Service Level Agreements, Jamba Juice and Vince
Those who know me understand that I have very special feelings for the town of San Mateo, CA, where I once spent many a day representing a certain startup. There is no emoticon I can paste here to adequately convey them. I thrill when I see the Fashion Island Blvd exit from the 101. I long for the rough, ever-so-slightly pilled sheets of the Hilton Garden Inn. My heart swells when the sun sets over Sam Mateo, turning the beige stucco, low-rise office buildings and temporary housing units to a burnished Tuscan beige, distinguishing them from all the other beige stucco low-rise buildings in Silicon Valley.
San Mateo is most special to me because it has my favourite Jamba Juice franchise, located close to the San Mateo Bridge entrance. If you look away from the road too long to check whether your protein boost has been thoroughly stirred into your Mango A-Go-Go, you will find yourself spending the next 7 miles driving across the bridge to Foster City.
My colleague Vince was a lousy navigator and as a result, he and I spent many hours taking the 14 mile round trip to Foster City after our morning Jamba Juice runs. Which is where we would debate Service Level Agreements, among other things.
It's important to know a little about Vince: he and I were the "wet ops" team for this particular startup's sales force. Once the sales people had finished overpromising, we'd come in and eliminate what we could so that the services we committed to deliver were more in line with, well, what the company could actually deliver. We became masters of padding the service specification and the service level agreement (the "SLA") so that they were, by the end, very forgiving documents.
In my view, new companies often prepare SLAs for their customers that are far more demanding (and potentially costly) than they need to be. This is partly because many startups hire from the customer side, and those new employees bring that "5 nines" mindset with them. Here are the basic principles I think new companies should bear in mind:
1. There is an increasing emphasis by acquirers on the support environment companies provide. The scope of service availability and other support commitments (and any penalties or credits provided to customers), is now being considered in due diligence and calculation of the purchase price.
2. No one needs 99% service availability. Think of your most co-dependent girlfriend; did she really need you to call her before she went to bed? Or to see if her cat's check up went well? A reasonable average availability in any 24 hour cycle should be adequate for your customers.
3. Termination of the agreement should be a last resort only in the most dire circumstances. I see many SLAs in which the company has agreed that the customer can end the agreement for any breach of the SLA. Termination should be available only for breaches of critical service levels for key services. If required, you can consider giving credits for other service level breaches. (Note: Vince advises that in Europe, contractual penalties for failures in network service availability are unenforceable. I am reviewing whether this extends to hosting and application service levels also).
4. If your company exceed customer support service levels, consider asking for a bonus payment.
5. Reporting forms change often in a company's early stage. Try to avoid agreeing to a specific format (agree to metrics but not form) so that you can remain nimble as you perfect your reporting operations.
My two cents.
San Mateo is most special to me because it has my favourite Jamba Juice franchise, located close to the San Mateo Bridge entrance. If you look away from the road too long to check whether your protein boost has been thoroughly stirred into your Mango A-Go-Go, you will find yourself spending the next 7 miles driving across the bridge to Foster City.
My colleague Vince was a lousy navigator and as a result, he and I spent many hours taking the 14 mile round trip to Foster City after our morning Jamba Juice runs. Which is where we would debate Service Level Agreements, among other things.
It's important to know a little about Vince: he and I were the "wet ops" team for this particular startup's sales force. Once the sales people had finished overpromising, we'd come in and eliminate what we could so that the services we committed to deliver were more in line with, well, what the company could actually deliver. We became masters of padding the service specification and the service level agreement (the "SLA") so that they were, by the end, very forgiving documents.
In my view, new companies often prepare SLAs for their customers that are far more demanding (and potentially costly) than they need to be. This is partly because many startups hire from the customer side, and those new employees bring that "5 nines" mindset with them. Here are the basic principles I think new companies should bear in mind:
1. There is an increasing emphasis by acquirers on the support environment companies provide. The scope of service availability and other support commitments (and any penalties or credits provided to customers), is now being considered in due diligence and calculation of the purchase price.
2. No one needs 99% service availability. Think of your most co-dependent girlfriend; did she really need you to call her before she went to bed? Or to see if her cat's check up went well? A reasonable average availability in any 24 hour cycle should be adequate for your customers.
3. Termination of the agreement should be a last resort only in the most dire circumstances. I see many SLAs in which the company has agreed that the customer can end the agreement for any breach of the SLA. Termination should be available only for breaches of critical service levels for key services. If required, you can consider giving credits for other service level breaches. (Note: Vince advises that in Europe, contractual penalties for failures in network service availability are unenforceable. I am reviewing whether this extends to hosting and application service levels also).
4. If your company exceed customer support service levels, consider asking for a bonus payment.
5. Reporting forms change often in a company's early stage. Try to avoid agreeing to a specific format (agree to metrics but not form) so that you can remain nimble as you perfect your reporting operations.
My two cents.
posted by Suzanne Dingwall Williams at
10:31 PM
<< Home