I am French, therefore morose, and I like to wallow in bad news. But this state of mind may offer some genuine lucidity. Not my words but those of the New York Times.
In brief, I am French, which gives me a good excuse to continually complain and ruin the atmosphere of the slight euphoria distilled by the champions of analytics. Data is hype, geek is chic and marketing will only be algorithms and correlations in the near future. Great! However, it’s more complicated than that. As a vendor, I’m probably shooting myself in the foot here, but I believe that it is our responsibility to give an honest view of the potential and limits of the technologies in place (or of the technologies that companies are trying to implement).
In fact, I have a piece of bad news: you thought that new technologies (tag management, behavioural targeting) and of course, above all others, Big Data (the Holy Grail, the panacea, the object of every fantasy), were going to come along and spoon-feed us, and allow us to take it easy whilst the hard and software would think for us at the speed of an in memory database. But no, unlucky! You are going to have to rack your brains more than ever before.
(Raw and) Big Data + dashboarding – “everything you have always wanted to know about data but never dared to ask”:
The not so good news: Big data can process enormous volumes of heterogeneous data. We no longer need to format the data. Raw data is the solution to everything, giving us an impression of security as it contains everything.
The worry: The idea of raw data is an oxymoron. Data is never raw, it is always slightly formatted in order to exist (and be useful). So, if raw data allows us to be idle from the beginning, once it is available you still need to ask the right questions and provide the right answers but on a much larger scale. With the raw data as a basis, you need to pay attention to not having to continually recreate segments which can become very tiring and counter-productive on the long term. Raw data is not a dashboarding tool which will structure (or simply rewrite) queries. The link between raw data and the bright dashboards you see on screen will not save you from very complex issues associated with the notion of visits (the 30 minute period and their dedup, a real joy!), visitors, multi-domain perimeters etc. Take shortcuts and you will have a dashboard full of absurdities. There have been countless(1) numbers of disappointments in the field and I fear that they will only increase over the next few months.
Targeting and correlation: “Minority report”
The not so good news: Thanks to Big Data and advanced targeting technology I can personalise everything and easily do one-to-one. I no longer need to think about my message, I can adapt to everyone as I will know everyone’s needs. Just like in Minority Report, I don’t need to understand why the event takes place since I know how and when it will happen.
The worry: This technology can work on the operational level in digital marketing within certain limits (privacy, capping, creative resources), the most relevant example being targeted advertising. However, beyond these very operational aspects, it is clearly more debatable whether this type of technology works when we talk about marketing strategy. If I want to speak to everyone, I no longer speak to anyone and my identity, my brand, become weaker. Since this data highlights efficient and effective correlations, I no longer need to look for the causes. However, if you lose understanding of the causes you also lose the intimate understanding that you have of your market and customers. Your vision is in serious danger as well as your company’s longevity. The correlation is relaxed, but dangerous in the end, and anyway, who wants to live in the world of Minority Report?
Tag management: “The towering inferno”
The not so good news: tag management lets me simply manage all of my tags and I no longer have to depend on the IT dept. These different technologies, whether they belong to analytics sellers or are third party solutions dedicated to tag management, considerably reduce the time required for implementation as well as the technical complexity of tagging. Tagging becomes child’s play and I can focus on analyzing.
The worry: If the IT dept is less involved in the process and the technical work is faster and more secure, the abstract complexity that provides a multi-dimensional representation of your business is only greater. A plumber’s work and a builder’s work is 100 times faster today than it was in the past, but the marketing structure in place remains more sophisticated. It is even becoming increasingly complex because if you are able to build a 15-storey building instead of a 2-storey building you will do it because you are ambitious (or that your manager is on your behalf). The architect therefore needs to be even more efficient, and the architect is you!
I could also talk to you about some of the unrealistic wishes which exist including wanting to implement multi-platform client monitoring without the pivot (2), the essential reference base, or wanting to carry out multivariate testing on all ergonomic choices that are made without the creative resources necessary to have valuable alternatives etc. Here once again I am going to upset the mood.
In sum, the more the system made available by your favorite seller improves, the more your expectations and those of your clients will be raised, and the more you will need to be imaginative, creative, rigorous and meticulous. Let’s be clear, as analytics solutions designers we believe that big data, tag management etc., offer fantastic opportunities and on a daily basis we are committed to making our systems increasingly simple and flexible. This is what we have to strive for: connected data which is simple and easy to understand, making it possible to make quick and useful decisions. But let’s not delude ourselves. Ultimate simplicity and the effective and efficient decision-making which results from it are far from being objectives that are easy to reach. Leonardo de Vinci once said: “simplicity is the ultimate sophistication(3)”. The miracle solution does not exist. Of course there are facilitators, but no magic.
But is this really bad news coming from the French complainer? I don’t think so. First of all, the senior analyst with all of their experience, and ability to consider everything in its context and interpret information, have turned out to be more useful than ever before. Furthermore the intellectual stimulation that results from these new possibilities is rather exciting.
And no matter what, I call the great Pierre Corneille to the rescue to motivate digital analysts faced with the scope of their task: “When there is no peril in the fight, there is no glory in the triumph.”(4)
(1) Even worse if the errors remain hidden and that the decisions made are based on incorrect results.
(2) A general detail which is neatly swept under the carpet whenever the theory of this type of analysis is approached.
(3) Or the same idea according to Bruce Lee: “Simplicity is the key to brilliance”, but the classics need to be brought up to date, don’t they?
(4) In French, the alexandrine : “ A vaincre sans péril, on triomphe sans gloire »