Your websites are probably on servers owned by a hosting company that guarantees you “no downtime” because they have (ostensibly) an intricate system of backups to keep your site up 24/7. Facebook, Twitter, LinkedIn and the rest have zero contractual obligation to you to stay online. Yes, they have good reasons to not go down. But what will we do if they do? It isn’t as if we have many other major choices of social networks where we can disseminate our messages on a massive scale.

Sure we can hop from Twitter to Facebook, but what if Facebook also goes down? I don’t see a mass exodus to MySpace happening. Maybe we’ll hunker down in our niche networks or micronetworks like Sprouter or BizNik or a Ning community waiting out the storm. Maybe we’ll check in furiously on Foursquare, Gowalla, Whrrl and the like, hoping someone notices that we’re still alive. Maybe we’ll sit in our blog and send out an email blast inviting people to join us for some conversation in the comments.

What Are We to Do?

If you’ve ever used Second Life for business, you’ve probably faced chronic downtime on a network where you’ve have put creative energies and efforts into building up a storefront or holding a major event, only for the Linden Lab servers to go wonky and you’ve lost sales or had your event ruined. With Second Life, those of us who were entrenched in making it a viable marketing and commerce play for our companies would curse the company that we loved for creating the platform but hated for putting our ventures at risk. Then again, it has never been Linden Lab’s fault that we’ve decided to depend on the servers of a single company for major business transactions.

We Are at Fault

What are we thinking? We wouldn’t build our company headquarters on somebody else’s land that they owned without a contract, some guidelines, guarantees and stipulations. And yet we spend hundreds, thousands, tens of thousands of dollars creating assets and putting them on somebody else’s servers — and not even companies that are in the business of keeping things their servers up 24/7. They are in the business of getting you to be on their networks, because the more numbers they gain, the more powerful they become.

Our tolerance for their outages just goes to show how dependent we are on them. We think of these moments as par for the course because we are on the “cutting edge” and dealing with “new technologies.”  We’ve been mesmerized by the promise of exponential reach. We’ve been hypnotized by the potential numbers. We’ve been sold a bill of goods and come out of these moments of outage when some of our marketing and customer service efforts come to a screeching halt, wipe the sweat from our brows, smile feebly, put out our hands and say “Please sir, may I have some more?”

Can We Stop this Madness?

Unfortunately, no. We’re too far gone. We’re in this thing too deep. We’re intoxicated by those moments when we get some genuinely impressive results with our Twitter outreach or our Facebook engagement, and we salivate at the possibility of more of that good stuff. We’re hooked.

So if we can’t stop it, what can we do? Here’s my advice:

• Don’t replace your stable marketing tactics. Look at social networks and social media marketing as an augmentation of your traditional marketing, not a replacement. Don’t give up the rest of your outreach and marketing efforts.
• Have a contingency plan. Downtime on the big networks is inevitable. If you can’t be real-time on Twitter or Facebook, where will you go and where can your customers find you?
• Support niche networks. Yes, Twitter and Facebook are huge, but they’re not the only places offering the same kind of messaging capabilities. Set up shop in a smaller network and support their efforts to be the next Twitter or Facebook. They’ll love you for it and give you the best darned customer service, because they want and need you.

And keep in mind this mantra: Don’t put all your eggs in one basket. It’s an oldie, but a goodie.

What are you doing to better distribute your social media presence to safeguard against the Fail Whale?

Fail Whale screenshot from twitter.com

Related GigaOM Pro content (sub. req.): Can Enterprise Privacy Survive Social Networking?

These outages can quickly become crises. How do you know what step to take next on a project if the project is entirely managed on a site that is currently down? How do you get in touch with a client if you usually only communicate through that same project management tool? You may not be able to work quite as fast without your tools, but these tips can help you keep the situation from escalating.

