If It Looks Like a Cow, Swims Like a Dolphin and Quacks Like a Duck, It Must Be Enterprise Software | Hacker News Reader

If It Looks Like a Cow, Swims Like a Dolphin and Quacks Like a Duck, It Must Be Enterprise Software

subtraction.com

15 pointscharzom6 comments

Open articleView on HN

That's consistent with my experiences. As several commentators mentioned, I think it's such a problem because decisions about enterprise software in big organizations are typically made by people who don't understand software, design, or information architecture. Smaller companies are much more responsive. It is a mistake (which I've made more than once) to assume that a large organization will make reasonable decisions.

It used to be that a successful business started with expertise in how to make widgets or rent cars or something and then tried to figure out information management. The new trend is for people who understand information management to start from there and get into some other business. Old-school businesses, even capital-intensive ones, which remain clueless are gradually becoming more irrelevant as newer information-savvy startups intrude on their turf.

And that is rather interesting.

I've always wondered why enterprise software is so shitty. The HMO Kaiser uses a DOS-based keyboard-controlled interface to schedule appointments. The interface consists only of white or green dots on a black background. But the computers are all fairly powerful machines running Windows XP.

Making an appointment requires the user (secretary, usually) to hold the arrow key down for several seconds as the screens scroll by. It is really the most time-wasting piece of software I have ever seen.

The interesting thing is that Kaiser probably bought the software at the lowest possible price, choosing the crappiest company to develop it. But in the end, it wastes so much time, confuses their employees so much, and pisses of their customers so much that it has probably cost them far more than it would have to just buy well-designed custom software.

In the future, I honestly think the best enterprise software will be in the form of web-based systems that run in browsers. It's easily updatable, uses standards-based technology, is fairly cheap to develop, and-- perhaps most importantly-- the users of the software have experience using interfaces on the web. Everyone uses the internet, even Kaiser's secretaries, so that will completely eliminate the learning curve of these retarded custom interfaces.

> The interesting thing is that Kaiser probably bought the software at the lowest possible price, choosing the crappiest company to develop it.

More likely, the software dates back to 1984, and they've been trying to upgrade it for 10 years, but since it now has 25 years worth of cruft in it, it will still take another 5-7 years to completely migrate all their internal and external partners to the new system.

See the comment from Jason Schulz on the original article. He's nailed it.

HMO? Kaiser? That explains it. It doesn't matter. They are in the health care, er a, money printing business. Nothing matters. Every cost gets passed on. Nothing makes sense in that industry. It is an oligopoly protected by a monopoly (the AMA).

Enter that industry at your own peril. Anything logical you ever learned in business no longer applies.

Why do you say that? Sounds like you have some battle scars from first hand experience in health care. Care to share them?

I wouldn't know where to start. But here goes anyway:

1. I wrote a billing system for an M.D.'s personal practice especially geared to collect the $1.6 million in outstanding receivables. She killed the project because "those office people are too stupid to use it anyway." It was more important to her to maintain her "superiority" than collect her bills.

2. An HMO whose legacy system I maintained decided to convert to something "more modern than green screens with color and a mouse". (The legacy system worked perfectly.) The new system never worked. Two years, 3 consulting firms and $20 million later, I was brought back in to "convert it back". By then, they had lost 60% of their revenue and were out of business a year later.

3. I doctor's health organization was acquiring about one clinic per month. Their Big 5 Firm brought me in to develop a plan to convert each clinic's software and data into the main system. It was taking 18 months to do each conversion. They were 29 conversions behind. I made recommendations that required drastic changes in their IT department. They were so upset, they refused the engagement. (Out of business 18 months later.)

4. Another HMO spent 18 months and $5 million implement HIPPA to maintain privacy. The running joke was that the files for the worst cases were kept in the break room for all to see for a laugh on their break.

5. The claims processing software in another HMO I worked at was configured to NEVER APPROVE A CLAIM ON ITS FIRST PASS. I don't know what happened to them, but prison would be too good.

6. At a PPO I worked at, I had to cover production errors between Christmas and New Years because the 2 primary programmers were on vacation. (I only had access to the test system.) I prodded my supervisor to get me a production login and password for 4 weeks. The first error (7:30 on Monday morning) and I couldn't get in. We argued in front of the call center (400 people) about my lack of access. "You DO have a password and here it is," he cried. "Yea, but I don't have a login." "What's a login?" That pretty much sums up the talent level in healthcare IMO.

Just part of the reason healthcare is so expensive, but the providers earn less.

I'll chew razor blades before I ever consider "healthcare" work again.

Reply on news.ycombinator.com