What's the (share)point?
February 2, 2012 7:35 PM   Subscribe

How do I evaluate the cost/benefits of migrating my org's front-end website from M$Sharepoint to an open-source platform?

I have been given one week starting today to evaluate the cost/benefit of migrating a ~3,000 page / ~2,000 visitor/day front-facing website from $harepoint to WordPress, Drupal, Plone or something else.

The idea is to save around $40K/year in fees that we pay a group to administer site hosting, domain hosting, and occasional header/footer/layout updates but nothing more than copy/pasting Google Analytics scripts in and out. I manage the content, design (pre-sharepoint), and functionality. (I actually iframe the functionality (forms) in from another site because the group cannot make time to provide things like email registration.)

I have never taken on a migration of this size before. I have also never contracted a job of this size out to a design firm. I have built maybe 20 websites in WordPress but nothing larger than 20 pages and a blog. After the migration, I would be solely responsible for all aspects of the website.

What are the major considerations I need to make? Are there guides for a project plan this specific?
posted by metajc to Computers & Internet (4 answers total) 3 users marked this as a favorite
 
First thing you want to do is document the current functionality, pain points and wanted enhancements (but only the MUST HAVES, ignore the nice to haves, and even the "this would be really useful"). The document doesn't need to have every tiny detail, but it should be pretty comprehensive.

Once you have this, find a good development firm. This part is tough. Get personal recommendations if you can. Obv. never pay for the whole thing upfront, make payment conditional on deliverables.

Show them the document, and have them describe how they would build this as well as how long it would take and the cost. Make sure you ask them what they think will be the more difficult portions to implement.

As with any subcontractor, watch them like a hawk. You'll have to strike that delicate balance between oversight and micro management. If you suspect incompetence let them know that that will not fly, and if they do not step it up, terminate the contract (make sure you write that into your agreement, you're better off losing 10% for deposit, then having to re-do the whole thing because it was written badly).

As far as which CMS? I'd imagine (as much as I like it) you'd need something more flexible then Wordpress for migrating that size system. I'm sure others can chime in with specific recommendations.

Good luck and Vaya Con Dios.
posted by pyro979 at 7:57 PM on February 2, 2012


Forgot to mention, try to make sure they optimize it for speed. Make sure that the developer knows that's a concern for you. This will include optimized images, using sprites or encoded images in CSS, as well some sort of caching solution like "W3 Total Cache" or something similar. You don't wanna overtax your servers and it'll be a much better user experience for your users.
posted by pyro979 at 8:01 PM on February 2, 2012


2,000 visits per day is essentially nothing traffic-wise, and is probably not worth paying anything for optimizations above and beyond reasonable programming practices.

That said, your first step needs to be figuring out the TCO of the new system. $40k sounds like an awful lot, but "open source" does not mean "free". You're still going to have server or hardware costs. If you get custom development, as pyro979 suggests, you'll have maintenance costs (every time wordpress upgrades, for example, I'd expect a bill from the contractor for the time they spend checking their custom work to ensure it works on the new version). That's to say nothing of every CMS's inevitable security holes that can be exploited before they're patched (doubly easy if the patching takes a while or the patches aren't applied quickly).

Also, 3,000 pages is a huge amount of content to transfer. The cost of doing that will be non-trivial.

It might end up being a significant time-sink with a relatively small cost savings behind it, and if MOSS meets your current needs (and people know how to use it), it might be a net loss, especially in the short term, to switch to a new system.

This also jumped out at me:

After the migration, I would be solely responsible for all aspects of the website.

This is a good time to be exceedingly honest with yourself. Managing a CMS is often more difficult than picking one . . . if you can't do the latter without help from AskMefi, you should probably spend a little time thinking very seriously about whether you can actually do the former by yourself, too. You are probably underestimating the work that goes in to administering your MOSS install.
posted by toomuchpete at 8:43 PM on February 2, 2012 [1 favorite]


If you have the technical know-how to do this yourself, you should do it yourself: either an end-to-end sloppy trial run to inform your interaction with bidders and contractors who take on the whole project, or better yet you should actually take central control and implementation of the project yourself and outsource very specific deliverables to others. (Probably things like sets of narrowly-specified, thoroughly-tested scripts to scrape sections of the existing site and create content objects / database entries in the new CMS.)

If you do not have the basic know-how to at least do all of the steps yourself manually, as toomuchpete is indicating there is no particular reason you should be looking at open source software. I would expect the licensing costs for a commercial product similar in capabilities to the ones you mention to be a small fraction of the overall bill for the migration and the ongoing cost of maintenance. (An open source solution could be the way to go but absent any specific functionality or infrastructure requirements on your part, in 2012 things like good project planning from you and the skill and experience of the team carrying out the migration are much more important than what flavor of CMS software they use.)

The flip side of that is, if you're talking with a commercial CMS software vendor directly and the sales guy is saying something like, "Why I can give you a license for $$K and we can hook you up with our professional services department / one of our partners who can handle that migration for probably the same amount or less!" then something fishy is going on, i.e. unless you have very particular needs you haven't mentioned, needs that some specialized commercial CMS fulfills exactly, you should not be seeing bids / quotes where most of the costs are going towards software licensing fees.
posted by XMLicious at 2:19 AM on February 3, 2012


« Older math quest   |   1 minute desserts for lazy cooks Newer »
This thread is closed to new comments.