1. Have an alternative method of communication for clients and co-workers. It may seem obvious that it’s important to have multiple methods of communication, but it’s worth going beyond just having an email address for the people you work with. Depending on just what tools have gone down, there’s a possibility that an email server has gone down as well. Having a phone number or another form of communication available — even being able to send a message through a social networking tool — may be necessary to be able to get through to the folks you need to talk to.
2. Have a backup of project data. All your project information is in the cloud, right? Even if the service goes down for a little while, you’ll be fine, right? Not necessarily. There are plenty of scenarios that end with you having to recreate your project from scratch (as could happen if the app doesn’t come back up in time, or it goes bust). Having a backup of your information can offer extra peace of mind.
3. Choose a tool to keep working. Many outages are open-ended, because it can be tough to tell exactly what is necessary to bring an application back online.If you’ve got deadlines, waiting around for something to happen is not the best option. Ideally, there may be some aspect of your project that you can work on without logging in at all. If that’s not the case, there are a couple of options: consult your backup documentation or your client to see what would be most practical to work on in the mean time.
4. Discuss a new deadline. If you’re in a situation where you really can’t proceed with a project — maybe the only copies of important files are inaccessible — you may quickly realize that your current deadlines just aren’t going to be doable. If that’s the case, talk to your client or manager as soon as you can and explain the situation. If you aren’t sure how much of an extension you’ll need (a typical situation if you don’t know when your tools will be available again or if you’ll have to rebuild the project from your backup), make that clear so that you’ll have grounds for future discussions, if need be.

Unfortunately, the best methods for dealing with any kind of application outage require some prep work while your most important tools are still up — getting contact information for a client who can’t respond to the only email address you have isn’t going to move things along. In the worst-case scenario, when an outage has caught you entirely by surprise, it may be difficult to come up with a next step that will actually help you with the work. The best option may be just trying to get in touch with your client or co-worker and check what can be done at all. In the event that you can’t reach anyone, you may just have to wait out the outage and deal with the repercussions down the road.

What do you do when your web apps go down?

Image by Flickr user p4inkiller

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• VMware’s Cloudy Ambitions: Can It Repeat Hypervisor Success?
• Infrastructure Overview, Q2 2010
• Defining Internal Cloud Options: From Appistry to VMware

In what has to be a developer’s worst nightmare, last week a “perfect storm” of events led to just that for task management service Toodledo. Thankfully, after what I am sure was a very long day for the Toodledo team, they were able to bring things back online and with no data loss.

As a user of Toodledo, I was impressed with the outstanding job they did not only recovering our data but for the way they communicated with us during what was really an unthinkable situation. Was it an inconvenience to be without the service? Absolutely it was. But in what really should be a case study in how to respond to such an experience, the folks at Toodledo provided frequent updates and explanations during the outage and a thorough debrief after the fact.

I suggest you take a few minutes to read through the 10 or so pages of user comments and responses to their explanation there. It is full of appreciation and understanding from users for being spoken to honestly and treated with respect. Rather than leaving in droves, more than a few people have suggested that they have upgraded from free to paid accounts because of the way that the situation was handled.

So what did they do right? Here are five steps on how to deal with your customers during (and after) a service outage.

1. Give us a descriptive error page and lead us to a forum where you are discussing the issue. Let us know there is an issue and that you are actively working on resolving it, especially if it is above and beyond something brief or minor.
2. Make updates regularly. We know that taking the time to write an update is keeping you from the actual “fixing” of things, but we’re scared about our stuff and want to know what is happening.
3. When it’s fixed, tell us what happened. We’re probably still a bit angry and feeling inconvenienced but we’ve had “stuff” happen in our lives, too, and will likely empathize.  Keep in mind, we may not be as technical as you, so give us the details in terms we understand.  Keep the jargon to a minimum.
4. Acknowledge that the outage was an inconvenience, and don’t minimize the effect it had on us. We use your service because it helps us and provides a benefit. When it is down, we lose that. I think that this is particularly important.
5. Tell us what is being done to keep the issue from happening again.

Despite everyone’s best intentions and planning, outages are going to happen. How it is handled and how you respond to your customers will define our response.

What do you expect from a provider during or after an outage? Tell us in the comments about your experiences.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Why iPad 2 Will Lead Consumers Into the Post-PC Era
• The Near-Term Evolution of Social Commerce
• Content Farms: The Players, The Benefits, The Risks