Winning Manufacturing Strategies

« iOS or Android: An 'Ecosystem' Footprint will decide the Winner in Future | Main | Honey, they "shrunk" the hardware, and have put that in a "connected box" »

What is your Business Case for Big Data

For any new trend or technology to be adopted in Enterprise there is a tendency to build a business case on the ROI or TCO model, justifying the cost and the potential returns that can be expected. While this is certainly a strategy that worked for so long for organizations, especially when the technology and the business problems were predicatable e.g. Packaged implementation of a CRM in which both process, technology are known & proven industry wide in each vertical domains.

Both Business and IT in any enterprise can build a ROI model for a CRM implementation with a fair amount of +- accuracy in the ROI or TCO. We witness the first level of challenges in building a right business case where ROI wasn't predicatable, and TCO just over shot budgets every year on year in case of BI implementations. Till date there are large number of debates, topics in large BI conferences that still talk about whether ROI or TCO is a better business case for BI & how to define the ROI or TCO on a moving target. Then came the concept of providing tangible outcomes, benefits and process metrics improvements in shorter time frames (3, 6 or 12 months depending on size of a BI implementation) which appealed to business a lot as they were able to justify their investments and get the desired approvals as well.


The core of the problem with those business cases always was, these were driven largly by the IT of Enterprise, and minimal to negligible involvement of business. The term "Business case" itself signifies that business benefit, improvement or need justifying the technology investment and hence all the more reason to have business define the case with inputs from IT on the Infrastructure costs and implementation costs.Now this challenge gets bigger with Big Data as now we are fighting unknown business problems, high frequency and voluminous data sets, faster decision capability which can have a large impact on the Enterprise business. Some of Big Data problems have infact significant impact on business e.g. Listening and responding to the social media conversations with speed about sentiments for the Enterprise Products/Services, Detecting fradulent transactions when they are happening, automatic diagnostics of manufacturing defects identified from the logs/sensor data before the problems get bigger or result in production losses.


The idea is not to scare you with the thought of no way to define a business case for such a large scale of problem, but to accept the such a challenge exist and has to be dealt with appropriately. Yes TCO/ROI are proven models for business case justification, however these models only help define the future value for a technology implementation or the cost involved if we adpot one. There's a better way to build or present a business case by what i call anti-thesis, "Present what we are losing (say X) if we are not adopting doing (say T)" model. Let me give few examples to illustrate how this model might be helpful in building a right business case, and get wider acceptability across stakeholders in an Enterprise.

Example 1 (Retail Campaign Management) :
Spending "A" Dollars annually on mailer campaigns of "B" products/services to "C" customers (keeps changing). From here two potential cases can be built depending on how mature organization processes and systems are built

Case 1 - Every transaction or customer spend can be attributed as "Reference of Mailer Campaign" or "Not". Best case scenario, as this can certainly provide insights going forward about the Lift in sales or impact because of the Campaigns v/s the cost "A" for those campaigns. Secondly this also tells you the response from customers towards those campaigns - if it goes down, either customers are losing interest or campaign wasn't effective or worst customers have transitioned to competition

Case 2 - If transactions can't be attributed to "Campaign" then also all's not lost, from the date of Campaign the pattern in sales can provide insights about % lift in the sales & over a period of time this can mature to knowing near about when the effect of Campaign impacts the sales. One can always argue that there might be other strategies like Products/Services getting better, or other channels impacting the sales hence difficult to gauge if it's only due to "Campaign"

Example 2 (Automobile Industry Social Media Sentiment) :
Auto industry has a heavy dependency on their retailers or dealers for sale of their vehicles. However one the critical pieces where Auto manufacturers don't get enough visibility is the relationship between dealers and the end consumers buying manufacturer products. This critical missing chain, if handled properly, can provide auto manufacturer a valuable feedbacks, inputs on the services, vehicle quality, features consumers desire or are not too happy with, dealer performance related inputs and many other such valuable information pieces. With advent of social media, this link infact has moved into Social world now and any feedback with regards to services, quality of vehicle or defects spreads quite fast therby denting the manufacturer's reputation & brand value in the market. Several manufactures have already woken up to this reality and started tapping the consumer sentiment being expressed on Social media channels, initiated by Listening services to pro-active response, campaign management and cross-sell offers to turn sentiments towards positive bias. Now the best busienss case to build such services of tapping into an additional channel of social media world requires identifying the brand value loss, customer churn at dealer levels, sales reduction in specific product segments, or defects reported pattern by dealers/consumers directly to the manufactures to name a few. Attributing a value to those, will help project what the manufacturer is losing if they are not adopting a social media strategy.

Thus, presenting, as value, what business is losing (revenue, brand, customers, service, defects, fraud etc) if x->y->z strategies are not adopted, and then translate this business case into technology investments attributing how each business value loss is going to be impacted by those technology implementations and IT strategies.

I would like to hear your experience and business case challenges that you faced or were able to overcome with better innovative ideas to present and gain acceptance with measurable returns.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.

Subscribe to this blog's feed

Follow us on

Blogger Profiles

Infosys on Twitter