Web design and development professionals will usually be asked to host a client's website or database at some point in their careers. It's not surprising; it's hard to find a reliable hosting company, and a top-notch hosting company with 24/7/365 professional support suitable for a non-techie nonprofit is usually out of that nonprofit's price range. If a client trusts you to build their site, they'll eventually wish you would host it for them, too.
When I posted a Facebook status on partnering with Rackspace to offer website hosting for nonprofits, Dan MacNeil of the Community Software Lab replied "In my small sample, (1? 2?), contractors reselling hosting regret it." And for good reason. Although most website development companies (including ours) have some level of systems administration talent, hosting websites is a high volume, small margin business model. Since we're a small consulting organization, we're not likely to have (or want) thousands of nonprofits on our hosting service. If we're trying to charge $10 a month for a hosting "product" for the nonprofit sector, how will we staff a round the clock call center for support, or cover one or more salaries for systems administrators? Dan also asked: "What value do you add between your customers and Rackspace? When things go wrong, how much power do you have to fix it?"
I think the answer is that we shouldn't sell or consider hosting the way we would a "product". Nonprofit website hosting isn't really a profitable or sustainable business vertical by itself, given the levels of support and guidance we'd like to provide. Instead, what are the ways offering hosting could improve our clients' consulting experience with us?
Economies of scale. Ever priced a Rackspace server? Their services are very high-end, high-priced, and much more high-powered than most of our clients need. If we buy hosting from Rackspace and divide it between our clients, we can offer clients the right amount of hosting, on world-class infrastructure, starting at $10/month.
Incredible support. Rackspace has built a reputation of having the best support in the industry, and my experiences with them suggest this is well-earned. Working with them allows our clients access to this level of support via phone, email, or chat...24 hours a day, 7 days a week, and 365 days a year. Support at that level is simply unavailable on similarly priced hosting packages.
Lower priced websites. This is the answer to Dan's question above, and it's critical. Every hosting environment is different, and every web developer or designer knows the pain of figuring out how that particular hosting company handles caching, or where PHP settings need to be tweaked to get Drupal running happily. Standardizing a hosting environment means we don't need to learn about a new one each time we start a project, saving our developers' time and your budget. And of course, many clients will save precious budget dollars by not having to migrate away from a sub-par hosting company midway through a project; sadly, this happens more often than you'd think. A few hundred (or thousand) dollars saved from your project's budget in not having to deal with a crappy hosting company can add up a new feature or three, some extra training and strategy sessions for your staff, or some design and usability enhancements.
Closer client contact. I am anticipating that clients who host with us will stay in closer contact with us after a project launch. This allows us to get a little closer to them as their website needs evolve over time, and lets us meet those needs more effectively. This is hard to quantify, but listening to our clients will make or break us over the coming months and years. Anything that lets us hear their voices a little more often is probably a good thing.
One support channel, that works. This one is more minor, but it's still valid. We're always telling clients "contact your hosting company to get that fixed"; sometimes, we even end up fixing it for free (when we can) when their host is slacking. Offering hosting services lets us create a single point of contact for hosting / email support, where issues regarding our work can be escalated to one of our project managers. This makes our clients' support experiences a little smoother, and gives them one less phone number to keep track of.
It's a complex decision involving a variety of factors, but I'm convinced that it's worth the time to explore, architect, and implement a hosting framework that works for our clients. Hopefully, this will lead to measurable reductions in project budgets in 2010.