From the Blogosphere
Innovation and ERP — Decision Making Not Made Easy
It is with some degree of wonder that IT decisions makers are still relying on vague data
By: Lev Lesokhin
Jul. 3, 2013 12:15 AM
Decision making is defined most simply as the act of choosing between two or more courses of action. It might seem straightforward, when only one person is involved and the choices are black and white. But add another person to the equation, and a few more choices, and the decision gets exponentially more complicated. I am reminded of this each time I venture out with my co-workers for lunch in midtown Manhattan. I have found it’s always best to claim, “I chose last time,” since that usually gets me out of the debate. But that won’t work when it comes to IT budgeting and planning.
So it is with some degree of wonder that IT decisions makers are still relying on vague data, questionable extrapolations, and subjective (and emotional) perspectives, making it impossible to gain consensus or get a comprehensive picture of the organization’s systems.
We hosted a webinar recently with Liz Herbert, a Principal Analyst with Forrester Research, on Innovation and ERP, and she knows a thing or two about the struggles IT decision makers face. During the webinar, we asked three poll questions to the attendees that shed light on some the challenges that organizations are facing with decision making, specifically as it relates to their ERP platform.
Poll Question 1: How much do you spend on operations vs development?
This one really galvanizes one of the key points made by Liz — how do organizations balance “run” with “innovate?” Organizations know they are spending more on operations (ie, maintenance) than on innovation (ie, building), with 60% of respondents spending over 50% and a full 1/3 of respondents spending over 75%. That certainly doesn’t leave much for innovation! Even more telling is the 20% that don’t know — a point backed up by Liz in her reference to a current Forrester customer who is struggling with understanding the allocation of their spend on a multi-million dollar program. (The fact that Forrester even has a viable consultation offering around this need further highlights the pervasiveness of the issue.)
Maybe it shouldn’t be such a surprise that organizations struggle to delineate their spend allocations, but if you don’t know what you are spending on maintenance, it sure makes it difficult to propose cost reductions or re-allocations.
Poll Question 2: How do you measure the extent of your customizations?
The responses here are not surprising, since knowing what to measure and how to measure are age old challenges — as old as development itself. It used to be simple, when systems were an aggregate of linearly-developed programs on the mainframe. But with the advent of different coding practice paradigms (thanks a lot OOAD and SOA), the task became more challenging and obtuse and much less accurate.
Those 25% of respondents that said they have no way to measure customization are doomed when it comes to innovation initiatives. And many of the methods that are being used do not take into account the complexity of what is being measured. Without insight into the quantity, complexity, and quality of customizations, any attempt at bringing in new functionality or features will be a lot more difficult than you anticipate, if they can get completed at all.
Poll Question 3: How many lines of ERP custom code do you have in your organization?
Even if you are among those who know where your spend is going, and you are making a serious attempt at measuring customization, is your measurement the right measurement?
The response to this question that stood out the most for me was that the highest percentage (31%) believes they have less than 100,000 LOCs. We have done analysis on 254 ERP applications in our Appmarq database, and that response on the surface is seemingly a valid proportion based on the median of our sample (putting them in the lower half), but it certainly puts them at odds with our analysis that shows an average of 806,000 LOC per ERP application. This is backed up by my experiences during 18 years in the industry and the hundreds of customers I have spent time with, much of that that doing ERP customizations, integrating with ERP, or testing ERP. Without strong measurement in place, virtually everyone underestimates the level of customization in their ERP platform and my gut tells me that a good portion of these 31% have a lot more than 100K.
So what to make of the 27% that believe they have over 750,000 LOC? Many make the opposite assumption of those who think they have less than 100,000. I can hear the internal conversation, “Lets see, we went live with SAP back in 2001. We then did that large program to integrate an acquisition — boy that was a doozy! And then we had that other project in 2003, 2004, and oh yeah, another in 2005 and 2006 … I guess we have a lot.” Odds are, those respondents aren’t any more sure than the rest of you — even if they may have made a more realistic estimation.
The fact remains that most don’t know how much ERP customization they have — and for too many the answer is “a lot” and “more than we would like.”
It is all about the data — it tells the real story
Now tie that back to the original intent of our webinar – ERP customization killing innovation. There are a set of decisions around ERP as it relates to innovation — modernize, consolidate multiple instances, upgrade to newer versions, continue to build out and customize further, etc. And each decision requires a strong command of the current state of ERP customizations. So, let’s reflect on the opening scenario around decision making — when there is one person making a decision between black and white, pretty simple process. When you have an organization making a future-altering decision, with representatives from the business, from IT and even sometimes from third party advisors and consultancies, the decision making becomes infinitely more challenging. There is too much at stake to use subjective information. You simply must get everyone working from the same set of objective, fact-based data points.
Tell us in the comments what you think of the poll results — are you surprised?
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week