by jer0me on 9/18/24, 4:08 PM with 205 comments
by tqi on 9/18/24, 6:30 PM
This is an interesting perspective. From what I've seen / heard from others, it's generally better to adapt processes to the off-the-shelf tool than to try and customize or build from scratch to accommodate your bespoke processes (especially in the business operations realms). For one, the organization is likely less unique than you think, and those bespoke processes are as often a function of some early employee's preference as it is a genuinely good reason. For another, customizing software is not just a one time cost, since every subsequent update / upgrade is likely to require additional work (or at least testing). And finally, in most cases the closer you are to a vanilla, standard process, the more likely you are to stay in compliance with local laws and regulations.
Though I suppose it's possible that the imbalance is just due to the fact that it is much harder to quantify the costs of using a suboptimal (for you) process than it is to look at the procurement contract for a custom solution.
by alberth on 9/18/24, 5:31 PM
a. I use to work in this space, even a $6M deal would be massive (let alone something 100x bigger).
b. The ENTIRE cuny budget in 2013 was only $2B [0]. This isn't their IT budget, this is literally the entire budget to run the entire university system across multiple campus, faculty, buildings, etc.
c. Because Higher Ed is known to be so cheap, especially in the late 00s - big tech companies charged accordingly (which was at a massive discount relative to most other accounts).
d. even if this $600M figure was an aggregated figure over multiple years, staffing and auxiliary costs - it still wouldn't come close to this figure.
e. an expenditure this large would definitely be called out in CUNY annual financial reports, and I can't seem to find any reference to it.
[0] https://www.cuny.edu/wp-content/uploads/sites/4/media-assets...
---
EDIT: I did find reference last year to a $175M funding request to migrate from on-prem PeopleSoft (Oracle) to cloud.
Though, what I've historically seen is that only 10-20% of the funding request actually go to the software vendor. Organizations typically add 3-5x additional to either "pad" their request (in the event it doesn't get 'fully funded') and/or this is an opportunity for the university to higher for a bunch of roles they wouldn't have been able to get funded in the first place - so lots of things get buried in these numbers.
Lastly, the figures are also typically multi-year. Like 5-years being asked up front for approval.
Said differently, it wouln't surprise me if the true annual migration cost from onprem to cloud PeopleSoft might only be $10-20M
https://www.cuny.edu/wp-content/uploads/sites/4/page-assets/...
by JCM9 on 9/18/24, 5:00 PM
The whole thing is kept alive by the student loan program. Modify that or take it away and academia in the US would implode.
by gkoberger on 9/18/24, 5:29 PM
They obviously went with Oracle. I'm sure they spent a ton of money, and I'm sure it was the right choice. I would have gotten bored of it pretty quick. You don't pay Oracle because it's a good deal or a good product... you pay Oracle so you never have to think about it ever again.
I don't really have a point here. I wish there were better options in the market. But I certainly don't want to build them, since it's a boring problem with bad customers (edtech is a horrible sector to sell to). Oracle has a price point that makes it worth it to them, and they have customers willing to pay it.
Hopefully someone sees this blog, and rather than be annoyed at academic/government waste, sees a big market they can dominate with a better product. But given how it was written in 2013, I'm not so sure.
by jer0me on 9/18/24, 4:12 PM
---
This is a post by Prof. David Arnow on the blog of the Brooklyn College’s professors’ union about CUNYfirst, a PeopleSoft-based course registration and HR system sold by Oracle. Posting because the system recently got some attention on Twitter: https://x.com/ChocolateyCrepe/status/1836171439965446441
by bluedino on 9/18/24, 5:36 PM
You want 1,000 licenses for it? That will be $5,000 a year, for a total of $5 million
It's going to us a year to implement it, we're going to send out 25 people to get it up and running, train your users, etc. That's another $25 million.
We'll spend the next year building integrations to all of your other software and systems, that's going to be another $25 million.
These quotes will all vary +/- 25% depending on the vendor. Schedule a 200 hours of meetings, trainings, etc for 500 people involved with the new software. That's another $5 million.
