10 Digital Footprints with Value to Insurance
What are Digital Footprints?
Digital Footprints are the electronic recording of an event, activity or state that captures the fact that something happened at a specific point in time. Having said that, I’ve always found the best way of understanding what Digital Footprints are, is to use an analogy.
Consider you’re wandering along a sandy beach and after a while you look back to see the impressions your feet have made in the sand. If you follow your footprints backwards you get a very good understanding of where you’ve been, the depth of the footprint provides an understanding of the effort you’re making and the distance between strides shows whether you were running or walking. Digital Footprints are a similar concept inside the digital world in that there are plenty of times when, as we’re going about our daily lives, we interact with a digital system or service in some way and the fact that we’ve done so gets recorded. Each recording of an interaction is like a single footprint – it tells us something about what we were doing at a point in time.
So each digital footprint tells us something and by combining footprints together, it tells us something far more valuable.
Let me give you an example. If I browse my Insurers web site they can track how I got to the website, what parts of the website I visited and for how long, any searches or trial purchases made then where I go to when I leave the site – this is all standard web analytics. Now imagine I then call my Insurer about a claim or I instigate a secure messaging request to the support team. In this scenario I’m probably interacting with multiple systems, each of which records my activity in different ways and in different formats.
Insurers are increasingly looking at how they can capture all of these digital footprints, accurately align each to a specific customer/prospect then map the journey taken with them. In the example, my searching around an Insurers website then instigating a messaging session means there may be a link between these two activities. If they are linked, what does that tell an Insurer? Does it show some form of intent, some form of interest or is it part of a known pattern of behavior? Whatever it does show, using this approach we can accurately capture the digital customer journey via digital footprints and enable analytics to determine relevance and insights to drive better outcomes.
10 types of Digital Footprints
So here are ten types of Digital Footprints and how I think an Insurer can utilise this data to better understand a digital customer journey. Many of these could be classed as ‘disruptive’ as they are individually having a profound impact on the Insurance industry.
- Web site visit footprints
- Most Insurers already utilise Web analytics on their websites, so there is a great deal of information already captured about exactly where you go, what you do and how you do it. The information captured is usually about the route you take through the website and about any decisions or non-decisions you make.
- Web applications typically run on Application Servers which have their own log capabilities, typically about the technical environment, which is separate from the Web analytics data
- An example of combining these two is the rich picture you get of not just which option you picked on a website but also the ones you didn’t pick.
- Mobile App usage footprints
- Mobile Apps tend to simplify or restrict the engagement process for a user to reduce complexity, so they tend to capture a different version of the journey. As there are different interaction types available for a mobile app, such as physically moving the device to make something happen, this information also needs to be captured by blending the journey through the app with information from device sensors about how the device was used.
- Vehicle Telematics data footprints
- Vehicles generate a great deal of data about how a vehicle has been driven, as well as where and when it was driven. Vehicle black boxes capture all of this over time to support vehicle usage based pricing models. As more and more sensors appear in vehicles we’re starting to see even more data collected. Simple things such as also collecting the temperature outside the vehicle and whether it’s raining, or not, help to build a better picture of the environment the vehicle is operating in. These yields new insights such as whether a driver drives more slowly, or not, when it’s raining or very low temperature; both of which provide insight for usage based pricing.
- Health data footprints
- Devices, such as fitness trackers, generate a great deal of health data about our bodies and how we use them. Some devices have a huge number of sensors, some less so. Basic information about movement (i.e. steps taken) and how this equates to distance travelled over time is a common base facility. More and more fitness trackers are now capturing other important data such as heart rate, exertion/exercise levels, sleep patterns and the types of exercise/movement we’re making.
- Facilities to capture certain types of health related data as also important. There are web systems used to capture weight, food and liquid intake. These sites provide additional insights into other key health factors.
- Visiting a health professional will also generate more specific health data which is large is range but could include cholesterol levels, blood sugar levels, identification of disease, health concern factors such as dental hygiene or diabetes indicators, physical conditions and treatments.
- Social Media footprints
- With vast amounts of content available, much of it in an unstructured form, the challenge is around finding the portions of it relevant to needs. Finding the social media content for an individual, group, organisation etc. can be quite challenging as identification requirements don’t always mandate enough information to enable the link from the content to an identified recipient (i.e. Customer). Text processing techniques are applied to extract specific content plus generation of associated data such as sentiment.
- Travel data footprints
- Travel itineraries are increasingly popular and available online. These provide insights into modes and methods of transit, future confirmed events, overall time management and with the option of the generation of preference data. Preference data provides insights into what a customer likes/dislikes so an Insurer can use this information to respond in the most positive and sensitive way possible.
- Location data footprints
- Typically geo-encoded content with something like a latitude/longitude set of coordinates plus a timestamp. Recorded over time this shows a highly accurate picture of precisely where something/somebody was, at what time they were at that point and is often accompanied by altitude data (often called a Track). This combination creates a ‘3D’ view of a track.
- Internal application capturing unstructured data footprints
- Many Insurers receive either paper written documents or hand entered form documents – both of which may be scanned with OCR/ICR. Scanned hand written documents are notoriously difficult to extract text from due to the high variation in individuals hand writing styles whereas form data is much easier. Either way, textual content needs to be extracted from the unstructured text then text processing applied to generate insights and data for linking to an identified recipient.
- Physical journeys recorded digitally footprints
- Going to visit an office or contact centre generates a number of physical contact points with people, processes and systems. Where the contact is non-digital, this needs recording in a digitally interpretive way for inclusion in the customer journey.
- External data vendors footprints
- Many third party organisations charge money to collect, collate and distribute data to Insurers. This data may contain information not readily available (e.g. bank account transaction) or an enrichment of a data set with additional attributes or metadata (e.g. estimate of location from an IP address).
- Many non-profit oriented organisations distribute data through Open Data concepts. Whilst generally having low/zero cost models, key controls on the data (accuracy, completeness, timeliness etc.) are often hard to ascertain accurately.
What are the benefits?
As individual sets of footprint data, they have value in that they shine a light on specific aspects of a customer’s journey. Where these individual data sets become immensely powerful is when they are placed in defined sequence to show the customer’s journey across different environments and channels.
Time sequence is the most likely one as it also shows the interrelationship between the channels and environments with resulting actions and next steps.
All this footprint data, when a sequence is applied, highlights not just the customer journey but provides a huge data source which can be mined and used for predictive/prescriptive analytics. Most data mining solutions require large amounts of varied data to enable them to draw out conclusions around why specific actions occurred and how likely they are to occur again. Sequenced footprint data provides a great source of data for mining.
So an Insurer can use all these insights to work out the journey a customer took, why they went on a specific journey, what decisions they made/didn’t make and what appears to be of interest to them. With all this information, Insurers are now able to better target offers of products or services to customers with an increased likelihood of the offer being taken up.