Why you may be the only thing stopping your AI team delivering real insights

Real insights

One of the most amazing outcomes of Data Analytics is that you can quite often discover real insights about your business that you didn’t know.

These insights could be the kind that is intuitive but difficult to spot, but in some cases, they can be real mind bending insights that would never have occurred to you in a million years.

You have to dig deep to find the gold

Using deep AI processes gives you tools to dig deeper than you could ever do manually.

That’s the irony, many of these processes are mathematical and only explore bytes of data to find mathematical patterns.  The machines have little or no understanding of the business processes, customers, people, or the market, other than what it can find through the data. (there’s a caveat here)

And yet that’s the brilliance.  Deep analytics is incredibly impartial as it has no agenda or preconceived bias.

It can find or uncover patterns in your data that aren’t always apparent or visible with the naked eye.  It can also prove or disprove your hypotheses.

The challenge is that some of the outcomes from analytics can be difficult to comprehend or accept at the outset because they aren’t intuitive to us.

Was Brad Pitt a genius?

The movie Money Ball is a great example.

In the movie, Brad Pitt’s baseball team was woeful.  He had to deliver a better team on a budget so went about it with an analytical approach.
The results were, a squad made up of a bunch of mediocre players that couldn’t compete in the big league.

In reality, they were a phenomenal team, when you combined the skills and experience of the individuals.  The whole is greater than the sum of the parts, right?

Remember that Money Ball was based on a real life story.  Remember too, that optimisation has been around for years and has been used by many companies to create products or deliver projects that would appear to be ‘punching above their weight’.

Trust the Black Box

It takes trust in the ‘black box’ to make the magic happen.

This is where you come in.

Giving the ‘black box’ and the team who built it the trust they need to deliver results is not an easy thing to do.  Where the outcomes of an analytical project produce results that, on the face of them, do not look intuitive or sensible, it’s hard not to pull them up.

Importantly, success in Data Analytics depends on your appetite to embrace something that appears to be crazy.

These results would likely have involved a lot of resources and time.  The results you’re presented with would have been tested and verified beforehand.  Robust in the analytical sense is about ensuring there is a high probability of them standing up to scrutiny (mathematically) and being reproduced with similar results given an extended data set.

One of the risks is that you stymie your team when discovering valuable insights by not allowing them to explore at the depth or diversity they need.

There’s a Ying to your Yang

The opposite can also be the case.  You expect your team to deliver results that are simply not possible.

One of the benefits of Data Analytics is that given the right data and analysis, you can produce valuable uplift to your business such as increased revenue, margin, retention, and so on.

But pushing a team while they are trying to explore and discover these results won’t force them to find answers that haven’t been discovered yet.

Analytics projects thrive under a scientific approach where analysts have time to explore and investigate.  Putting unrealistic time pressures on these teams will most likely results in sub-par performance if any at all

Collaboration is the key

Work with your Data Analytics teams and collaborate with them on both their approaches and findings.  You may not match them with their mathematical skills, and they may not match you with your business skills, but together you have all the parts needed to make the whole.

Share this post