Building Effective IT Strategy - part 2

In yesterday's post, I introduced three elementary categories for IT Strategy:

1. Product 
2. Portfolio
3. Transformation

I then pointed out that each of these follows a basic cycle of:

1. Determination of Strategic Approach
2. Goal-Setting
3. Assigning actions to goals

So, now as promised, I will use this to take a look at a specific scenario. But before we begin, let me mention an important caveat to the three categories listed above. While all strategy tends to fit within these, not all strategy has to exist at the highest level. In other words, there are various levels of Strategy that are still abstract enough to remain differentiated from Tactics.

In this case study example, we are going to look at Big Data. Big Data is something that many organizations might consider important enough and large enough to develop a strategy around. However, just looking at the moniker "Big Data" and one might instantly wonder - well, doesn't that belong as part of a larger "Data Strategy." Yes. And then wouldn't that also imply that the Data Strategy would be part of a larger Strategy. Again the answer is yes - and in this case the whole thing lines up neatly like this:

• Portfolio Strategy
•  Data Strategy
• Big Data Strategy

So, this begs the questions, just how we might first differentiate the lower level strategy from higher level strategy and then perhaps even more importantly, how do you ensure they stay in alignment?

Differentiation:
This tends to be managed something like this - you begin at the top level with the superstructure of where everything is supposed to fit as well as common capability / design principles and objectives. Then as you move down the Strategy levels, things progress from conceptual expectations to logical descriptions. The top level is the most open or flexible; the bottom level the closest to expectations regarding solution execution.

Reconciliation:
Side by side with the differentiation activity is integrated road-mapping - each level below fitting neatly into the one above. The other big component here is alignment of Strategy with Architecture which provides the other key reconciliation tool (if used properly).

So, how would a Big Data Strategy fit into an Enterprise Data Strategy? First, it obviously in most cases extends something that already exists. This then implies either replacement of existing capability or addition of new ones. Now let's jump back to the process we mentioned earlier - Step 1 - assigning portfolio strategy is complete. How would we attack goal-setting for Big Data?

This by the way, is where perhaps more than half of the organizations trying to adopt Big Data solutions are getting tripped up right now. A poor example of goal-setting would be - "let's do a POC without a clear path of how to exploit this technology yet" (mainly because everyone else seems to be doing it. A better approach might be:

1. Define the set of possible Use Cases associated with your organization (where Big Data might make an impact)
2. Choose one or two that can be effectively demonstrated and measured - let's say one might be the rapid development of a user driven BI solution based on unstructured (web/social media) data. 
3. Develop a clear path as to how; a) the initial capability could be rolled into the larger existing ecosystem to avoid silos or solution-fracking and b) add new functionality to the emerging Big Data solution - consistent with overarching organizational goals. 

Step 3 is assigning actions to goals. We'll take a look at that in our next post...

Copyright 2014, Stephen Lahanas


#Semantech
#StephenLahanas
#TechnovationTalks