Bigdata is doing really well

Why big data only works well with your own data

Even the largest database won't help without the right analysis. Understanding the data collected is a prerequisite for a targeted strategy. Real-time decisions are key. The merging of data from business intelligence, CRM and web analysis helps with a better view of customer needs. Professional data analysts are in demand - and the trend is rising. How do companies prepare for increasingly complex tasks?

These and other questions were discussed at the Werbeplanung.at Summit, moderated by Eugen Schmidt (AboutMedia) the experts Jürgen Seitz (Stuttgart Media University), Michael Vaclav (.brandREACH), Susanne Koll (Omnicom Media Group) and Georg Kalandra (Hutchison Drei Austria GmbH).

In his keynote speech, Seitz presented a study in which the focus and opportunities of Big Data were outlined. According to Seitz, data in general, but above all one's own data, are at the top of the list of priorities: "Data ownersip is again important to decision-makers," he says. The focus is also on customer centricity, i.e. the customer per se. Because customer-oriented behavior is also the top priority on the agenda, even before more sales. The most important challenge for this is a holistic picture of the customer, including above all the cross-device view and the integration of offline data, including CRM data and, in the future, offline touchpoints as well. That has to be solved in the future in order to get a holistic picture, says Seitz. The willingness to invest is generally high in the companies, above all the priority is on the expansion of in-house capacities and the integration of data silos - this should create the basis for benefiting from big data. Today, economic potential is seen primarily in the personalization of the user experience, which is intended to improve conversion optimization. Of course, the GDPR also plays a role here, companies must obtain an opt-in from their customers. “You need that to guarantee data ownership,” he says. Data leakage is also an issue: Where does data flow to third parties in the value chain?

Data Ownership and the Problem of Silos

In the discussion that followed, Schmidt took up the topic of data ownership. Does it even work to break up the data silos? “It's a very exciting time,” says Koll euphemistically and earns laughter for it: “As a media agency, we're used to dealing with data, but it's getting more and more complex. But you also have to question the data, and above all the quality of the data. “First party data is absolute priority - because anyone who provides data from the company has not understood the subject, says Koll.

With regard to the GDPR, Kalandra emphasizes that telcos have been subject to strict laws for years, now more effort has been added, for example the right to deletion. The data is in-house at Hutchison Drei and used anonymously: “We can form user groups, but we are not allowed to draw any conclusions about individual customers,” he says: The IT systems know about the individual customer, but the data is then hashed and anonymized - the employees don't know anything. “So we're only doing a fraction of what would be technically theoretically possible,” says Kalandra.

"You can use the data you have yourself", emphasizes Vaclav - although in view of the GDPR it is questionable what you can actually use from it. One problem is that some companies collect a lot of data but don't use it and form silos in many places. "Data can rot," emphasizes Vaclav: Ten years later, the data of a 14-year-old is useless - this can be seen in the cinema and entertainment industry, for example, where customers create a membership card and indicate which category they are interested in - interested as a 17-year-old man the customer for action films, ten years later he is a father and would rather see children's films in the cinema. “It is exciting to see how the affinity matrix of individual people changes and what can be deduced from it,” says Vaclav.

Digital as a supplement

Is online the power of big data supplanting the classic? "You always read that print and TV are dead and only count digitally," says Koll: "But that's not true, it complements each other." The question is whether you shout louder or do it smarter to get through to the customer: “I'm for smarter,” she says. You always argue this in front of the customer, but they often argue that the customer's son has specific consumer behavior - a sample that is not necessarily representative. "The way companies think has to change, the breaking of data silos within or with the agencies still has to take place," says Koll.

Seitz: emphasizes that the negative view in Europe is the problem with the efficient use of big data - China, on the other hand, was recently described by the “Economist” as the Saudi Arabia of data.

Wanted: data analysts

And what about the shortage of skilled workers, i.e. the search for suitable data analysts? Education is the problem, says Seitz: Although parents send their parents to relevant courses of study, there are still far fewer data experts than are really needed.

“Many people only do data analysis once”, criticizes Vaclav in relation to the approach: “Data analyzes are only hypotheses, not yet a result. Only through ongoing analyzes and repetitive tests - even with things that did not work before - can we really use the potential. "

“Many call themselves data scientists if they can even know a little Excel,” criticizes Kalandra: But he has to be able to do the art of algorithms. For example, a bakery wanted to know whether a location would be interesting for a new branch - the criterion was whether there were enough families in the area. “For example, I could see that a school is nearby and that the right target group comes by,” says Kalandra: “To do that, I need someone who thinks outside the box and looks at local conditions.” That be also the asset to Google and Facebook: "We are anchored locally and do it specifically for the segment."

“We have to rethink, break down silos, set up hypotheses and then A / B testing to the point,” says Koll: And you have to have the courage to admit to the customer that something didn't work.