Problem Cost Model

[cproctor]

Hi everyone,

I was just wondering if anyone had implemented a problem cost model. I'm stuggling to develop something that I think is fair given that many elements could be subjective.

Any tips or examples on methods that have worked for anyone would be great.

Thanks

[Marcel]

We have set up a costing model that is currently used for some parts of our organization and that we plan to further improve and expand.

The costing calculation is done by gathering a number of metrics that each translate to a dollar value. Areas currently covered are:
- lost productivity (hours lost on the business side x labor cost per hour, with differentation for certain functions)
- lost revenue (downtime/customer impact x associated value per unit of time, based on parameters already used in other areas of the business)
- support cost (incident volume x average cost per incident)

These calculations are typically done shortly after a problem has been identified and play a role in determining impact. We do not estimate repair cost at that time, nor do we currently track repair cost after the fact on a per problem basis. At the time we also do not look at potential legal cost (lawsuit costs, fines, etcetera) since these are very difficult to estimate up front.

Keep in mind that cost does not necessarily mean out-of-pocket expense. It is just an expression of the dollar amount associated with a problem.
_________________
Manager of Problem Management
Fortune 100 Company
ITIL Certified

[Diarmid]

cproctor,

could you flesh out your question with some context including what you want to use the model for and how you define "problem"?

There are costs associated with managing the problem, the problem management system, problem investigation, defences to mitigate the problem (workaround and prophylactic measures), resolution requirements (could be hardware, software, training, etc.), implementation (could include project management, use of contractors, service downtime, etc.), problem review...

If you show what you have already done, it might be possible to comment.
_________________
"Method goes far to prevent trouble in business: for it makes the task easy, hinders confusion, saves abundance of time, and instructs those that have business depending, both what to do and what to hope."
William Penn 1644-1718

[Bluesman]

Interesting subject. I am also curious. We try to show that some problems really are worth solving, since the (financial) impact of recurring incidents are magnitudes larger than people think.
But:
A university is in some respects very much a "business", and in other ways not. It´s hard to build a model based on this mix.

We tried to make a very simplistic model for cost of downtime of a service, including 3 main user categories: students, administrative staff and teachers/professors et al.

Already in the student case, the parameters are so many and so widespread (inter-dependant and rubberband-factored) that they are more or less hopeless to model correctly.

And the staff cost: a nightmare. Just looking at lost "productivity" raises a lot of issues, such as a) if one essential service for one group of users fails for half a day - can the person still be "productive"= do something else of value to the "business"? 80% productive? 20%? b) what if 2 services fail? 3? Where is the point where people just give up and go home?

Actual cost - on the staff side - is a bit easier. Wages can be averaged, office/location cost can be averaged.

Teachers: nightmare time again. A teacher who cannot use service X for a lecture means downtime for himself, AND for lots of students, who maybe travelled far to attend. And the lecture will have to be rescheduled, meaning even more expenses to all parties. Again, wages and office/classroom waste/cost is the easy part.

Then there is the badwill factor. If a uni with 15000-20000 students and 1500 staff start moaning about constant printing problems and how bad this and that service works..how much direct loss of income (less applicants=less government money) will this generate?

So - this is an interesting subject. I am curious to see what the more knowledgeable amongst you will come up with.

/Richard
_________________
---------------- bragging line -----------------
V2 Combined Practitioner (service support)
Problem Mgr

[IceCreamMan]

We have started down this road on a simple level based on hours of personnel utilised to solve the problem. While I do not know the actual cost (salaries of individuals) we can get acurate results using hours.

Problem managers time
Engineers time (junior middle senior)
Consultants time if applicable.
Cost of solution.

Simple thus far but its a step forward on the right road.

Cheers

[Bec74]

Perhaps the key is having a good view of the cost of incidents first. Whether you measure less than optimal productivity (staff working on unplanned work), your resolver teams cost, customer contact when zero touch channels fail etc does not matter overly...as long as you have good assumptions and buy in..& apply the same logic consistently.

Then the figures associated with Problem cost (whether you choose time or another metric set) should able to be netted against incidents (i.e by reducing the likelihood or impact then you should be delivering a cost avoidance benefit..).

as problem records are often fairly open ended when you are still in investigation (til you find the root cause hard to tell how long the whole thing may take.. and not SLM bound) be sure to check your NPV along the way- don't throw good money after bad.

This approach should help you determine whether you are possibly over-investing in PM; i.e where the ongoing cost of the workaround or likelihood of re- occurrence so low that PM is not justified.

In the credit crunch the only vacancies my org of 5000 has approved in four months for new hire have been in SM- largely because we have credible metrics about the $ value we deliver.. and we are only at month seven of implementation of IM & PM.

Approved now for change, release and CMDB to progress based on our initial demonstrated cost avoidance return from IM & PM activity.
_________________
If you cannot find a root cause, make one up. Either you will be right or everyone will scramble to disprove you.. either way you will be further ahead