Really random and quite dull but are there any Payroll people about?

[catgirl1976]

My payroll team (just taken them over and know little about payroll) run a variance report every month which dogs where people’s pay has varied from the previous month so it can be checked. Makes sense

But the variance level is set to 60 percent. This seems crazily high to me. Most people are on standard contracts with very few doing variable hours (maybe 20 percent of the workforce). Is 60 percent variance standard as a tolerance? I would have thought a lot lower but Google’s no help and payroll is not my area

would be really interested in what other organisations use ? Thank you

If it's your team then try it with a lower tolerance and see?

It is and I will but I wanted to check that wasn’t standard (or what is) before I was like “That’s crazy Janet, make it 20 percent”

60% is standard for us however in our line of work people get lots of overtime, enhancements and allowances so anything lower would flag up normal variances and be a pain in the arse to check. Try it at 40% and take feedback from the team?

I run ours at 20% but most salaries are stable and not much overtime. I just apply different filters and can adapt it if I want to just to check from time to time

It should be easy enough to analyse the spread of variances across the last few months payrolls- have a look at what the standard deviation is and set your tolerance accordingly. However I also believe 60% is too high (I work in finance and deal with payroll monthly but most of my postings are for those on standard contracts without overtime or other upcharges).

Thanks that’s useful. 60 percent just seems crazy. When I asked why it was so high they said “well there’s be more work if it was lower. Well….yes….but there might also be less errors which are a bit of an issue at the moment hence you all moving over to me to sort out 🙈

I’ll do some analysis of the last few months and see what would be reasonable

We look at ANY variance from gross salary/12. That's for about 280 staff. Vast majority are on the same each month but we probably get around 40-50 differences. Filter on the diffs and make a comment next to them to say what the reason is. (They might be a starter/leaver in the month, or get on-call pay if they're in the support team, or be given some sort of bonus/one off extra payment as a thankyou).

🤦‍♀️ good luck OP!

I used to be an auditor (many years ago when it was still largely transactional testing). For a salaried payroll we would variance test at 10%. For an hourly paid and overtime led payroll we would look at seasonality and test at 25% but based on up to 60% spread on an annualised pattern. 60% on a salaried payroll is only going to pick up extremes not give any sensible analysis.

I’d drop your checks to 10% and have a very close look. I’d also run a few reports looking for duplicate bank details and do a check for ghost employees, but I am cynical like that.

I think maybe ten percent to start with to flush out any and all issues and then maybe we can look at upping it. I wouldn’t be surprised if there’s all sorts once the box is opened. And don’t get me started on the mess pensions are in 🙈. I must have been very bad in a previous life to get this wonderful “opportunity”

Mmm, if you are looking into pension contributions I would analyse employees and employers contributions and check against policy. I’ve definitely seen employees drop their matched contributions to zero, but keep the employers at the maximum .

I’ll check that too - thank you!

Seems to be a massive issue with leavers being processed properly and waiting ages for their pension. Plus all the salary sacrifice stuff looks wrong to me - got a horrible feeling it’s not been taken out of gross but net. Praying it’s just a few bikes and not everyone’s child care vouchers going back a million years

<wine>