It is time we talked about declared dataPublished: Wednesday, 31st October 2018 by Matthew Tod, Head of Analytics, D4t4 Solutions plc
Declared data I hear you groan, what is that?
It is personal information that an individual willingly shares by filling out a form, answering a question or taking a direct action. It is the first party data that you freely and knowingly pass over to an organization as you interact. Examples in digital channels would include cookie consent, forms submitted, product reviews, survey answers, chat text and more.
Declared data is the third leg in a comprehensive data collection strategy for mobile apps or websites. Initially the focus was on capturing behavior, the swipes, taps, clicks and page views that are generated when consumers interact with digital channels. To this core dataset, data related to the experience is added, such as page speed, promotions offered and price presented. Declared data is the final element; it is everything a consumer writes, says or chooses, for example preferences, feedback or answers.
If you are not collecting all three types of data, in a single data model and across all digital channels, all your advanced analytics, personalization or algorithm-powered tasks, such as scoring models or fraud detection, are unlikely to be optimized.
Those of you who read my last blog - Personalization MUST mean relevance in mobile banking apps - might recall my frustrations when my bank failed to make a relevant response to my declared data. Despite overtly searching on numerous occasions for a child account for my daughter, my bank continued to promote irrelevant products to me. In scenarios like this, there is a clear opportunity to use declared data to gain an immediate benefit - the simple answer to a simple question. If a customer is browsing the savings account section of a bank’s mobile app, it makes complete sense for the consumer to be asked “are you looking to apply for a new savings account this week?” and to be offered a very clear “Yes, this week/probably next week/ later thanks” choice of answer.
Why is this important? Well it is clearly an actionable piece of data from a customer who trusts the bank with this information, so an appropriate response would be to help the customer choose and buy. Opening a dialog which enables the customer to decide and indicate their agenda, will always be more powerful than the bank trying to calculate when to target an offer.
Also, this answer is vital for artificial intelligence powered systems because it associates a customer intention with browsing data, a pre-requisite for machine learning techniques such as supervised learning.
So how do you collect this most valuable data?
Celebrus clients use the real-time monitoring capability of the product coupled, with a rules engine, to trigger these questions, collect the answers and interpret the resulting signals.
A simple rule might say:
- If a customer looks at savings account pages for more than two minutes
- then ask the question “Are you looking to apply for a new savings account this week?”
- show three answers “Yes, this week/probably next week/ later thanks”
- record answer to customer record
The key, of course, is to write rules based upon deep insight, that trigger well-crafted questions at the right time, so they are not annoying. Testing helps achieve this, and should be part of any plan to accelerate capture of more declared data.
With this data in-hand, the path is clear to deliver better customer experience or deliver a whole range of data powered outcomes.
Find out how Celebrus enables leading enterprises to Capture, Create, Connect and Control their customer data. Click here to watch our 2-minute video.