# What spam really costs, Part II

I hope you did the homework I assigned last week. I wanted you to find out things like what your data services cost a month, what percentage of your available WAN bandwidth is used and what percentage of that bandwidth is used for email.

I hope you did the homework I assigned
last week. I wanted you to find out:

-- What your data services cost per month.

-- What percentage of your available WAN bandwidth actually is used.

-- What percentage of that bandwidth is used for email.

-- What storage costs you.

-- How much spam each user gets each day.

-- How long it takes a user to deal with a single piece of spam.

Those were the basics. To complete this exercise you'll also need to know the average fully burdened cost of an employee (taking into account insurance, tax, benefits, etc). Now divide that by the average number of work hours per year to find the fully burdened average employee hourly cost.

For my calculations, I'll use some averages derived from a number of sources that might or might not be consistent with the values for your operation.

First we'll calculate productivity costs. If the average user (working 220 days per year at a fully burdened annual cost of \$US71,440) receives 100 messages per day of which 30% are spam and the average time required to handle a piece of spam is 5 seconds, then the average cost to handle spam per user, per day, is \$US1.69.

This means that for 1000 users the total cost of productivity loss per annum will be (hold on to your seats) \$US372,083 or \$US372 per employee! But while productivity loss is by far your biggest cost, the others are nothing to sniff at.

If the cost of internet connectivity per month runs to, say, \$US6000 (servicing internal users and telecommuters and mobile workers) and email uses 75% of the bandwidth and spam uses 30% of that, the total cost of bandwidth used by spam (spamwidth?) per year will be \$US16,200.

Next, if the average message size is 50Kbytes and the cost of storage per month, per gigabyte, (including management costs) is 50 cents, then it will cost \$US9000 per year to store spam.

Finally, support costs: Assuming the average cost of support per user, per year, is \$US250 and 5% of that can be allocated to solving spam-related issues, we'll be spending \$US13 per user, per year, for a total of \$US12,500 per annum.

If we add all of these annual costs - connectivity at \$US16,200, storage at \$US9000, support at \$US12,500 and the big one, productivity, at \$US372,083 - we get a total cost of spam per year for a 1000-seat operation of \$US409,783! That is more than \$US409 per employee!

I've made a spreadsheet available at www.nwfusion.com, DocFinder: 6644, so you can plug in your own figures. My spreadsheet also goes further by extrapolating the model over the next five years for growth in connectivity costs, storage costs, support costs and spam.

Note that if spam represents 30% of messages today and the spam growth rate is 35% per annum, spam will become 99.6% of email messages by 2007!

The point of this exercise is to build a case for action. Armed with this kind of analysis, you can approach the CEO and present an argument for investing in messaging management services. Along with the financial argument you should be able to present the liability issue: Your organisation could find itself in a sexual harassment case for not actively stopping spam.

Note that I used the term "message management services" rather than "anti-spam services" because just minimising spam is not enough. That is just one component of your messaging problems. Other issues include confidentiality, personal use, improper content and violations of your acceptable use policy.

That's this week's homework. Hand in your assignment as soon as you can and I'll give you a grade.

Report to the principal at backspin@gibbs.com.

Or

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Tags spam