[booth]

Hello!

I'm graded at my job by how much I sell in seven categories. We lose points when we miss quota, and gain points when we beat quota, weekly. My scores fluctuate wildly, because I'm part-time and only make 20 sales a week, where full-timers make 200 or so sales per week.

So, my bosses contend that over the course of a few months, my weekly averages should represent my ability when averaged together. I contend that, as it is near impossible to hit all seven graded categories in 20 transactions, the points I automatically lose will artificially lower my score, especially in "rare sales" categories with a quota under 10%, if not on "common sales" with quotas around 60%.

For example, if I had only one sale per week, I would fit no more than three categories at best, which would be a great sale. But, having a zero in any category causes you to lose points, and although I had one great sale, I would be in the negatives every week. Thus, even if you took 200 weeks of one sale, my average would be artificially low.

Who's right, and why? And what formula could be used to re-balance the equation?

Thanks!
Booth

[reinsuranc]

The full-timers sell 200 widgets a week. You sell 20 widgets a week. If the full-timers are measured weekly, and you are measured over a ten week period (10=200/20), wouldn't this be fair to both of you?

[booth]

The full-timers make 200 transactions per week, I make 20. The sale could include any number of widgets, AND/OR gimcracks, geegaws, or knickknacks.

More specifically, it's a video game store. I'm ranked on how many used games I sell, how many system warrantees I sell, how many magazine subscriptions I give away, how many discount cards I sell, how many game reservations, and other things that slip me right now. You're ranked on your percentile in sales; they don't care how many transactions you make.

You gain/lose up to five points either way on seven categories, so the score range is -35 to +35. A good employee averages hitting one or two categories on most sale.

To take the eaxmple from my first post, if I only had one sale, and hit three catergories, that would be a remarkable sale, but I'd still be -5 for the week, because I still missed four catergories. So if you took 200 transactions, yes, it would be fair, but if you took 200 weeks of one transaction, it would be still be -5.

I asked to be graded monthly, and they won't do it. It's not a big deal, it's not my main gig, but the statistics question won't leave my brain alone: is there a formula to compensate for the small sample size, or must I be content knowing there is a huge margin of error?