Does one size really fit all?

Some time ago in one of my blogs I had talked about the product deluge that was going to hit the market with  solutions in  advanced analytics. To be very honest, it’s arrived faster than I had anticipated.

Not very far in time when the ERP revolution  engulfed enterprises in the west, beginning from Europe to North America and finally Asia a lot was  spoken about how ERP packages were designed and built considering the best practices of  successful companies.

It was not that implementing an ERP system was in any way bad for the enterprise, since seamless automation across an enterprise did bring in a high level of transparency and accountability which  automatically impacted  performance for the better. The only limitation that I quite didn’t adjust well to was the fact that  while every successful company had its own set of best practices   which were so  intrinsic for its  sustaining success and differentiators in the market  ERPs could not serve them 100%. The automobile industry is a perfect example.

Eventually best practices gave way to flexible configurations and customisations to the extent that it is now a global norm  that ERP implementations will have 30% – 40% customisation depending on the industry.  Some ERP product companies have gone to the extent of selling pre-customised products for specific industry segments and guess what they need to be customised too.

I wonder if  we are going down the same path with  analytics too? I really do not understand how will “one size fit all approach” work with a science  that fundamentally focuses on helping industry leaders build unique differentiators in the market. Will the ERP story repeat  here and more importantly are we once again targeting  customers with the same customisation  logic that  created  a multibillion dollar services industry?

Using  an example to explain my point   , let’s take the telecom industry. There are umpteen products in the market that talk about analytics for this industry, each explaining how their algorithms are better at predicting customer churn. So two telecom companies will end up using the same product  leading to the same output – then where is the differentiator? Well for that we will need to get our software customised  for you. Doesn’t that sound familiar? Product companies cannot be blamed here, because they come with legacies of having built ERP type of software and continue to sell their analytics  software the same way.

Analytics is not about running data through a set of algorithms written by some scientist sitting in another part of the world. On the contrary its all about small wins and specialised outcomes each with its own size, colour and style. It is time  the industry moves on  to change the world of analytics for the better, I hereby submit my pennyworth for this larger cause.