Big Data: the end of long cycle rollouts

We recently shared the findings of our global thought leadership study – Big & Fast Data: The Rise of Insight-Driven Business.   

I think that there is one thing the market agrees on – that big data has entered the business mainstream. Different industries find different value: from one end of the spectrum as a cost reduction and efficiency tool to the other as a game changer to change the entire market segment.
What is success for one is clearly not success for another and there is no Big Data “panacea” for growth – every business is different, and there will be much discussion on each of these axes. 

So what does Big Data adoption look like?

One key finding is that current analytic development cycles are way to long – we know that the business leaders want the pace to match the market opportunity, but that they feel held back by IT. There seems to be an emergence of a new bravery from IT for big data – a mind-set characterised as:
“quickly building proof of concepts, “failing fast”, and then, where value is found, scaling rapidly – in weeks and months – rather than launching multi-year programs”

This echoes my own experience with clients – the adoption track is becoming clearer – at least in 90% of cases. Business stakeholders are looking for early, proven value, not future, promised value.

The report findings introduce this adoption approach as:

  1. Proving value – the most successful programs start by addressing a few real business use-case opportunities.
  2. Expansion to pilot: Companies often choose an entire line of business to migrate to the new approach
  3. Enterprise adoption and uptake: Migrate specific business units or functional areas, with a focus on expanding use cases and enhancing platform capabilities.

Platform, what platform? 

This brings us to a central platform question for the Enterprise. In our own TechnoVision 2015 perspectives, there is a clear exploration of the emergence of the 3rd platform and the shift to a digital platform; as explored in platform thinking.
The end goal; the scale-out environments in which the business will need to deploy these insights needs to be a day one discussion point; a clear understanding of the deployment platform characteristics for big data workloads needs to be understood.
The business needs to be asking:

  • How do I create functional tests of my use cases – with real data – and at pace?
  • How quickly can I test – my competitors may already be on this track?
  • If we find value, how can we move from proof of concept to production at pace?
  • How do we make the insights as accessible as possible to the business?

The risk is having the business demand in massively in excess of IT’s ability to deliver; IT risks irrelevancy and the business stakeholders will bypass and go direct to those that can execute; a topic I will explore in my next blog.