduplicating customer data and then setting up the right attributes by following a well defined data cleaning governance structure.

Data governance? That sounds rather complicated and dry…

A common misconception! In parallel with a retrospective data cleanse, data governance offers you the  key to embedding and maintaining good data management going forward. Yet this exercise does come with risks of being “over-intellectualised and made too complicated an exercise” as our CEO, Matt Cheung, discusses in detail in our podcast – Data Chaos, and Sexy Data. 

Data governance is not complex, because not all data is equally good, bad or important, therefore does not need to be governed all at once. Furthermore, data governance effort can be tailored to your organisation’s level of data maturity, starting with the simplest form – a well-governed data cleanse. For companies who are more advanced in their data journey, data governance can progress to include preventative elements such as a data strategy, roles, technology, documentation and process including automation through RPA and leveraging AI, if you are ready. Even for those who are comfortable with data, governance can remain simple at its core, as Matt talks about in our podcast, from a lead to cash (L2C) perspective:

“Data governance can simply involve tracking what is consumed at what step in the L2C process, and knowing how that data evolves through the pipeline, and map to systems. As a lead matures, collect more data. If you have more than 10 attributes of a contact, you are doing something wrong”.

Why not take up our data governance assessment to see where you land on your data maturity and how we can help you?

Down with the data? It’s still down to you, the people…

Matt’s quote above touches on another key aspect; that even with data improvement, consideration of the user journey and customer journey is key to success. To ensure good touch-points with data, create clear ownership and accountability of it. This also emphasises to the organisation a commitment to high quality data. For example, a CDO role could put data on the executive agenda, but tailor the specific intervention to suit your organisation.

To summarise…

Bad data is like bad lubricant in an underperforming engine. It creates unnecessary friction within the business, creating unseen inefficiencies and costs, such as manual workarounds and processes, or assumptions that lead to unreliable insights and therefore incorrect decisions. Poor data creates excess work to compensate for it, and this effort can be eliminated through a prioritised approach to data cleansing through a well-governed process. Data governance doesn’t have to be more complex than this, and can evolve to suit your level of data maturity.

Overall, it is key to get your data to a point where it is trustworthy, both objectively and culturally, so that it can be embraced as a core asset which empowers the business. We believe that this will only get more important as we progress in this information age where data becomes more readily available and easy to manipulate, and customers demand a higher quality, personalised service. For more on this, check out our blog on how to use GDPR regulation to empower, not hinder your data strategy and governance framework.