Simplifying data science is a dangerous game

Simplifying data science is a dangerous game

There might very well be a place for science to go with adland’s art – but M&C Saatchi’s ‘magic formula’ entirely misses the mark.

M&C Saatchi, an agency feared by peers as much for its chutzpah as its ability, makes a big deal in its credentials about its core philosophy: brutal simplicity of thought.

As these sorts of flag-waving exercises go, it’s pretty good: bold, useful, workable (i.e. it can be rolled out to explain pretty much any piece of work created by the agency), memorable and immutable.

Now I fear M&C has rather undermined itself with its attempt to capture another piece of high ground: data science.

The thinking behind this is fluffy, confusing, simplistic and impermanent – i.e. everything M&C’s manifesto isn’t. You could say they’ve dropped brutal simplicity for something that is just simplistic – and not very brutal either.

You will remember that, a couple of months ago, as it celebrated its 20th anniversary, M&C Saatchi – (tongue-in-cheek, I assumed, part of an elaborate joke) – announced it had discovered the holy grail of marketing: this was the marriage of art and science, the results of which could be boiled down to a magic formula.

Yes, it was all about:

Magic formula

But it wasn’t a joke after all, because now the Saatchi Institute (a think-tank, no less), has produced its first ‘paper’ on the subject, claiming that it is all about maximising differentiation and minimising deviation.

This is based on a project with Nielsen and Unilever, and may (or may not be) about Dove. The equation, apparently, maps the data when it is plotted on a graph.

I am always deeply suspicious of ‘Institutes’ and think-tanks: they are a tactic favoured by dictators and charlatans to give their nefarious activities a patina of respectability.

The Saatchi ‘Institute’, which appears to have neither a digital nor a physical presence and may therefore exist only in the imagination of its founders, thus lives up to all my prejudices. Who runs it? Who’s a member? What does it do?

Anyway, you can watch this video here, or read this short release here if you want more.

Naturally, the initiative has stimulated much discussion and more than a little scepticism. My former colleague Jeremy Lee thinks that it only encourages clients not to take agencies seriously, which is a good point. To which I might add that it does data scientists a disservice too.

My own reading of this story is that it’s a PR stunt that has somehow taken on a life on its own. You can see how this happens: ageing agency looks for something interesting to talk about and make itself relevant again.

It fastens on data science as the answer, sells the sizzle by talking in terms of ‘Newton’ and ‘holy grail’, and then finds itself scrambling to find the meat to go with the sizzle.

Now, if you read last week’s Campaign piece, you will see M&C Saatchi UK chairman Tim Duffy desperately back-peddling.

Indeed, he concedes that the formula, far from being the answer to everything, is not necessarily the “answer to anything”. But it might give a CMO the confidence to say to their CEO “if I spend money on the following, it could have this effect.”

I laughed when I read this. It’s just what media agencies (and management consultants, programmatics, data scientists etc) have been saying for some time.

And it’s never really been true. They can have a good guess, and sometimes they’ll be close-ish, but never with the certainty promised by M&C’s equation.

I think I could write my own equation to cover its efforts: R=MCS (D) + PR x B.

R stands for rubbish, MCS for M&C Saatchi, (D) for added data, PR for, well, spin, and B for bollocks.

Nevertheless, I absolutely believe there’s a place for science to go with adland’s art. It’s just that I don’t think anyone has worked out how to combine the two effectively, even though media agencies (see below) seem optimistic they know how.

Media agencies: the answer to all our prayers…

Last week, Jamie Toward, head of content at MEC, published this interesting piece claiming media agencies hold the key to the future.

Mr T is an old colleague of mine from Redwood, who a) has a thoughtful and considered view of the landscape and b) has never been shy about challenging the status quo.

His thesis, brutally simplified, is this: millennials are becoming more dominant culturally and economically.

They don’t like traditional ads, but they are more responsive to marketing messages that are authentic, transparent and encourage engagement and dialogue. This sort of style of messaging is best served up as content.

Technologically savvy, they like to move seamlessly across digital platforms. And media agencies are the ones who best understand how to distribute this content effectively because it requires an understanding of data and platforms, both of which they have.

Moreover, he says, with the death of the traditional full-service model, media agencies have moved into the creative space, whether through co-operative relationships with, or ownership of, content specialists (e.g. iProspect’s deal with John Brown or Havas and Relaxed News).

It’s an interesting argument. It falls down in one area, though, as far as I’m concerned.

I don’t think they can really do creative content. Yes, they can white-label stuff from content agencies, or buy them, and therefore give the appearance of doing content.

But I don’t think, in their DNAs, they genuinely have the ability to understand the creativity required, or how to turn a data-led consumer insight into a powerful creative proposition in a way that content or mainstream creative agencies can.

Will they ever? Maybe, but so far the evidence is limited. Even if they white-label stuff, they really need to control or own the creative thinking that lies behind the best content. Of course, they can hire the right people, but so far they haven’t (to the best of my knowledge), and even if they do, the cultures are essentially alien to each other.

When push comes to shove, they default to their corporate comfort zone: planning, buying and distribution.

Let’s see if any of them prove me wrong.

Source: Simplifying data science is a dangerous game

Via: Google Alert for Data Science