In other news, water is wet. This has been a time-honored practice (along with cost-plus) for decades.
I used to work for a defense contractor. My favorite time of the month, was when some manager from a completely different division, would come by, and give me a time sheet to fill out.
Did you at least get paid for the extra “time” or was it purely going to the company?
Makes me wonder how the manager (or anyone else doing a similar thing) justified to themselves that doing this sort of thing was ok, since most people don’t like the idea that they’re just straight up committing fraud.
It was straight-up fraud. I passed it by a friend, who told me the deal. My time was billed on a cost-plus contract. I could have gotten in trouble for filling those out. By the time I figured it out, they stopped it.
No, I didn't get paid for it. Back then, I was a junior test engineer, and was getting paid peanuts (which was probably why they had me do it. I'm sure that the hourly rate for me was upped a wee bit). I wasn't working on any particular contract, myself, as we were sort of "floaters," at the time.
The US Department of Veterans Affairs is spending a lot more than that to switch from their in-house healthcare system to Cerner. It is not going well.
Organizations might not have the ability or money to hire an entire organization to build a product from scratch.
An upfront $200,000,000 bill is much harder to budget for than a 7-8 year $25,000,000 bill at an organization generating around $1 billion in revenue and with extremely low margins.
For most organizations, software is a cost center the same way HR, Payroll, and Accounting is within tech companies, so whatever limited money exists is better spent on their primary mission
Contracting out functions can work quite well, but the issue is organizations might not hire the right kind of management staff for techncial projects. Also, in most organizations IT/Infra falls under the CFO organization, and as such budgets are much smaller
Also, procurement rules are often written in such a way that it's often unfeasible for newer alternatives to compete simply because there is a massive upfront cost to meet RFP requirements.
Contrary to popular belief, enterprise sales doesn't want to target high revenue alone, you want to target high revenue PER dollar spent. If I made an efficient competitor to Cerner or Epic, why should I spend $50-60mil over 3 years to get FedRAMP qualified if the TAM federal agencies can provide me is maybe $100mil? For Cerner/Oracle and Epic, that $50-60mil isn't that expensive to provision because it can add additional lock-in or deal sweetners like other adjacent softwares or services getting discounted, thus making that $100mil into $130-150mil, or leveraging economies of scale to make that $50-60mil infra cost down to $20-30mil.
Money doesn't solve all problems, and unfortunately these health systems do not have the IT staff required to even hire competent engineering leadership and then build out dozens of teams required to in-house their EMR.
EMRs are incredibly complex, mainly due to all their integration points, billing systems, staffing etc.
You could ask the same of enterprises using Salesforce, AWS, or SAP. At a certain scale, it’s just not feasible to have the org consumed by technology. It’s much easier to solve problems by throwing money at vendors. Arguing otherwise is a nauseating consulting exercise that doesn’t guarantee cost, efficiency, or innovation improvements at the end of the day.
I worked for a company that did state gov contracts. We probably built legislative publishing stuff for more US states than any other company in recent years.
We never ever locked anyone in, not even little bit. We bult everything on open standards (odt and later docx), django etc.
A typical cycle was they would hire us to do joint development for 1-2 years, then a juicy support contract for another 1-2 years, then they gave us the boot.
My feeling after doing that many times is there's no money in it really. Government clients are impossible to work with. They have utterly perverse incentives that defy any rational mind and they don't pay if they can get away with it.
> One example of vendor-lock the study cites is a procurement process in which the Department of Agriculture in 2021 spent $112 million more to buy Microsoft Office than Google Workspace to avoid switching costs that it perceived to be even higher.
This just shows how incredible ingrained the preference for MS is and how ridiculous expensive it is. Technically a Ubuntu with a browser and Libreoffice should suffice for 99% of a their work and even if one department or office wanted to try this, they likely wouldn't be allowed due to some regulation pushed from the top.
MS had first mover advantage in the Fed and hired much better Sales and PS Engineers than Google or AWS, and also believes that the customer is always right.
For example, when Agencies were deciding between AWS and Azure almost a decade ago, Azure offered to make their entire infra FedRAMP compliant by default, but AWS felt the cost would be way too high upfront so they built GovCloud from scratch.
Also, organizations want someone to yell at when their entire infrastructure went down at 2am on Saturday. Using free/open source sofrware doesn't come with support or services, unless you hire a company, hence why RedHat/IBM Linux, SUSE Linux (most conglomerates in DACH uses this because of their partnership with SAP), and Oracle Linux exist.
I used to work for a defense contractor. My favorite time of the month, was when some manager from a completely different division, would come by, and give me a time sheet to fill out.