One of the main challenge of a successful digital transformation is customer centricity: understanding what the customer wants at every second and be ready to deliver him the best service accordingly. To achieve this, companies need to know better their customer, implying for them to gather more information about him, or at least in a more effective way.
That is where the famous “Big Data” idiom pops up immediately, bringing with it at the same time the challenges of volume handling and data quality on one hand, and the privacy aspect on the other hand. Recent news about China hackers having stolen indentifying information of 4 million US federal workers makes the topic very actual by the way. As a consequence, a company using obviously customers data could be very quickly seen as “big brother” and would suffer of a negative image.
On the other hand, we see the value of Big data while improving daily life of smartcities inhabitants like in Songdo or savings lives by using social media info to react more accurately during natural catastrophies. So, why not profiting from them as individual in our daily consumer life as well?
We then reach a dilemma: should we use customer data to offer him better service, or shouldn’t we? Is it by the way profitable?
Is there any value in gathering customer data?
There was a buzz few weeks ago based on the fact that major actors in some channels Today are actually not coming from this area: Amazon has never been a book shop but sells Today more books than any bookstore; Uber has never been a Taxi company; Google enters in the Insurance market without having any background in this area.
How can they do this? What is their strength to be able to fight against well established specialists?
They “just” manage to gather the right information and to combine them with efficient digital processes.
So, yes, gathering the right data, and in sufficient volume to guarantee some financial effects has some business value. BCG explains it very well in their report: “Big Data’s Five Routes to Value” but I would rather advise interested people to go through this article from CIO.com : “8 ways to make the most out of your customer data“.
Investing in this customer data may have many purposes and therefore many benefits. The most obvious argument, used by all is customer profiling to deliver better service. But you may also think on cost savings for fraud avoidance: by knowing better your customer, you may better and quicker notice bad behaviours and detect either fake identities or fraud attempts. You may even anticipate them by using predictive models.
So yes, go on! And you maybe don’t need that much money to start. Start small… and internally!
There is data you already own
Actually, starting through your own databases is for sure a wise approach. Data is already belonging to the company; customer has shared it openly; and investment to gather it is hopefully smaller if your IT architecture has been properly urbanized. Have a look for instance on this interesting article.
Most obvious starting point will be your CRM, if you have one, or alternativelly your customer database. Many information can be already stored there like the frequency of customer calls, that you may combine with period of the year or seasons, for example. Many organisations have by the way multiple customer databases. One rationalisation action should become mandatory for obvious reasons. And then, the challenge will be to connect your business applications databases to it in order to 1. gain in data quality by aligning similar content to the right value, and 2. increase the volume of connected information about your customer (and his relatives!).
But in when we speak about Big Data, we use to consider the “3 Vs“. With the actions considered above, you will get volume and variety. But the kind of information you’ll get will be for most of them not that volatile. Volatility will actually come from another source, usually not that much used:
The way your customer is behaving, for instance, while visiting your website brings lot of information about his interests and his intention. Companies may exploit such information to gain in accuracy on setting up their webpage or delivering the right content to its’ visitor. Huge progresses have been executed in the past years in neuroscience and open doors to many opportunities. For the ones interested to explore further this area, I would suggest to have a look on the work of Netway. Their founder, Marc Van Rymenant delivered useful explanations on the approach during a very interesting presentation at Google Think Performance in Paris last year.
And then, it will still be time to think about external sources. The challenge will then to get customer acceptance.
Get commitment from the customer!
Today, most of people are aware about the danger of sharing information over social media and are therefore very sensitive when they notice companies are using it. But they don’t realise yet how far: generally speaking, if it’s for free, your are the product. I still love using this video to increase even more awareness on the subject.
We understand that people may have some reluctancy to let companies getting information about them. According to Accenture, “nearly 60% of consumers want real-time promotions and offers but very few — just 20% — want retailers to know their current location and even fewer 14% care to share their browsing history“. On the other hand, Infosys noticed that “70 percent of Americans are willing to spend an average of 13 percent more with companies they feel provide superior service“.
So all in all, consumers are willing to get more, but seem not ready to give in return. Really?
Actually, by searching a bit further, it appears quickly that people do bother about their privacy, yes, but not on every area (see BCG report: Earning Consumer Trust in Big Data: A European Perspective) and, actually, “only” 44% of them are really concerned.
Then, the issue is more based on communication. If the relationship between the consumer and the seller is based on more transparency, if the customer is aware on what he gets in return of his data, most probably he will likely be more willing to share. Some trends tends to prove this: the “pay as you drive” insurance concept being quite standard now in the UK, or the rising concept of Quantified self that people are applying on a full voluntary basis. In the same way, what about getting a better customer experience while visiting a shop in return of providing some information about your consuming behaviour? It is not only in movies, it’s for real like CGI tends to demonstrate it via the implementation of a showcase of shop of the future in Lille (see press release – only in French).
The demand is there. Customers want companies to deliver ore services. Accenture has even a word for this: the “Internet of Me” instead of “Internet of Things”. We have to reply to it.
Regulations may remain a topic. But for how long? Today even the United States start to get involve in the Open Data approach. With Bluemix, IBM & ThinkData are pushing quite right in this direction. Of course, data privacy will keep its limitations and that is good so. But it should still let a lot of options to gather enough information.
The right approach could maybe be to have in mind very basic rules on data management:
- Be more selective on data sources, taking only the ones customers don’t bother that much;
- Be transparent by explaining to the customer what you use and for which purpose;
- Guarantee protection of the data collected (like for instance this French startup, Bress, setting up a data Cloud for Health information)
Adapt your infrastructure
From a technical point of view, some choices have to be done to put all these in place. Having the right tools is of course a key success factor that is already recognised by all. If IT fails, nothing will happen. That is for instance what Cap Gemini reveals in its’ study: “Big & Fast data: the rise of Insight-driven Business“.
Many companies, like Forrester, advise to setup a data lake in which it may be easier to gather information. But there are some pitfalls that Gartner has identified and that IT needs to keep in mind.
The main tool that everybody knows Today is of course Hadoop. It has become a kind of standard and for this reason should probably appear in any shortlist during the selection phase. But this is not the only one: With the recent investment decision of IBM, followed by the announcement of Amazon, we may foresee Apache Spark becoming a strong challenger, thanks to its promising performances.
As Tableau mentions in its “Top 7 trends in Big Data“, the way you’ll reach data may still be a subject of discussion between SQL and NoSQL fans, which should be delegated to appropriate specialists in your organisation. At the end, only result counts.