Reversing the 80/20 rule for IT investments
posted by Marty Menard on December 27, 2006
Have you read the Information Week article about what HP is trying to accomplish in shifting their total IT spending from 20% on new capabilities to 80%? You can find the full article at this link. This is the second time this year that Randy Mott has made headlines about how he plans to shake up HP and the IT industry by significantly changing the rules of the great game of IT. The first time was the $1B in savings from Data Center consolidations.
Everyone in “the industry” will tell you that 80% of all IT spending goes to KTBR, or Keep the Business Running, and that only 20% of total spending is focused on new capabilities. Actually 80/20 is a best in class facility, many companies spend ~10% on new versus existing.
So what Randy and HP are suggeting is a complete reversal, spending 80% on new capabilities and only 20% on sustaining or KTBR. It seems impossible. How do you reverse the course? Intel isn’t significantly different in it’s spending patterns. Most of our total budget goes to sustaining our past investments and we spend lots of hours debating where we spend our precious dollars on key programs and projects. Do we invest in manufacturing, engineering, marketing, sales or services IT projects? We can look at Net Present Value (NPV) or an ROI calcaulation or which exec yells the loudest! But at some point a decision needs to be made.
One thing we’re committing to at Intel this year is a 3 year effort to replatform our entire ERP system. It’s a significant portion of our total investment and to make this happen we must actively stop other interesting but not compelling activities.
How does this relate to the reversal of the 80/20 rule? Most of our KTBR spending is due to poor implementations of ERP. By improving the environment, simplifying it, using SOA principles, we will reduce our KTBR costs. I’m doubtful this one activity can be a complete reversal, but moving to 50/50 seems feasible.
What IT can’t do is look to the operations and expect the whole reduction to come from them. Operations is a cost of past implementations. If the design and implemention is poor, cost after implemenation will be high.
So I applaud Randy. He’s a leader not only for HP but also for our industry. He’s challenging the norms that the IT industry has had for years and forcing many to stand up and take notice and also challenge what is possible.
Good computing
Comments
Dec 28 | Terry Davis said:
I’m an engineer. I thought the 80/20 rule was, “20% of the effort on new products yields 80% of the results”. Polishing a product takes all the effort.
Dec 28 | David Witte said:
Thanks for the post, very interesting article and perspective. The link to the Information Week story was broken in your post, here is the corrected one: http://www.informationweek.com/showArticle.jhtml;jsessionid=1KSNELAJBQYWWQSNDLPCKHSCJUNN2JVN?articleID=195300001&queryText=CIO+uncensored
Dec 28 | Marty said:
To David, thanks for the help with the URL. I have cut and pasted and tried 4 times to make this application take your changes, but I’ve been unsuccessful. I hope people will go read the article from your link!
To Terry: we’re both right. My comment was about the TOTAl budget, and yours is also true about project budgets. Randy Mott is challenging the IT industry to reverse the TOTAL spending some of that is due to project spending and how we should be more ‘roughly right than perfectly late.” Marty
Jan 02 | Micheil said:
I too am an engineer, of the hardware variety. To me, a resultant 80% of a budget being spent on KTBR issues is a manifestation of poor planning and/or poor execution. Like Intel’s recent efficiency analysis, I think HP is also looking at how business is conducted, though the HP 80/20 initiative speaks directly to IT. In the end, both HP and Intel are saying: Let’s stop and think, plan well, and execute better.
In hardware engineering we are going through the same thing. Maintenance is cumbersome when it should not be. We are stuck spending so much time in validation as opposed to development. Development is creation and progress. Validation is stagnation, though a necessary self check of the development cycle. I would argue that the 20/80 split exists in development/validation areas of hardware design, if not more toward the 10/90 side.
My point? I believe that the technology industry as a whole is realizing that we ought to be able to move forward better. I think it is invigorating to know that such introspection is taking place.