WHAT IS IT WITH DATA ANALYSIS? – Or, more specifically, why does it nearby always seem to trigger the same response from clients: ‘The data you’ve used is unreliable.’

On the one hand you could easily argue that if this is a stock-standard response you shouldn’t be surprised by it. When it comes to crunching numbers, Pavlov seems to marry Newton: Actions seem to cause (same) reactions. And for me, yesterday was déjà vu all over again.

Five weeks ago I started yet another data modelling adventure for one of my clients, this time involving more than 1.7 million records of disparate data, sourced from three large databases. One set was received from Human Resources, another from Operations and a third from a pool of long-serving business analysis who, throughout the years, created their own repositories.

Other than the magnitude of the entire set there was nothing that could steer us off rails. The three sets could be matched through two types of primary keys, and after a bit of re-formatting the data looked clean and usable. What we set out to create didn’t exist – we were truly on the road to make something new.

Of course, we pre-presented along the way. We checked and re-checked every third graph with managers, analysts and supervisors to make sure that all up’s-and-down’s could be accounted for and explained. Each fifth graph presented a Normal Curve to keep track of unusual occurrences, potential flaws, out-layers and other possible weirdness. We put managers to the task of helping us clean data further and further to finally arrive at a set of just under 1.2 million lines.

We felt like the ancient alchemists must have felt when feeling – no, knowing – that we were onto something quite extraordinary. The data seem to support much of our initial thinking in many ways, thinking our client persistently dismissed as incorrect – or unbelievable at best. The numbers literally stacked up in our favour and we doggedly continued to create ever better models. The further we advanced the more we started to realise how vast and expansive the minefield was in which we’d entered.

One of the first things we realised was that the client had sent its share and stakeholders information that was based on false data. Silently, those who were responsible for gathering, collecting and modelling that data confessed they’d known for long that this was the case. To change it would cause a major uproar as next reports would show graphs that wouldn’t look like anything published before. The repercussions would be severe; heads would roll immediately. It was better to keep things the way they were and to make slight changes along the way. It would take them more than three years to get to the point where they could sleep again at night, if they kept on working on it in stealth mode.

Equally severe: people had been sacked for blowing whistles – for warning top and senior executives that their reliance on all this data was wrong. To some of what must have been the most astute managers it was clear that things were fundamentally amiss with the graphs and reports they had to put their signatures to. Those who closest to the coal mines, who actually observed and noticed what practically happened daily within the machineries of the organisation saw it for what it really was. When their knowledge became an uncomfortable reality to those who wished to remain oblivious to the facts, the messengers were removed and hastily but carefully replaced.

One of the biggest risks to ourselves was to be shown the door without being given opportunities to help the organisation change course. Some hopeful employees who dared to meet with us to pledge their allegiance to what we were doing warned us that we were running toward an Abyss of Doom in which many other consultants had disappeared previously, and how our report could easily become just another fire starter – literally and as a figure of speech.

This truly is one of those points where consulting becomes difficult. It is one of those situations where we become wedged between Ethics and our own Hopes – hopes that see us help employees become more successful, organisations happier places and businesses better world citizens. We know that we can help businesses become exactly that – but also know that all that potential can be yet again suppressed with one quick flick of an executive’s pen.

We’re lucky to have found a client who actually accepts most of our findings – perhaps all after he’s been given to chance to mull things over this weekend. This time we’re dealing with professionals who are courageous enough to see things for what they really are. Unlike what happened in the past, this time there’s change in the air. The presentation went well – after we were told, or course, that the data we’d used was – you guessed it – unreliable….