.webp)
I’m talking about all the other inputs: benefit files (read: local spreadsheets), separate tickets, emails, and perhaps a few lost documents (e.g. severance agreements, onboarding documents, employment contracts, sick leaves). You have since then validated the data (going back many times to the submitter) and transformed it into something your providers can process.
Now, you have to WAIT… sit there and wait for the results to come in. Don’t get me wrong: I know this wait doesn’t equal not being busy. You are saying no to last-minute inputs (you are, right?) and catching up on all the work you couldn’t do in the squeezed part of the payroll calendar from cut-off to submission. But how long do you have to wait for the providers (assuming you have outsourced payrolls) to get your trial results?
We asked our followers this question:
“How many days, on average, does your provider have to process payroll from submission to submitting trial results for review?”
And here are the results:
- 48%: 3-4 business days
- 27%: 1-2 business days
- 25%: 4+ business days
I know, it would be interesting to understand the complexity of payrolls, the headcount of payrolls, the provider model (local, regional, global), and any other context. But let’s accept it: this is a benchmark for you to explore. First, let’s explore it together!
The Art of the Payroll Calendar
I have said it before, but designing and implementing a global payroll calendar is a true work of art. You need to balance the needs of so many stakeholders, and among those is the payroll processing SLA (Service Level Agreement) with your payroll provider(s).
This part of the calendar is crucial. It sits between the submission (that’s after the cut-off) and the payroll approval. The steps after payroll approval are getting employees paid and the GL files into the Finance ERPs. So when processing days move, the rest moves… but those can’t move!
Welcome to the life of a (Global) Payroll Manager.
In my (very recent) days as a Global Payroll Manager, I wrote the SLA myself (Can you believe it?) as my provider at the time didn’t have a processing SLA (Can you believe it?)
Well, you better believe it!
And I didn’t use GenAI to produce it either - Come on, we can write an SLA ourselves. These are the main topics I included:
- Grouping of countries: The landscape included payrolls of 1 employee and payrolls of, e.g. 750+ employees, which naturally called for different SLAs. This ended up separating countries into small, medium, and large.
- Countries in groups: The next step was to map countries to those different groups and group them by region. This helped identify time zones so we could set SLAs on days AND times per a central regional timezone. If, when countries moved up and down in headcount, this meant they also moved to a new group, the processing SLA would change too.
- Processing SLAs: The actual days the provider could take to provide trial payroll results for approval by my team. This spanned from 2 days for Small, 3 days for Medium, and 4 days for Large.
- First Time Right: The FTR percentage was another main driver. It’s great to get payroll results within a certain processing SLA, but if those results are rejected all the time… ouch. That gets you into some serious trouble. So, I would argue that FTR should at least be 90% or above. Go and fight that battle with your customer or provider.
I hope this got you thinking:
How have I agreed on my processing SLAs with the provider?
As I was writing the above, I also wondered what I could have done differently. I think I was too kind to allow 4 processing days for Large payrolls as, in fact, we should have applied more automation to force a 3-day processing window. But hey, that is, as we say in Dutch, “de koe in de kont kijken”. Feel free to Google Translate that!
So, what’s next?
I encourage everyone to reevaluate their processing SLAs: as a provider and as a customer. It takes two to tango, remember? The levers I recommend pulling:
- Work on the pre-payroll processes. How can you get better and cleaner data? How can you automate what matters? Do a Pareto analysis; focus on what turns the needled. Engage with stakeholders, and convince them of the importance.
- Automate your payroll controls to return trial payroll results into spending < 2 hours of checking, reviewing, and reconciling. Remember, that is the benchmark.
- Push your provider to automate their processes and tech stack: remove non-value-added processes and automate the local payroll engine and their own controls before asking for approval.
- As a provider: I guarantee you that offering compelling FTR performance and short processing SLAs will set you apart from the competition. Invest in rigorous improvement programs here.
You are up next. Let me know how you get along with this, my friends!
