-
Hosting Downtime and Competition
July 9th, 2009 : Rich Miller
Hosting and colocation are highly competitive businesses. That competition gets heated whenever a prominent provider experiences an outage, as rival providers seek to poach customers who are unhappy their site is offline and searching for news about the outage.For many years “rescue marketing” was commonly seen in Google keyword searches, as hosting companies would purchase text ads tied to keywords, including trademarked brands of a provider experiencing an outage. Frustrated customers Googling for information would encounter hosting offers from competing firms.
In recent months the trend has shifted to Twitter, where rival hosts have begun pitching their services using hashtags, the topical phrases included in Tweets (such as #fisherfire) to aid topical searching. By adding an outage-related hashtag to an “adverTweet,” some providers have sought to market their services to customers affected by outages.
ServInt brings an interesting perspective to outage-related marketing in a blog post today, in which CEO Reed Caldwell discusses the recent outages at Rackspace and reflects on the post-downtime scramble for customer loyalty.
Caldwell notes that ServInt has been in business for 15 years and suffered a major outage in 2004. “We have had honest, healthy competition from friends and colleagues and we have seen disingenuous poaching by hundreds of companies who – surprise – aren’t around anymore,” Caldwell writes. “What we are seeing now is a reminder of how a responsible company handles a serious issue, and how some companies try to take advantage of that.
“(Rackspace) had a bad week, but are holding themselves accountable and encouraging their customers to do the same,” he continues. “Ultimately, the true test of a company is not how well it does at the top of its game, but how quickly it gets back up. Rackspace will get back up and we look forward to it.
“There are several companies, if you can even call them companies, who have been in business for less time than a stale pot of coffee and are throwing mean spirited, transparent promotions out to justifiably angry customers. We feel this is not only in bad taste, but it is unethical and an excellent testament to how they view their fellow hosting providers. You attract customers by providing great service and thereby earning it, not by bashing someone else’s.”
Is “rescue marketing” a winning strategy or a shady tactic? Do customers respond to these pitches? In our experience, customers’ willingness to jump ship is closely tied to the details of the incident, and whether it was avoidable. But it is also influenced by what they learn about their provider during an outage. We’ve seen plenty of examples of both good and bad customer communication during downtime.
What’s your take? Share your thoughts in the comments.
Nice article. I agree that competition is good and that organizations should focus on the delivery of their services instead of trying to step all over their competition (especially when they’re down).
Ken Cameron
Posted July 9th, 2009First, the outages at RackSpace, Fisher Plaza, and Google were inexcuseable for companies who want to grow up and be enterprise class service providers. Enterprise class data centers have completely redundant environmentals, usually at an N+2 level. There should be NO-SUCH-THING as an outage caused by a single UPS or Generator failure, or even ANY failure in a single electrical room in a data center (where was the redundant electrical room?)
Second, the handling of all of these situations pointed to a compete lack of Disaster Recovery. This is, most likely, partially to blame on the customers who didn’t demand clauses about DR, but that is no excuse. Authorize.net processes credit cards! The RTO (Recovery Time Objective) for credit cards (esp. Authorizations) must be minutes, at most, an hour. Every second of downtime is lost revenue. That is a lot of justification for good DR.
Sure, competition is steep in this service provider arena, and, so they all cut costs to the bone. If these providers want to catch the big fish, they will have to figure out how to offer realistic enterprise level service.
Ken Cameron
LinkedIn: kencameron
Twitter: itklcameron
Thanks for your post. We applaud ServInt for both their sentiments, and the fact that they are an honorable competitor. You have to respect honest competition!
When we experience customer issues we focus on customers first, communications a very close second. Get the servers up, get the message out. But we also realize that real-time status is becoming more and more expected – so we are delivering that via Twitter, our blog, etc.
We will NEVER respond quickly enough to customer outages. We want to eliminate them. But when they do occur we want to be open, and honest and we want to do so as quickly as possible.
Thanks again,
Rob
Rob La Gesse
Director of Customer Development
Rackspace
210-845-4440
Chuck Smith
Posted July 10th, 2009While the spirit of Reeds comments are well intended, they are not really realistic. Research how Ford has stepped up thier marketing and incentive programs to exploit GMs bankruptcy. Attempting to “swoop in” when an industry leader trips has been a part of the competitive landscape for centuries. Like others I agree that Rackspace handled thier outages as well as can be expected and should be applauded for thier openess, but that certainly doesn’t give them a pass from normal competition.
At the end of the day, tier power chain (or components thereof) failed. In an industry where we all claim “N+ this” and “Tier that” and “five nines”, you have to expect to be the target of a marketing attack when your “stuff don’t work”.
This is an excellent discussion on a very relevant and timely topic. I have to wonder however if we would be discussing this today if clients demanded to see certifications by The Uptime Institute before selecting a hosting/colocation provider in the first place.
