$228M Blown on Customizing Salesforce

Thanks to my colleague David for bringing to my attention the Federal Government's $228M blowout in trying to customise Salesforce for the aged care industry.

Here’s a quick summary:

• 8x blowout by Accenture ($18M -> $157M)

• 3x blowout by Salesforce ($11M -> $34M)

• 2x blowout by Capgemini ($37M -> $74M)

And it’s far from over:

"But what we also know is that the system for registering these contracts is incomplete, and there is in all likelihood a lot more money flowing into people's pockets that we just cannot see."

As a taxpayer, it’s frustrating that no public servant ever loses their job over these things, yet they happen repeatedly, every year. And these are just the projects that get completed—there are plenty more that are halted after massive overruns.

I worked on projects like this back in the '90s with Ernst & Young, implementing Oracle and Siebel (the predecessor to Salesforce in terms of CRM market dominance).

It’s the reason I left instead of taking up a partnership role. The waste was staggering, and I didn’t enjoy working on a project for two years, only to feel like I’d achieved nothing of value for the taxpayer. It was incredibly disheartening.

This is a prime example of the "get your foot in the door" approach I've warned against, which happens just as much in the charity sector as it does in government and corporate..

I’m confident (enough to lay a wager) that the 8x blowouts from Accenture and those from Capgemini are for "configurations" (aka customisations). I suspect the 3x blowout on Salesforce licensing/support is a direct result of extending Salesforce functionality—which often involves rolling out the extended functions to users who don’t need the base CRM functionality but still incur an additional licence.

What’s going to be really interesting is seeing what happens in a few years when these customisations start getting in the way of the government’s ability to upgrade to a major new version of Salesforce, held back by those same customisations. This is what often happens. That’ll be another $100M to remediate, straight into the pockets of Accenture or Capgemini.

This reminds me of the IBM fiasco with the 2016 census website. It failed on census night because IBM managed their own infrastructure instead of using Cloud v2.0 technologies, which were solidly proven by that time. It also cost $9.6M, which is ridiculous for an online form. I would have happily given a fixed-price quote for $1.0M and bought a Tesla Roadster with the profit. I’m not being flippant here—I’ve built high-peak systems like this a few times.

The problem is, it doesn’t matter what government I vote for; the people responsible for these decisions remain the same. We need a democracy v2.0 ("direct democracy").

Question: Have you ever worked on a project where you felt the costs were unjustifiably inflated? If so, how did you handle it?

Andrew Walker
Technology consulting for charities
https://www.linkedin.com/in/andrew-walker-the-impatient-futurist/

Did someone forward this email to you? Want your own subscription? Head over here and sign yourself right up!

Back issues available here.