DDMA Data Day 2017 Hot Topic Applying A.I. and Machine Learning in Marketing

27 November 2017

Article written by Nieck de Groot, Medior Consultant

DDMA Data Day 2017 Hot Topic Applying A.I. and Machine Learning in Marketing

‘Enabling customer experiences’ is the theme of the DDMA Data Day 2017 at Pakhuis De Zwijger in Amsterdam. “The customer must be happy to do business with you. Data is the catalyst with which to make it happen,”, explains Jan Hendrik Fleury, Chairman of the Customer Data Committee& Talking at the DDMA.

A few years ago the focus was still on gathering as much data as possible, but it has rapidly shifted to how data can be utilised optimally. Or even better, how we can teach computers to do it themselves through machine learning and artificial intelligence (A.I.)..

The topics vary from how important it is to have the basics in good order first to which jobs are going to disappear first once machine learning and A.I. are further developed. It goes without saying that we looked at privacy, the GDPR and how compliance with this regulation is not the same as handling consumer data in an ethical way – which is something algorithms do not allow for.

Simon James from SapientRazorfish showed us how you can already use A.I. to process large amounts of raw data. Using satellite images to assess the productivity of companies, by identifying the number of cars in the car park or the oil level in oil storage tanks, unlocks valuable data, for example, for investment firms.

Have you ever done a video interview (e.g. Skype) when applying for a job? If so, it is possible that an algorithm by HireVue based on image recognition has already assessed whether you meet the personal characteristics required for the role based on body language, facial expressions, voice intonation, etc. before a recruiter has even seen your CV.

This raises issues about how far you can go in improving processes (e.g. in marketing) without this compromising your customers’ trust. Simon James advises companies to adopt a proactive and open attitude, such as by incorporating the company’s position on ethical use of data into their vision, as they did during the emergence of CSR.

The winner of the DDMA Award (KNVB) uses the inevitable football analogy to show how they managed to successfully aggregate into a single database all of the data from membership databases to football scores for all amateur and professional clubs and sales data for ticket and merchandise sales for the Dutch national team. A good example of how having the basics in good order leads to 100% higher commercial results. They also use this data to support local clubs in identifying opportunities and threats.

One of the other nominees, VacanceSelect, applies data analytics by using offline data (booking data) and online data (website visits) in order to give customers personalised results from their searches. With the help of Google Custom Search API, a search for “child-friendly campsites” is converted into a landing page that dynamically displays the right photos showing teams of entertainers and playgrounds.

In addition, VacanceSelect uses predictive modelling to select which visitors are more likely to convert and to only send follow-up marketing to these customers. In this way, data can also be used to stop investing further in certain leads.

This brings us to a form of data-driven decision-making that is less frequently discussed: disproving learned truths based on data analysis results in order to dispel resistance and clear the way to take new steps forward in the digitisation of the organisation. The Marketing Director of Manutan argues that this is also the key to bringing the management and organisation to the point at which they will embrace and promote choices based on data analytics.

Beyond data-driven applications and digitisation in the organisation, the role that artificial intelligence and machine learning are going to play in our society was a widely discussed topic.

Bart Fussel from aFrogleap talks about the impact of the rapid development (including A.I.) that the world is going through, illustrating it with ‘the law of accelerating returns’. If you took someone from 1750 and brought them to 2017, they would be mentally unable to comprehend the amount of technological change. At the current rate of technological change, the same would be true if you took someone from 2017 and suddenly dropped them into the year 2035. What can we do about it?

Sagar Savla from Google makes the point on this that the discussion surrounding A.I. is currently focused primarily on polarisation: either it will be our biggest ever opportunity, or it will destroy us. However, the discussion should be more aimed at how the synergy between humans and computers can help us move forward. A.I. hasn’t appeared from nowhere – even though recent media attention gives the impression that it has. Actually, it has been developing for decades and even now Siri still doesn’t get it sometimes. Just like past innovations where the potential wasn’t immediately clear, A.I. and machine learning too will create new jobs that we haven’t thought of yet.

He argues that it is better to ignore the debate on what the difference is between A.I. and machine learning and to just start on a small scale to discover what these two technologies could offer your organisation. Trial and error is the only way to get results from the new technologies that are going to enrich our lives: that was the conclusion of many of our speakers at DDMA Data Dag.

Click here to see the presentations and photos from the day.