Reducing the Costs of Online Outages
Intuit’s major outage last week served as a harsh, yet valuable reminder for all of us about the serious risks which underlie our growing dependence on third-party, ‘cloud’-based services.
The irony about the timing of Intuit’s service failure is that the company has been escalating its marketing efforts to position it as a leading provider of cloud services. For instance, Intuit’s SVP/CTO, Tayloe Stansbury, was a keynote presenter at last month’s SIIA All About the Cloud Conference in San Francisco where he boasted about Intuit surpassing $1 billion in cloud-based service revenues.
The company’s CEO, Brad Smith, did a good job quickly apologizing to Intuit’s over 300,000 customers who were affected by the outage. His blogpost also explained the cause of the disruption and made a public commitment to put new policies and procedures in place to prevent the problem from occurring again.
Smith’s response was certainly better than that of BP’s CEO, who has succeeded in alienating public officials as well as the public at large with his arrogant actions and unconvincing statements.
In order to counteract any potential backlash as a result of this incident that could derail the rapid growth of the market, the cloud computing industry needs to take a number of immediate steps to reduce the likelihood of similar events and alleviate the impact on customers.
As a strategy and marketing guy, I’m not qualified to recommend specific technical steps which Intuit and other cloud vendors should take to mitigate the risks of software or system failures and safeguard against manual process mistakes.
Instead, I will suggest some simple measures which vendors must take to reduce the occurrence and repercussions of these incidents:
- Invest in redundant systems even if it adds operating costs. Smart customers are willing to pay more if they are convinced they will be safer.
- Train and certify employees as meeting industry standards. Smart customers will select vendors who boast skilled employees and have the certifications to show for it.
- Act fast to notify customers when something goes wrong and keep them informed of your efforts to resolve the problem. Customers understand that ‘stuff’ happens and appreciate vendors who are proactive in responding to an issue and communicate about the status.
- Ask for customer feedback and suggestions. Customers are loyal to vendors who seek and respond to their input.
- Create public notification systems and performance scorecards.Salesforce.com has pioneered this idea with its trust.salesforce.com website which provides real-time reporting of its service availability. Customers love transparency.
- Develop Service Level Agreements (SLAs) with teeth. Customers want contractual agreements which clearly state the provider obligations, outline their escalation policies to remedy issues, offer meaningful penalties, and prescribe grounds for termination.
Although a strong argument can still be made that this type of outage is rare and most cloud computing providers are far more reliable (and secure) than the majority of enterprises, this latest incident makes it imperative that the cloud computing industry quickly develop off-line solutions which augment their online services to enable their customers to continue to operate even during a service outage.
I’ve been advocating for more than a year about a new definition of location independence which permits customers to take advantage of off-line, on-premise SaaS alternatives such as appliances and applets.
These on-premise platforms and tools are becoming increasingly feasible from a functional and economic standpoint as the technology, service delivery methodology and business models mature. They also are in keeping with the growing interest among corporate decision-makers in private or internal cloud options.
While the idea of private or internal clouds undercuts some of the potential economic and operational benefits of the cloud computing ideal, it is still a step in the right direction and can alleviate the angst many business executives and IT folks have about the current state of the cloud, especially in light of Intuit’s latest problems.