Big Data has evolved from a buzzword to an important concern in every executive's strategic orientation. The traction is real, Big Data is living up to the hype, and everybody wants to use it--even if they often don't have a clue on where to start.

According to this year's GE Global Innovation Barometer "70% of businesses see big data as critical to optimize business efficiency, 61% believe it will be a real challenge to implement. For those who utilize it, 69% see added value for the innovation process".

Jean Baptiste Bouzige CEO & Founder of Ekimetrics--a niche strategic Marketing optimization and Data Strategy consultancy firm--says that "fields such as analytics and data science are rapidly growing in budget and interest within established firms, it is promising, as long as business executives don't repeat the same mistakes they have been doing since the CRM breakthrough in the early 2000".

Lessons from "CRMania" for the BigData enthusiasts

According to a recent study from Gartner, the CRM market is still growing and will reach a $24 billion record this year. At this level of investment, organizations will need to extract as much value as possible out of their systems.

However, many in the data industry have forgotten the "CRMania" that started a decade ago. People seem to be making the same mistakes with Big Data that they made with CRM: focusing on tools and software instead of understanding the needs of executives and business teams for accurate and actionable analysis.

Bouzige points out that "a CRM strategy can't be based on a software integration or customization [...] CRM means customer relationship management, it is unfortunate that it became a generic term to designate systems in charge of targeting campaigns, implying a general confusion between methodologies, databases and tools".

Today there are many technical options on the table when it comes to BigData, these include: Hadoop, NoSQL systems, data lakes and graph databases. Understanding the different alternatives is important but the core objective should remain to architect scalable information systems capturing meaningful intelligence, even from unstructured information.

Volume and automation block out the noise

"A major industry leader was not understanding why the best tool on the market was not able to perform well" tells us Bouzige, CEO of Ekimetrics "we found out after investigating the process, that the datasets behind the tool were polluted with bad data quality". Choosing the "best tools" or the most "popular ones", is not enough. The main challenge is to implement tools efficiently and react properly when they are not delivering in the right way.

High-tech, banking, insurance, telecommunications, pharmaceutical, consumer goods, IT manufacturing and IT services vertical industries are the largest spenders on CRM, because they have a large volume of Data and need automation. These industries are thus naturally the most inclined to embrace the BigData wave.

"The whole point is to bullet proof the process by understanding all datasets for a better business accuracy, the automation comes afterwards, too many firms rush themselves into automation before getting this first step properly done" considers Bouzige, "I heard too many CIOs say: no matter the noise in Data, as long as we have volume and automation"

Scoring: The numbers speak for themselves

One of the big advancement implied by the CRM industry innovation was the scoring system. Bouzige explained to us that "a score is a way of calculating a probability to purchase a product."

Scoring is an efficient way to reduce the target size for a CRM campaign, and therefore an important lever for cost effective oriented strategies". Out of the many CRM features that CMOs fancy, there is the propensity scoring, it is used to optimize the operational task of targeting; "this kind of statistical method is not easy to use for enhancing business understanding" reports Bouzige.

As a result, propensity scores are only bringing a relative strategic value and can be misleading. "I have seen experienced senior marketers from a top tier worldwide retailer, buried under a library of hundreds of scores. They were unable to draw a simple mapping of key customer's profiles and were therefore struggling to build a frequency-oriented strategy based on analytics"

The problem with new Big Data techniques is that plenty of algorithms have the same downside as scoring: they are powerful but difficult to link to a strategic business understanding. "The risk might be to miss a breakthrough while focused on tactics instead of strategy. In the worst case scenario, you might miss your target while blindly relying on complex algorithms instead of expressing your business sense".

To avoid the traps of the "The numbers speak for themselves" belief, one should remember the '90/10' rule of Avinash Kaushik, one of the main analytics evangelist: "put 90% of your resources on human and only 10% on tools". Once understood that, a step by step approach based on customer knowledge discoveries is easy to deploy, because every IT/BigData implementation is backed with a deep understanding of usages and constraints.

The competitive advantage will lie on the human factor

The Big Data market is predicted to be valued at $100bn by 2020 according to Rob Bearden, the CEO of HortonWorks, and half of this will be driven by Hadoop the open-source software framework. "Open source isn't trivial, it is a paradigm shift, and the biggest point of differentiation between CRM and Big Data" indicates Bouzige "since most of BigData competitive tools are open-source, you don't need to rush and overpay software solutions which doesn't fit to your business."

What you need, is the right human resources to understand your objectives, in order to switch from reactive analytics to proactive customer interaction. Many organizations lack a coherent, manageable structure for the data they're collecting, Big Data technologies will help structure these, even though the value will come from analysts who can step back from the tools and combine accurate methods.

Bouzige concludes "At Ekimetrics, we have deliberately chosen not to address the Big Data issue through the lens of tools, because we believe that 90% of the added value and thus all the competitive advantage will lie on the human factor". After all, it is the human interpretation and strategic decision-making that truly empower big data sets.