CMS/Content entry resource recommendations 20K pages multiple languages

December 29, 2014 9:40 AM

I am looking for a recommendation for outsourcing a grunt task. A prospective client is asking for recommendations for porting over 20K pages. It will require someone doing it manually since the site is a bit of a mess. I realize that there are freelance sites where users bid on projects like this. I am looking for a recommendation (either for a site or a firm to do the work) if you have one, let me know. Just to be clear I am asking about content entry and not the Drupal development portion.

The biggest issue is the pages are built with tables. The tables aren't just at a template level. They are embedded within the content of the page. Because of this my recommendation is we setup the new site in Drupal 8 and hire a low cost resource to port over the content manually.

The service or firm we use wont be able to be automated because the existing pages aren't consistent. Because of this the content porters/entry people will have to be smart enough to pick an appropriate template. Content types are TBD but there will be options to cover various needs. All will be responsive and setup to hold content and imagery in specific spots. We aren't giving users control to add in any layout they want.

The other complication is these pages all cross link.

So here are the steps I see happening:

Step 1 - Setup site in Drupal 8 in a totally separate environment
Step 2 - port over all old content into the new multi-site install programmatically. Each page would get a flag noting that it hasn't been updated.

Then we would hire a firm at a cheap/reasonable hourly rate to:
Step 3 - select a new template for each page and cut and paste just the content and links into a new structured template. The URL and page type would need to not change so all links continue working. After the content is transfered over the flag would be updated by the partner.

Step 4 - The client would then work on QA and testing to make sure the content came through as expected. After the client approves the page they would update the flag as final.

posted by rdurbin to Computers & Internet (2 answers total)

Have you checked with a service like Kapow? This sounds like a mess, but I've done this before. Do they really have 20k content pages? I doubt it, your best bet is to aggressively cull content. Use analytics data to back you up. Are the other languages straight translation or is it different content?

Also it sounds like they are not getting new creative and this is a straight port? That makes it worse. You really need a content audit to identify templates that are implied if not necessarily consistent from a DOM structure.

Otherwise, no I would not recommend using a service like Mechanical Turk. You need an on or offshore team of front end web developers and QA to go through each page by hand more or less. This is not an easy task.

This assuming these are actual static content pages. In my experience, in CMS systems that were legacy or didn't tightly control how content is entered you also get hacky JS and flash that ports poorly. I'm ignoring that one of these pages ends up being a complex single page app that takes a couple days to replicate.
posted by geoff. at 10:06 AM on December 29, 2014

Best answer: Step 4 - The client would then work on QA and testing to make sure the content came through as expected. After the client approves the page they would update the flag as final.

My experience is that clients do not want to do this and it is fuzzy whee responsibilities lie. Assume you'll need the pages near perfect before showing them, despite what they say. Otherwise they see a few pages with content and no styling, do the math on 20k * 8 languages and get angry.
posted by geoff. at 10:08 AM on December 29, 2014 [3 favorites]