What does your Service Level Agreement look like?
-
@Ambarishrh said:
So if i understand correctly, the best approach is to charge an hourly rate no matter if you do servicing for desktops or servers, based on calls from the customer. What i didn't really understand is what about the regular maintenance works you do for the clients? Like what mentioned previously, av scan, patch management, other maintenance tasks etc which we do as part of our ongoing management. Does it also covers under the hourly rate? If so how do we agree with the client on the billing, just calculate the hourly rates used for all these and send out a bill on monthly basis? The only difference on the bill would be when there is something broken, or new additions to the existing setup i believe?
So I would say that best practice is to tailor to the situation. That being said... overall I believe that hourly is the best with options like minimums, pre-arranged work and estimates and similar. So having, say, a forty item maintenance task list and a pre-arranged minimum and range of hours and some flexibility for addressing issues or special items (we found something that needed special attention but was minor so escalation would be ridiculous) would often be ideal.
But this is "optimum for the best relationship" and still has caveats - the service provider still is encouraged to find extra work to do, for example, and is not incentivized to automate what can be done manually or to invest in tooling to reduce labour as labour means profit and tooling does not.
There is no singular best answer. Part of the goal of moving to an MSP or ITSP model is to improve over internal staff. Internal staff are paid on an hourly system with minimums (e.g. hourly) or flat rate like an MSP (salary.) The same issues exist internally but with slight differences.....
-
Internal staff when paid hourly are generally either incentivized to reduce workload so that it falls below their minimum hours so that they have "free time" to relax, increase productivity, deal with other issues, or whatever. If they get 40 hours a week, this makes them want to find ways to work less than that or, at least, to find ways to need to work less than that. It doesn't mean that they blow off the remaining time, it could mean using it for more fun projects or simply not pushing resources to the wall and having no buffer.
Or internal hourly staff is encouraged, if they desire more money, to have overruns as extra hours, especially extra hours with overtime, goes totally into their own pockets.
-
ITSP arrangements can mirror employee situations whenever necessary by moving to dedicated or one to one staffing options. So, for all intents and purposes, you can always have an ITSP arrangement meet or beet an employee one in this way as long as the relationship is a good one. You give up no options. But you gain others, as a company bringing in an ITSP.
The MSP flat rate model or a hybrid of the flat rate model plus hourly is not completely ideal, but it is also far from the worst model. But it requires very clear scoping and a good relationship to really make work.
-
The worst model that I have seen is the project model. That's where things really get bad. In the MSP model you general deal with very common and quite straight forward scope that is spread over all or nearly all customers keeping scoping overhead down considerably. With project work all scoping cost is done on a one time basis making overall cost skyrocket and introduces risk at an unbelievable scale.
-
@scottalanmiller said:
...So having, say, a forty item maintenance task list and a pre-arranged minimum and range of hours...
I have seen IT Shops that will automate the maintenance, and still bill for 30 minutes or an hour of their time per task. Things like Scheduled Snapshots and then automatic Windows Updates... It is good for the customer because their equipment is kept patched and up to date... Good for the ITSP when the updates don't cause issues as they get to bill the client for the task and they don't have to spend but a few minutes confirming that it completed correctly.
There is no singular best answer.
It is rare that is a single, always, absolutely correct answer... (to go with your best practice comment above)... Tailor it to the situation.
-
@dafyre well one way that that can make more sense is that the MSP or ITSP can, if the arrangement is a good one, invest time up front in automation as long as they know that the customer is going to stay around and bill for the automation work that was already done coupled with monitoring and maintenance of the automation rather than for "hours doing the work" or whatever.
-
One of the hardest things to do is to figure out how to get both parties, the vendor and the customer, to mutually benefit from the same things. You want automation to be good for the customer, of course, but it also needs to make sense for the ITSP. Getting the ITSP to make investment decisions and doing the investing is important, but if you don't do it right then the customer ends up paying way too much for things that they could have just paid to have had automated. So you have to balance it in a way that allows the ITSP to work to improve things while the customer still benefits to some degree.
-
@scottalanmiller said:
One of the hardest things to do is to figure out how to get both parties, the vendor and the customer, to mutually benefit from the same things. You want automation to be good for the customer, of course, but it also needs to make sense for the ITSP. Getting the ITSP to make investment decisions and doing the investing is important, but if you don't do it right then the customer ends up paying way too much for things that they could have just paid to have had automated. So you have to balance it in a way that allows the ITSP to work to improve things while the customer still benefits to some degree.
This could be a billable project for the ITSP to do for the customer as a trial run to help build the relationship with the customer. One of the first things an ITSP or MSP should do when they onboard a customer, IMO, is take an inventory of what the customer has, and a short list of simple things that relieve pain points for both the customer and the ITSP.
For instance, a client of mine had some issues with their servers randomly rebooting during the day. It turns out somebody had set windows updates to deploy every day at like 2pm instead of 2am. They also had issues with one of their workstations with a flaky hard drive that was throwing errors in the event viewer...etc.
As part of our initial walk through, I pointed out those issues and we talked about what to do and worked up a project that entailed fixing those issues and noting any new ones that we found. My Pops and I did the work, and the client brought us back a few more times for fixing other issues that came up.
-
One thing, I think that is most excellent for an ITSP / MSP is to provide a face to the clients. Make sure you work out in the agreement whether or not you need to come on-site once a week or once a month or something like that.
-
@dafyre said:
One thing, I think that is most excellent for an ITSP / MSP is to provide a face to the clients. Make sure you work out in the agreement whether or not you need to come on-site once a week or once a month or something like that.
That's a tough one. It makes a lot of sense but it also increases the cost. It makes it easier for another MSP to undercut by not offering that. It's always tricky in a space where so many customers look at dollars and not results. Not that that is good or that you want those kinds of customers, but it only takes one weak moment of someone seeing cost cutting to jump ship and the relationship is lost. It is a careful balancing act between raising cost to babysit customer's emotional needs versus simply providing the services that they need.
-
@scottalanmiller said:
@dafyre said:
One thing, I think that is most excellent for an ITSP / MSP is to provide a face to the clients. Make sure you work out in the agreement whether or not you need to come on-site once a week or once a month or something like that.
That's a tough one. It makes a lot of sense but it also increases the cost. It makes it easier for another MSP to undercut by not offering that. It's always tricky in a space where so many customers look at dollars and not results. Not that that is good or that you want those kinds of customers, but it only takes one weak moment of someone seeing cost cutting to jump ship and the relationship is lost. It is a careful balancing act between raising cost to babysit customer's emotional needs versus simply providing the services that they need.
It also depends on your relationship with the customer. Do they automatically jump ship to the new guys, or do they say "Hey New Guy is offering to do this for me for 50% less than you. What can we do?"
"Well, we can stop coming on site once a week and then we can match that," Says the ITSP.
With switching to a new ITSP, they are moving to an unknown -- this guy is CHEAP... but is he worth his salt?... Giving their current ITSP (assuming a good relationship) a chance to match it would be more beneficial to the customer, I'd think.
-
@dafyre said:
With switching to a new ITSP, they are moving to an unknown -- this guy is CHEAP... but is he worth his salt?... Giving their current ITSP (assuming a good relationship) a chance to match it would be more beneficial to the customer, I'd think.
If you deal with enterprise customers, there is no problem here. The problem is in dealing with the SMB market which, as I often say, remains SMB for a reason mostly. So making illogical and non-business sound decisions is sadly a hallmark of the market and as an ITSP in that market you have to account for the fact that with the majority of customers, logic, good service, protecting them are not things that will often be taken into account.
