Putting Inside Out Analytics Into Practice

A few weeks ago, I published a blog called Inside Out: Flipping Your Big Data Analytics Perspective. In the blog, I discussed how the business processes that will ultimately improve the organization’s business objectives should help prioritize IT and Analytic initiatives, versus the other way around.

Imagine an organization has IT and analytical teams that continuously do things that are “cool” or “innovative”.  Sure, what they produce makes for great presentations, but, at the end of the day, if they aren’t used and implemented to improve a business process then what was actually accomplished? This type of misalignment, which happens all the time at businesses large and small, causes massive amounts of inefficiency and ultimately lost opportunity and revenue.

As a recap, IT and analytical priorities should start with the business objectives, the opposite direction of the actual process flow that happens in most businesses today.

1.     Business Objectives:  What are the most critical business objectives?
2.     Business Processes:  What are the processes that can best drive each objective?
3.     Business Decisions:  What decisions are needed to feed the process?
4.     Analytic Processes:  What analytic processes, aggregations or business rules are required to feed each decision?
5.     Data:  What data is needed for each analytic process?

To illustrate how this might work, suppose an organization wants to develop two new business processes.

1.     Improve Email Conversion by Creating a regular customer email with product recommendations
2.     Improve Retention by Developing a call-center retention program

Improve Email Conversion:

The analytical, IT and business folks can now map out the ideal process for this task. I recommend you do not constrain yourself with what is available or possible, just what would be ideal. The process might look something like this:

Now you can quickly visualize the components that are needed, and then map that against what is available.

Improve Customer Retention:

Now you can do the same thing with the call center retention program, by starting with the business objective and working your way out to the data you need to achieve the most ideal result.

Bringing it all together:

Now that you have combined both processes you can quickly view possible synergies, and prioritize your efforts.  Here are some of the takeaways from this example:

• Prioritizing Data Acquisition Initiatives:  Suppose you realize that the clickstream history is not available. Well, you can see that this feeds two analytic processes (Customer LTV and Customer Churn) that ultimately are needed for the corresponding business processes. This data source is critical for both business processes, and therefore should be prioritized as high.
• Working efficiently: Often organizations will have separate analytical teams that will independently build, implement and maintain different customer churn models for these two processes. And, yes, I understand that there might be slightly different requirements. But, wouldn’t it be more efficient to build a single solution, creating a robust solution that meets all requirements?  It would then be a “single source of analytic truth” and ultimately it would be a much more powerful and flexible solution since many teams would have input, and ultimately easier to maintain than three separate solutions, right?
• Coordinating efforts:  Similar to the above, couldn’t the business teams then coordinate on all their decisions?  For example, the Customer Value Segment feeds two different business processes.  Why can’t they get together and build a robust set of rules ensuring that the teams are making decisions off of the same infrastructure?

So what to do next?

1. Identify your most critical business initiatives and then the business processes that feed them. This will need to be done with key business stakeholders. Some of these processes may already exist in some form – others may be brand new.
2. Then brainstorm with the business, analytics and IT professionals what each process would ideally look like, without any consideration of what is currently available or what is even possible. Nowhere in this session should the statement “we can’t”, “we don’t” or “it won’t work” ever be used.
3. Map out each process – identify synergies, and opportunities, and begin to prioritize the initiatives. Include all teams in this discussion (IT, Business and Analytics), as now you can map what is and is not possible, and the efforts it would take to deliver each one.
4. Finally… Become exceptional:  Just imagine how effective the organization could now become? By having the business, IT and analytics all brainstorming, identifying opportunities and agreeing on prioritization, the possibilities are enormous!

I will be giving a deeper dive session on this topic at Teradata’s upcoming Partners 2015 conference on Tuesday, October 19. My talk is titled “Becoming Analytical Innovators: What Analytic Leaders are Doing.” If you happen to be there, please feel free to introduce yourself.

Originally Published on Oct 13, 2015 by TeradataVoice

http://www.forbes.com/sites/teradata/2015/10/13/putting-inside-out-analytics-into-practice/