11 Oct Putting Data To Work
It’s estimated that 90% of the world’s data was created in the last year, which means growth in data volume is exponential. Many businesses have mountains or oceans of data that just sits still.
The advent of open banking and the imminent arrival of PSD2 has had me thinking about the treasure trove of data that banks currently sit on without really putting it to work, compared to say the way tech giants like Apple, Google or Amazon use their data for uplift in customer experience or revenues.
I’ve long believed that the tech giants, the customer centric businesses of Apple, Amazon et al, would eventually step into the Banking sector in a big way, and now they’re meeting with regulators.
Amazon already lends $Billions, Apple is ravaging the payment transfer market, and PSD2 is about to make it infinitely easier to decouple bank consumers from traditional banking infrastructures. In the coming months it’s conceivable that you could see your money being held with Barclays but your banking experience with some start up or the big tech companies that have integrated with the bank. #FinTech start-ups and banks will need to hurry their innovation along or face being overwhelmed by tech giants with cash, appetite and the resources to leapfrog them.
I’m a Monzo card user and love the service and the experience the company is creating. They have an API that drives innovation externally, which has long been a solid strategy for many, and quite rightly they’re focussed on growth and their core product offering – current accounts.
With a little time on my hands this week, I decided to take a look at this API and see what value added opportunities I could find in the data. I’ve spent a lot of time this year thinking about ways to analyse personal spending data for a variety of reasons. Not least of which is we rarely analyse our spending behaviours until it’s started to run away from us. I use my Monzo card the way I use cash – to keep a lid on spending.
I thought, “wouldn’t it be great if something as easy to use as Siri or Alexa could analyse these things and proactively make recommendations?” So I set about testing a few scenarios with Alexa, and the initial results were interesting.
I started with two simple goals:
- When should I next top up my card? (Monzo doesn’t pay interest on balances so there’s little point in having a stash of cash sat on the card)
- Where am I spending frivolously? (Could this be better targeted towards personal goals)
Here’s the result:
Here’s how this works:
When I need to top up is a fairly simple calculation that looks at spending volumes over a period, takes account of the current balance and adds a suggested top up amount.
No. 2 is slightly more complex, and I have to admit to taking a shortcut for the above example. What it should be doing is looking at spending patterns based on transaction amounts (small similar amounts), vendor names and categories,and then identifying patterns. As a shortcut I’ve asked it to look for coffee or cafés as I’m partial to a cappuccino, so I knew roughly what to expect. It’s then analysing transactions related to coffee and looking at pre-determined tolerances for transaction volumes, frequency and values, and producing an output based on the results against those tolerances.
Both of these can be achieved with some straightforward maths or by using TensorFlow to do a little pattern analysis. NLP or just structured taxonomies could also be used for entity recognition.
The first point here being, the banks are sitting on a treasure trove of data that is probably going to waste, and the big tech companies know this, hence their interest in PSD2 and deregulation. The advice given to me in my second example could just as easily be to pay off a mortgage early if that was a personal goal. Organisations of all sizes are sitting on data goldmines they just need to tap into in a different way to unlock the value.
The second, more important, point is allowing others access to your platform and tools can unleash innovation far quicker than you can yourself. Digital adoption and growth has been fuelled by integrations – the API is king! Many large companies do this and then acquire tech or teams that look promising, or may just copy what others have done with their APIs.
Last of all, this maze of data is perfectly suited to modern AI technologies, and the conversational user interface. Now is the time to be experimenting with this new interface paradigm, as this #VoiceFirst tech wave will accelerate towards you faster and harder than it’s predecessor did.
If you’d like some help with an AI or Voice application please get in touch, I’d be happy to help.