Dark or hidden data is defined as any difficult to understand, undervalued, or ignored data your organisation collects that offers businesses opportunity for improvement or added value. Dark data tends to come in the form of open (unstructured) text.
Dark data is collected by almost every business, in many cases unknowingly as the nature of an organisation’s dark data tends to mean it’s appended to information that is being analysed, or it exists outside the purview of the part of your organisation responsible for customer analysis. It may also be more difficult to analyse than the traditional data you collect and in a format not supported by analytics tools, so is simply overlooked
The implementation of AI solutions combined with regular database health checks reveals opportunities to integrate this previously inaccessible dark data with other analytics channels to improve customer outcomes, reduce churn, and provide organisations with a clearer understanding of what customers want. It can also reveal a greater insight into organisational issues or lack of engagement in your marketing as well as build a deeper understanding about what data is most useful to collect and analyse from customers moving forward.
Using a combination of Touchpoint Group’s AI customer analytics tool, Ipiphany, and powerful marketing automation platform, TouchpointMX, we’ve isolated the top three most common sources of easily accessible dark data in an organisation and provided an analysis of what you can expect to find when analysing this data:
Most organisations ask the question ‘why did you unsubscribe’ when a customer chooses to opt out of a marketing list. This question takes the form of a list of options followed by an ‘other’ field that a contact may use to explain why they’ve unsubscribed: These ‘other’ fields are a gold mine of information, but the survey as a whole is critical to understanding customer issues like churn.
Our client - an enterprise organisation in the finance industry - captured responses to this question but wasn’t monitoring the results. In this case, an average of 39% of database contacts unsubscribed when they simply needed assistance with something else. At an enterprise scale, this is upwards of ten thousand contacts per year who didn’t mean - or want - to unsubscribe. Customers with queries like “I want to update my email address” could be redirected to update account details and customers asking to receive only account-related information could be added to exclusion lists.
In regulated industries like finance and insurance, this data plays an even more critical role. The ‘other’ field can also contain evidence of vulnerable customers - those who are unsubscribing from a lending service because they’re experiencing financial hardship, or from a bank’s emails advertising mortgage rates because they’ve defaulted on a mortgage. Analysing the text data in these fields is a critical aspect of identifying customers in need and streamlining the process of getting help to those who are asking for it - no matter the scale - is critical for businesses who intend to retain their customers.
Support logs are often used to record issues customers experience and are then closed and abandoned. These data archives can turn into vast repositories of untapped customer feedback, detailing how the issue is presented, what steps were taken to offer solutions, and how resolution was achieved.
Using Touchpoint Group’s AI customer analytics tool, Ipiphany, to read and categorise the unstructured text data in support log tickets, we can identify themes to help an organisation develop greater context around inefficiencies in customer support systems and provide a framework for a more proactive approach to improvements. It can also help to identify areas where further staff training or education may be required, and delve down into the root cause of specific issues that customers experience.
The experience we have with our clients is that this data is only analysed for quality control or post-incident follow up, which leaves clients in the position of finally uncovering the cause of issues months, or even years, after the fact. Importing and analysing support ticket data with a regular cadence will ensure that trends are captured and rectified before they become issues.
No-reply email addresses are often used to convey important information like account updates and notifications to customers. Although these emails may state that their origin is an unmonitored inbox, customers invariably reply. These replies often come from people who lack the technical skills to understand that their replies will go unread, which can lead to frustration and churn. If an automatic response comes back to notify the customer that they’ve emailed an unmonitored address, the customer is more likely to start an ensuing customer service conversation in a more frustrated state as they’ve already encountered a hurdle. If a reply doesn’t come, they may never realise the address was unmonitored.
In addition to providing a more streamlined customer experience, analysing and optimising user interaction points like the no-reply email address can provide valuable information on the themes of new customer enquiries, more efficient routing of information requests and sales or retention opportunities, and even allow for the setup of an alert system to capture vulnerable customers who may be reaching out for assistance.
Does your organisation use a no-reply email address? If so, do you have clarity over the types of emails that inbox receives?
Using AI customer analytics to read and understand customer data - in all its forms - is the key to unlocking the gold mine of customer interaction data that was previously hidden away in vast, unused databases. This information - data your organisation has likely been collecting for years - is the key to customer-centric knowledge that can be used to make considered, data-driven decisions and implement measurable changes to improve customer experience. Ipiphany provides a powerful, efficient way to analyse thousands of rows of unstructured data per second offering immediate, granular results.
The main barrier behind dark data analysis isn’t the data itself, or even the technology required to analyse it; it’s the logical capture and organisation of that data. While Ipiphany’s powerful AI can analyse captured dark data, TouchpointMX can assist organisations in gathering customer data for analysis in a transparent, structured way that breaks down the last of the barriers between your organisation and a sophisticated dark data analysis strategy. Whether you're looking for an analytics, or help uncovering dark data in your business, get in touch - we can help.
Does your organisation have dark data that you’re struggling to analyse? Contact us and we’ll organise a demo with your data to show you the power of Ipiphany’s AI customer analytics.