Let’s say your data strategy is to create a Single Customer View (SCV). This sounds good and you’ve heard your competitors are doing it and, after all, many of the data consultancies and data tool vendors you’ve met have recommended that you invest in producing a SCV aka a 360° Customer View. But are you sure this is what your business needs?
In this blog I’ll argue that focus should be on identifying the real business needs and desired outcomes. Only then should the conversation move to solutions, and in many cases a Single Customer View simply isn’t the appropriate solution.
What is a Single Customer View?
Even a simple google search would reveal an assortment of definitions for the Single Customer View. Unpicking just a few examples, would result in very different solutions.
One is a visualised solution, one a database outcome, and the third a process. None of these solutions are inherently wrong but would any of these meet your company’s needs? Let’s be totally frank here. Do we really mean ALL when saying all? Do we really mean a SINGLE record when we use the term single? Do we really mean ONE when we say one? And when we speak in terms of aggregated, consistent, and holistic – how do we intend to measure these?
Begin with ‘Why?’
A seemingly simple question that has caused me no end of headaches when running large data operations has been: ‘how many active customers do we have?’. Is this beginning to sound familiar? When asked this or a similar question have you proceeded to ask your stakeholders to define ‘active’. “Go on”, you say, “define active”, ….. and the rest is history.
Definitions are really tricky. Just look at the trouble we have seen with defining when a death is due to, with, or related to Covid. Any definition that was used was found to be imprecise and riddled with contradictions. The definition that was eventually formulated (death within 28 days of a positive Covid test) was a construct that most people were able to accept as an imperfect but common working measure. It was practical and broadly consensual. It was good enough for its purpose of providing the public and government with a simple to understand metric on which to base decisions and influence public sentiment.
The key questions are ‘why is a definition needed?’ and ‘what will be done?’ when information based on that definition is produced. Returning to our ‘active’ customers’ example, it is often the case that different stakeholders will apply different definitions. Finance teams might want to know which customers are billable i.e. subscribed to or using a service, or who have purchased a product in the last n days/weeks/months; Customer Journey teams might consider as active someone who has been engaging with the company’s website in some way; Sales Teams (incentivised on acquisition) might define an expression of interest as an active customer or perhaps apply a strict definition of ‘active’ in order to credit a sale as coming from a so-called ‘new’ customer. All or any of these definitions could be valid for a given purpose. Would we therefore expect a single customer record to hold all of these? Or might we perhaps want actually to view our data through different lenses depending on the purpose for which we need to interpret the data. While these views should be contextually consistent for the purpose for which they are used (provide the same answer to the same question) they are not necessarily aggregated nor even holistic. They are certainly not a single view.
And what is meant by holistic? Is the SCV to include outputs from deep data and extend to include all data points found in big data? Would the SCV really aim to consolidate into a single record all the information the company has about a customer – every interaction, comment, purchase, web/social engagement, cookie, or detail of all the times they DIDN’T buy, click, or follow; information about preferences, attitudes, influences, and influence; situational vulnerabilities, derived sensibilities, particular risks; and more.
Change is inevitable and innovation is a necessity
The essence of data usage and insight provision is its inherent unpredictability. Change is inevitable and innovation is a necessity. If nothing else, this past year has proven beyond doubt that today’s data needs are different from those of yesterday and almost certainly unlike those of tomorrow. The demand for adaptation has accelerated and the market is unforgiving of inflexibility.
This was really brought home to me on the morning after London and the South East of England were suddenly placed under Tier 4 Covid restrictions and people were instructed not to travel and to stay at home. I woke up to an email from a well know travel recommendations platform cheerfully informing me they had found me ‘weekend getaways near you’. I have no doubt they know a lot about me and will have invested heavily in developing a SCV. However, on that particular morning it seems marketing campaigns could not be quickly tuned to exclude people who, by virtue of their locale, really should not have been targeted with weekend travel recommendations.
Do what’s valuable, not what’s futile!
Setting the creation of a Single Customer View as a strategic goal is often a futile and unattainable objective. There is no one view of the customer which is of value in all contexts. Focus should be on identifying the real business needs and desired outcomes – these are transient and contextual. The key to success is in ensuring that data resources are interoperable in ways that can rapidly, and flexibly realise the value of deep and big data.
I have seen many SCV or 360° view initiatives flounder as they stalled behind ever increasing data engineering backlogs and operational problems as they sought to pipeline and model vast quantities of data, derivations, inferences, aggregations, snapshots, data science outputs, reporting requirements and still failed to keep up with rapid market changes and immediate and pressing business needs.
I advocate clarity of focus on the ends and not the means. Why is the data initiative needed? What will the end user do with the data / insight / report /data application once it is delivered? Once these are identified (quickly and succinctly), and only then, attention turns to solutions and the application of well-tried data practice combined with principles of software engineering, contemporary technologies, and data science.
Article written by Avi Marco, Chief Data Mentor
In upcoming blogs you’ll hear more from us about Ammonite’s approach to delivering successful data solutions, an approach that combines lean product thinking with data expertise.
At Ammonite we mean it when we say, ‘Data Done Differently’. And we continue to look forward to discussing ideas with you.
Get in touch
To find out how we can help you achieve ROI and gain competitive advantage from your data, contact Ammonite.