Why IT projects fail: underspend

Here's my theory on why so many IT projects fail to deliver the expected result: we underspend on what matters.

In a comment here on this blog, Cary King said

Gartner says that IT Service and Asset Management is 80% people and process, and 20% technology. Forrester says it is 70% / 30%. Let’s split the difference: IT Service and Asset Management is:
75% people and process;
25% technology.

I agree with Gartner and Forrester (which is a worry) and Cary (which makes me feel more comfortable). And I'm not just talking about asset management or even ITSM - the proportions hold across IT. At the recent itSMF Australia conference, my keynote talked about why IT projects fail: because the spending is disproportionate.

We spend like this:
Image

When we should spend AT LEAST like this
Image

and I think optimal is like this
Image

so I guess I went for
50% people
30% process
20% tech

To clarify, I'm not saying we overspend on technology. I'm saying we underspend on process and especially we WAY underspend on people; consultation, testing, training, coaching, feedback... all the aspects of cultural change.

So if the technology spend remains the same amount but is 20% of the budget rather than 75% , what is the change in the overall budget? An increase of 275% If it was a half-million-dollar technology project it is now a 1.9-million-dollar transformation project.

The reason IT projects fail is because they were underestimated by about 70% by neglecting people and process aspects.

If the project's business case doesn't stack up at this higher spend, then it was a bad idea. Because it will end up costing that much anyway to change the behaviours of the people. Or if we stick to the lower spend it will fail. Fail slowly over years or fail spectacularly, but either way it will fail because the behaviours weren't really permanently changed.

For example, as often happens your technology vendor writes the business case for you:

Software license         $350,000
Vendor services         $120,000
Staff training            $25,000
Vendor support        $63,000p.a.

Half a million up front and 63 grand every year to the vendor. The vendor will then find between $500k and $1M in benefits depending on how hard they are prepared to push the envelope of believability.

Here's the reality of what is needed to make it work (whether you budget it up front or end up having to find it later):

Software license         $350,000
Vendor services         $120,000
Staff training            $25,000
Requirements            $60,000
Software changes      $100,000
Process change          $300,000
Test and accept          $300,000
User training            $120,000
Rollout                  $60,000
Early life support    $100,000
Vendor support        $63,000p.a.
Inhouse support        $30,000p.a.
User education        $60,000p.a.
User feedback        $20,000p.a
Process improvement  $20,000p.a.

1.5 million up front and a couple of hundred thousand a year.

Comments

The article I wish I had written

Well done, Skep. The third graph is accurate, of course, whether it is recognized and budgeted up front, or it is revealed half way into the project. In the latter case, the investments are slipped surreptitiously into the budgets of everyone else who has to deal with the mess. Hence the project may appear to be in budget even when it isn't.

There are probably many reasons it is done this way. The various costs are contingent and hard to calculate. The people are a fixed expense and considered "free" for the purpose of absorbing incremental workloads. The project would never have been undertaken had the stakeholders understood the true costs of the effort. Risk management (identification, controls) are weak or non-existent. Hubris blinds the various stakeholders to the costs of cultural change.

I am frustrated by the consistent "blame the tool" message in the blogosphere of late, but that is another discussion.

Absolutely, yes!

"If the project's business case doesn't stack up at this higher spend, then it was a bad idea. Because it will end up costing that much anyway to change the behaviours of the people. Or if we stick to the lower spend it will fail. Fail slowly over years or fail spectacularly, but either way it will fail because the behaviours weren't really permanently changed."

Absolutely.

Some of those costs, however, may be absorbed by the day-to-day work of the people in the functions. So, your ratios may be a little aggressive. The overall concept is, however, absolutely correct.

IMO Admiral Hyman Rickover said it best, "Good ideas are not adopted automatically. They must be driven into practice with courageous impatience. Once implemented they can be easily overturned or subverted through apathy or lack of follow-up, so a continuous effort is required."

Recommend Marshall Goldsmith's book, "What Got You Here Won't Get You There..." and David Maister's "Strategy and the Fat Smoker." Both disuss the reasons people fail to succeed. It is people that make decisions about projects. Therefor all the reasons people fail to change are able to be seen in projects. Starting with optimism bias....

Cary King
Minerva Enterprises
Managing Partner
www.MinervaE.com

Syndicate content