It is a mystery to me why anyone running a data center would spend the tens or hundreds of millions to build it, and stop short of the tiny cost of a qualified third party review by the people that designed the Tier standards in the first place. If hosting providers want to talk about a more professional competition, shouldn’t this include being transparent with the customers?
The Poor Customer
You can’t simultaneously stand up for RS’s and claim you’re customer-centric.
We feel terrible for the many customers who were promised and sold uptime (self-healing networks, world-class data center etc.) only to have power to their servers lost. Our hearts go out to them. We do not feel sorry for RS and we think it’s an arrogant and self-centered to do anything but talk about the appropriate damaging of credibility to the entire industry.
Every vote of confidence for RS in this event is a vote AGAINST the customer and against the industry as a whole. We as an industry have to hold ourselves accountable for uptime or the entire industry will wither.
It’s not just RS customers who suffer when they have an outage. All of us lose too. How? Because the pie (the market) is growing faster than any of us can take clients from each other. That’s the good news. Unlike really competitive industries (airlines, telecom, automobiles) we don’t have to worry too much about trying to eat each others’ slice of the pie when the pie is growing. So, when RS, the “gorilla” in the industry, with by far the largest balance sheet on the planet in this market has multiple power outages, everyone asks “Maybe we should bring our servers back in house. Heck if RS can’t deliver power reliably how can anyone else? Maybe it’s just not a good idea.” The pie stops growing and all we can do is begin to go after each other’s slice of the pie.
When the biggest dedicated provider of managed servers to the world can’t keep the power on the entire industry hurts. We should be irate with RS and their failures, not apologizing for them. I feel sorry for their customers but not for RS.
So are we pissed at RS? Yes. Should they be incredibly trite, apologetic to all us competitors and all our clients as well? Yes. Without customers, there is no RS, no Online Tech and no managed hosting industry.
Watching this play out has been quite inspiring in terms of a more positive spin to marketing and business intelligence. I work for a boutique data-center near Naperville, Illinois and our customer base is such that Rackspace isn’t a direct competitor, but seeing how their situation has pointed to general aspects of the industry is very interesting.
Watching how Rackspace has managed their accountability in the situation and seeing examples such as ServInt’s response has given me a new perspective on business communications. Social networks have sped up the communication loop and we are all subject to failures when it comes to operations, whether it is down time or bad customer services, some things are just out of our control. It’s inspiring to see companies supporting each other and driving each other to better service rather than going in for the kill when things get tough.
After all, isn’t it about the service that the end user gets rather than who provides it? The best rise to the top, and the more ethically they do it the better it is for everyone.
Yan Ness: It’s nice to see a comment submission from Online Tech that isn’t a blatant sales pitch including the company’s phone number. Pretty clear where you guys stand on this issue.
There has been a number of problems this year, from the attacks on government websites that shut down their service, data theft, and power failures due to the aging electric grid. Rather than become angry over it we all need to step up the damage control.
These challenges are only going to increase in frequency as we become more reliant on data-centers. The e-Health initiatives alone are going to cause a major upsurge in data center activity. If we don’t help each other sort this out now things are going to only get worse.
There’s no excuse for failure, but don’t you think in-fighting and anger will only lower credibility in the long run?
RackSpace CEO makes a very accurate statement that should not be underscored when he says “Ultimately, the true test of a company is not how well it does at the top of its game, but how quickly it gets back up”. Every single data center and data center professional has seen some sort of emergency downtime in their career, so blame for one mistake is a sign of inexperience. Blame for the same mistake, that is another story. RS has earned their position by years of un-interruption and great service, and how they handled the event and communicated with their clients is how their clients should judge them.
In response to the question posed by the article – the fact that some will kick you even when you are down is inevitable. It has been going on for year and those that use less than ethical marketing tactics might have a few short-term wins, though they won’t be around 5 years from now. We should all work hard to educate our clients to the different costs of risk mitigation (Tier1 colocation is way less expensive than Tier4 colocation), how we will communicate with them in the *very unlikely* event of a problem regardless of how large or small, and we should all support our fellow data center professionals that are faced with the same challenges that we are today.
Everything should be 2N+1, including your customer communication. Using multiple channels to communicate with customers, in the worst event of downtime, is crucial.
Twitter provides another external communication delivery mechanism to clients.
I used to run a tiny hosting company (2 servers, about 100 customers). IMHO, there’s no such thing as 100% uptime. I certainly don’t know of a hosting company anywhere that has never had a major outage.
A responsible hosting company can put in all kinds of redundancy and backup policy and equipment – only to get hit by the Black Swan six sigma event.
I think RackSpace’s response is a credit to a company that is transparent and customer-centric.
RESOURCE LINKS:
