LinkedIn is a brilliant learning resource. Recently, several interesting articles made me reflect on my career experience.
One was by Don Pepper – When to trust your gut – definitely worth a read. It concluded:-
The lesson is that a gut decision will probably serve you well when you already have a great deal of experience dealing with the issues involved. If, on the other hand, you don’t have substantial experience in something, then data and analysis are more likely to provide the answer.
The second was by Ophir Tanz, CEO of GumGum, an Artificial Intelligence company with particular expertise in computer vision - "Why the future of deep learning depends on finding good data" Seems that bad data in = bad data out. Obvious, really - but I guess what we aren't so good at is spotting bad data.
In my career, I’ve come to realise this: when you have no experience, you have to carry out data analytics using accurate statistically significant detailed datasets and, at the same time, associate this with less tangible observational data – what you see, what you hear. Actually, it’s more than that; it’s going through the raw data and developing associations from it – understanding how what you see, hear and smell is reflected in the data. Developing these associations means good decisions today and valuable experience for the future.
Use guesstimate data, or interpret data in isolation from what you see, and you’ll get bad decisions today, missed opportunistic associations and poor experience for the future. Bad data in = bad data out.
How I learned this
Maybe almost 20 years ago, I wrote an article as a sales manager expressing my frustration that my MD, who was in his late fifties, could somehow know ‘from his gut’ what was going on – often before I did – without reference to my myriad spreadsheets.
Some of my spreadsheets/data analytics were particularly innovative, I felt, and useful. For example, using the telematics data from sales cars combined with company fleet ownership data, I identified numerous major industrial areas not visited at all by sales people for an entire year, leading to focused campaigns in the following years. (Even though, when quizzed, salespeople would argue vehemently that they had been to the ignored parts of their territories.)
Here’s another favourite of mine (I’ve changed the names to protect the innocent).
At the time, I had a team of field based sales people supported by an internal appointment making facility. When discussing performance improvements, a typical reason given for under-performance was, “Well, I don’t get as many leads from Tele-Appointments as Dave does.” The MD would tell me this was the result of the under-performers not communicating with TeleApps. I would contest that I’d witnessed – with my own eyes – communication happening on field accompaniments, and the sales people vehemently denied any allegation they were not collaborating with the TeleApps team.
To prove or, as it turns out, disprove my point, I went through mobile phone bills and produced the above – clearly showing that under-performers did not communicate and connect with the TeleApps team. I also realised, from the billing data, that Dozy and Titch only called in while on field accompaniment, and that Dave, Dee and Mick started calling from 08:00, whilst, typically, Dozy and Titch would make the first business call around 09:30, etc. etc. etc. [Are the etcs necessary?]
Of course, on reflection, the MD didn’t have or need a spreadsheet. The TeleApps team was in the same office as the MD, and he’d been quietly ‘soaking up’ what was going on just by listening to activity. He rarely heard anyone say, “Hello Dozy”, but frequently heard “Hello Dave”.
What I learned from this and other similar pieces of data analysis and observation was that when a salesman puts forward a more compelling argument for something than they would to a customer for the product or service they are selling, then something is wrong. In fact, this is just one element that you build up over time through observing different sales people, which ultimately enables you to sense successful sales behaviour. Indeed, being involved with Telematics, I came to associate driving patterns with successful sales performance. But that’s a different story.
Gary Klein identifies a similar effect with his Recognition Primed Decision Model, coming from the study of decision making under high pressure conditions.
So, on balance, I agree with Don Pepper’s conclusion and Gary Klein. There is a connection between data analysis compared to observed behaviour and building valuable experience whatever industry we serve.
Looking back, I guess I was lucky to have my spreadsheets given the acid test of experience.
Don Pepper - When to Trust Your Gut