Where's the other $540 million coming from?
by kcb on 9/18/24, 5:19 PM
by coliveira on 9/18/24, 5:35 PM
by Salgat on 9/18/24, 5:06 PM
by neilv on 9/18/24, 5:50 PM
Some bad ways that big-ticket purchasing decisions are made:
* Committee of people who don't know what they're doing, and/or who can't coordinate to arrive at a good holistic decision/solution.
* Person who wants this done for good reasons, but doesn't know what they're doing.
* Person who wants this as an accomplishment credited to them, but doesn't know what they're doing.
* Person who is mainly thinking "nobody ever got fired for buying [old big-name vendor]", and everything else is secondary.
* Person who is bribed by vendor (e.g., immediate cash, quasi dates with attractive salesperson, career rotating door with vendor).
Other ways?
(I haven't directly seen the bribery way, though heard of it in news stories. I've definitely seen all the other bad ways happen.)
by textlapse on 9/18/24, 5:11 PM
by pje on 9/18/24, 5:15 PM
https://www.youtube.com/watch?v=-zRN7XLCRhc&t=2047s
> There has been no entity in human history with less complexity or nuance to it than Oracle. [...] This company is about one man, his ego, and what he wants to inflict on humanity. That's it.
by whimsicalism on 9/18/24, 6:50 PM
by vondur on 9/18/24, 6:37 PM
Full report here: https://information.auditor.ca.gov/pdfs/sr2004/2002-110.pdf
by CSMastermind on 9/18/24, 5:11 PM
With that said this example strains even my ability to justify the incompetence angle.
by game_the0ry on 9/18/24, 5:53 PM
Why not let the computer science students take a crack at building some of those systems? Maybe not the an HR system, but why not a course registration system? Maybe not for the whole university, but maybe for just the computer science department?
The university would get work for free and the students would get real-world practice with building production code.
I get that there would be risk, but if it was under the supervision of professors (who hopefully are good at building, not just lecturing theory), I think there is an opportunity there.
by MangoCoffee on 9/18/24, 6:12 PM
In this case, I don't think Oracle is to blame.
They explicitly warned about the 'limitation,' yet the project proceeded nonetheless.
In my opinion, for this project to succeed, it should have been built from scratch. With the $800 million invested, they could have assembled a team of seasoned and junior developers to get it done. Just my two cents.
by insane_dreamer on 9/18/24, 5:04 PM
But the above numbers are hugely generous. This is not building an ERP from scratch. Do you really think it would take 500 people (say 400 engineers and 100 non-engineers), to build and deploy such a system? I would imagine you could get it done (and done right) with half that many, or less.
Anyway, just mind-blowing.
by lo_fye on 9/18/24, 5:25 PM
by db48x on 9/18/24, 5:27 PM
by ok123456 on 9/18/24, 7:01 PM
by lvl155 on 9/18/24, 5:15 PM
by zitrussaft on 9/18/24, 5:12 PM
by jonathaneunice on 9/18/24, 6:22 PM
by santoshalper on 9/18/24, 4:54 PM
"Why Enterprise Software Sucks" https://x.com/random_walker/status/1182635589604171776?lang=...
by llm_nerd on 9/18/24, 6:44 PM
The system was a disaster. It never worked properly. So the government is throwing hundreds of millions towards its replacement.
by motohagiography on 9/18/24, 5:38 PM
I'd bet huge that there is a layer of managers who don't see themselves as being accountable for domain competence in anything they manage as that's what the consultants are for. Consultants mean headcount and budget to manage- which is the definition of success in an institution. they run an organiation, the mission is little people IC problems.
It's not broken, it just works for people you can't see.
by narrator on 9/18/24, 6:04 PM
by gcanyon on 9/18/24, 5:47 PM
1. Had multiple installer applications on it, with no indication which was “the” installer application. 2. On opening the installer, asked me to select the install file to act on, again with no clear direction. 3. Had “help” files on the disc, in HTML format, which contained broken links to other files on the disc.
At the same time my coworkers, experienced Oracle DBAs and developers, with full paid-for support from Oracle, spent an entire summer trying to install an Oracle development environment, and failed.
All of which to say, yeah — $600 million sounds about right, as long as it turns out the software was never successfully implemented.
by dancemethis on 9/18/24, 6:21 PM
Certainly one of the names of all times.
by zemariagp on 9/18/24, 6:31 PM