Should my stuff be on the Web, or on my PC ?

Two short stories (there is a point to this):

I read yesterday that an online RSS reader, SearchFox, is suddenly closing down. On their website/blog they say "Please export all of your links and an OPML file with your RSS sources" (whatever that means). As it turns out Yahoo! may be buying the smoking wreckage.

Also yesterday, the accounting software I use on my PC blew up, and apparently I shouldn't really be relying on a version of the software that is five years old and no longer supported. (As it turns out Sage probably will sell me an upgrade and I may be able to recover my data.)

So which is better/worse – to be screwed when your Web 2.0, constantly updated, don't be evil, really cool Web app goes offline, goes broke or gets bought, or to be screwed when your desktop computing, old-hat, locally-running Windows app gets tired, freezes up or enjoys a head crash ?

I used to want everything on my laptop, because connectivity was uncertain and slow, (and because it was 'mine, all mine') but now you can be online 'anywhere', 'anytime' (yeah, right), I was just beginning to come round to the idea that it might be nice to have everything 'out there somewhere'.

But actually, comms is still very flaky, in my experience, and maybe some of these web outfits are flakier still. Is your whole photo collection only on Flikr, or do you also have it on a disk at home ? Why ?

Ideally, I suppose, all the data and all the code should be everywhere all the time. Seamlessly. So if I log on to the library, I could be searching the local catalog and/or the Amazon bibliography and/or OCLC, whichever is up at the time. And I should be able to look at my loan history online or offline, because the data is automatically kept in synch between the library server and my PC (oh, and my PDA, of course).
Web 3.0 anyone ?

(I have to go now; I need to back up my blog and buy some shares in Intellisync.)

Why Campus Management Systems Cost Ten Times More Than Library Management Systems

Rob Abel made some interesting comments in his Higher Education Insight blog a while ago (October 2005), on the question "Why Won't We Pay for Learning Technology ?".

He asked "Why are institutions willing to spend $300K to $500K or more on student information systems when it is like pulling teeth to invest $25K-$50K in a course management system? " He had some interesting thoughts on whether this is due to resistance to change by tenured faculty, and also on why institutions feel it is acceptable to charge students for textbooks, but not for learning management systems.

My experience is similar in the library world: Bob Walton, former VP at Innovative Interfaces, Board-member of Dynix, and current VP for Finance at Wooster College, recently said that a college like his typically ends up paying ten times more for a campus management system than for a library automation system (RMG panel at Midwinter ALA in 2005).

I think much of the problem is with the 'benefit case' – both with the figures themselves, and with whoever is preparing them. For administration systems, it is relatively easy to compile financial benefits from the proposed systems – tangible or not. Management will come up with a report that says how the new system will save 5% of this and 10% of that, and will show that only a 1% increase in enrolment, due to better systems, means $xm more income.

But if you sit an educator or a librarian down and say, "So exactly how much cash would we save/gain by replacing/upgrading our learning/library system ?" you might get blank faces all round. Partly, it is just hard to do. Partly, these people do not have the skills to do it. Partly it is sort-of not 'right' to put a price on better learning or research.

Yet it needs to be done. Sure, if higher prices are paid the money will have to come from somewhere. And, sure, some of it will end up in the pockets of the owners of the vendor companies. But think how much better these systems would be if they cost only twice as much as they do now.

The Cost of Selling Software - US$250,000 per deal

Jim Farmer has prepared a very interesting analysis of what it costs to sell an Enterprise Learning System, concluding that the sales and marketing costs of each Blackboard signing are around US$250,000.

Now open source advocates can sometimes over-emphasise the costs and failings of commercial software, but Jim's comments in the paper are very fair, reminding readers that "The high cost ... is influenced by the ... demands of potential clients." and that "... software suppliers have been asked to ... subsidize user groups, and support higher education initiatives." His paper is well worth reading in full.

(Jim told me three years ago that open-source search engines and campus portals would soon make library automation software - and libraries - obsolete; and it still scares me.)

I know something about the cost of sales issue, having sold library automation software for most of the past ten years (at Innovative and Dynix). I have long said that it costs the major ILS/LMS vendors an average of about US$50,000 per bid, which means that for a typical mid-sized university deal, five competing bidders will spend US$250,000 between them: and only one will win, so again that is US$250,000 per deal.

Of course an individual vendor's costs per deal are very dependent on the win-rate, something Jim has not really looked into.

As a deal like this might only have revenues of US$250,000 anyway, it means the library community has paid US$250,000 to the library software vendors and none of this has gone into development – it was all spent on sales and marketing.

I think all this raises three main issues:
- What are these costs and why are they being incurred ?
- What could or should be done differently ?
- Where does open source fit into all this ?
I propose to deal with each of these subjects in turn in future blogs.

Let me leave you with a intriguing thought:

Assuming that WebCT and Blackboard used to compete (with sales & marketing dollars) in every deal - once they are merged (if they still collectively win the same number of deals), they will collectively save US$250,000 per deal, and this saving could be passed right on to the customer. So reduced competition is a wonderful thing for customers, as it cuts sales costs and so reduces prices. Right ?

But hang on, when I was at Dynix competing with Sirsi, the libraries would 'shortlist' five suppliers and I had, crudely, a 20% chance of winning, and Sirsi had a 20% chance - so 40% between us. But now, SirsiDynix combined only has a 20% chance on that 5-vendor shortlist, so the costs per bid are the same and the costs per win are unchanged too - so no savings there.

Hummm.....more analysis needed

(Thanks to Michael Feldstein for drawing Jim's report to my attention in Blackboard by the Numbers in his blog, e-Literate.)