Skip to main content

You, yes you! ABSOLUTELY have to forecast revenue and labour BEFORE you start to write your roster.. Yep I said it; have to! You MUST know your key numbers if you are in the Restaurant business, and few numbers are more key than these!

Labour is your number one manageable cost and the one that kicks you in your bottom (line) the hardest when you are not looking. 

Next, you have to let go of something. You have to let go of your obsession of rostering Monday to Sunday. GO on, loosen your fingers, you can do it…

Here’s the why. Bare with me.

Understand this; each and every week almost every business loses money, your hard earned, from Monday to Thursday, on the HOPE that Friday and Saturday will dig you out of a hole. Now that’s craziness!! You are holding your breath for 4 days hoping there will be enough air to keep you going at the end of the week!

BUT, what if… (brace yourself for some crazy), we rostered Friday to Thursday instead of the traditional Monday to Sunday?? There, I said it. Out load!!

But why the hell would we do that I hear you say??

 By starting your week with your key revenue assumptions up front, you have the balance of the week to adjust your labour spend, a whole 5 days, if you DO NOT meet those assumptions.

So if, for example, Friday and Saturday tanks and we are $3k short on Sunday morning, and we have over allocated a labour spend by $750, we can revise the balance of the Sunday to Thursday to bring it back in line.

If we assume an average hourly rate of $25/hour, you just have to pull back 30 hours in 5 days (and continue to adjust for further revenue fluctuations) to bring that bonus home baby!!! (or just to keep you on budget).

You can do it, we all can. Every $$ we claw back is a $$ spent on keeping the doors open, on lifting up the industry, and on giving you more return for all the blood, sweat and tears.

To look deeper into the how, take a walk through the following; firstly as a brief super hack, and hack that will take you closer then you are now, and broadly give you what you need with minimal time, and secondly how to do it in more detail. There is a law of diminishing returns when analyzing data, and it’s up to you to determine what is the best balance for your operations.

The Super-Hack.

By looking at each week’s payroll we can gain enough information to be useful to apply to forecasting your labour costs.

Simply divide your payroll from last week by the number of hours worked. This will give you a general, average hourly rate across the entire week (average your salaried staff at 40 hours).

When you determine your labour target for the week, simply divide it by your average hourly rate, and you now have the number of hours you have to spend.

The Hack

A step further then the Super hack; we look at your payroll and determine the labour cost and the hours that went into it on a day by day basis, accounting for your penalty rates.

When you determine your labour target for the week, simply divide it by your average hourly rate, and you now have the number of hours you have to spend.

The detail

In a restaurant operation Labour is a critical, controllable variable expense, one of the ‘Key 3’ controllable pillars; labour, Cost of Goods (COGs) and pricing.

Rostering occurs in one of two ways; firstly as an operational footprint, in which we simply ‘need’ a certain number of people to manage the assumed demand, (3 staff on the bar on a Saturday night, 8 FoH staff on a Friday), or as a labour percentage of an assumed revenue ($40,000 forecast revenue @ 25% labour cost = $10,000 to spend on the roster).

Writing a roster based on bodies on the floor may work for some, likely smaller, operations, but will simply not cut it if you want to make money.

We are going to work through the preferred means of rostering as a labour percentage of an assumed revenue, as it is significantly better for your bank balance.

I will be using 25% as a base figure, to allow for the additional costs of super, applicable salaried staff entitlements, ( public holidays, leave), workers comp and where applicable payroll tax. Also we have of course removed GST from the assumed revenue figure.

Now a couple of points to focus upon; firstly how do we forecast revenue, and secondly what do we do if the forecast revenue isn’t realized, and finally how do we figure out what is the target in the first place?

 So how do we forecast our revenue?

Traditionally we base the success of our operations (misguidedly in my opinion) upon what we traded on the corresponding week the previous year (adjusted to match the days, not the dates). So, if in the first week of July last year we traded $43,000, that’s a guide to what we should assume this year.

More appropriately, and in brief as it is a topic all on its own, we should arrive at a revenue range that considers at least the following; average weekly trade, what we traded last week, the average of the month previous, and the trend of comparison weeks trade with last year (bare with me; if we trend as a 5% increase on week on week comparisons for this year versus last year, we can factor in a 5% increase in our forecast). This will produce a forecast range, such as $39,000 – $46,500. 

We write the roster based upon the lower range, with the roster designed such that labour increases in line with increased revenue (if we roster for $39,000 but trade at $41,500 we are not under staffed, as the existing rostered staff do a couple more hours). We can achieve this by building a shift that includes key staff for longer and additional staff at fewer hours, that are added-to based upon demand.

SO in our example we have $9,750 to spend on our labour to hit our 25% target, based upon the $39,000.

Now I break the roster into 3 areas; FoH, BoH and Management (salaried staff, not including kitchen); roughly targeting 10% FoH, 10% BoH and 5% management.

This will vary significantly depending upon the nature of your operation, your revenue, opening hours etc, and it gives you good feedback when determining the best employment mix (casual, salaried, Part-time). As an underlying rule it is better to have more variable costs than fixed costs, and this includes labour.

So we then breakdown our weekly revenue into day by day forecasts, and then determine the number of hours we have to spend across each day, with the weekend days incurring penalty rates of course, so increasing the value of each hour correspondingly, and reducing the number of hours available.

There are over 50 rostering software packages available in the marketplace, so find the one that best suits you, and that provides for penalty rates when calculating the cost of the roster.

After determining the number of hours you have to spend based upon hitting that $9,750 target, we then have a labour costs per day as follows;

Monday Revenue $2,350, labour $950 (40%), Tuesday Revenue $2,800, labour $1,000 (36%)….Friday Revenue $9,950, labour $1,900 (19%) etc.

The labour cost on your slower days ($950 in this case) is greatly affected by your fixed labour costs (as we simply spread the salaried staff cost equally across each of the 7 days), and on the operational design of the restaurant. On busy nights I have aimed for 19% and 20% labour, to offset the higher labour costs across the start of the week.

SO we arrive at a point where you have forecast a revenue, and you have written a roster with a targeted labour % in mind.


What do we do if the forecast revenue isn’t realized? When we don’t take the money we thought we would?

 Now the tricky part is what do we do if we don’t meet our forecast revenue targets?? If instead of the forecast $39,000 we take $35,000, and our labour percentage has gone from 25% to 28% (based upon the rostered hours); that is an over-spend of $1,000 in just one week!!

 BUT wait; rosters can be adjusted!! That is, we can adjust the labour cost in line with the revenue fluctuations! 

Sounds feasible. If we don’t meet our revenue target on Monday for example, we cut back somewhere else in the week to maintain our target %, or we send staff home a bit earlier.

Clever you say; well played says I!

BUT wait, what if it’s our peak trading nights that don’t come to the party; instead of the $10k we are expecting on Friday, and $11k on Saturday we fall $3k short by the start of Sunday???

Now that’s not so cool; how can we adjust our labour spend to accommodate such a large short fall, when we only have Sunday to do it?? Sending staff home earlier never seems to cut it!

Well, you can’t, and I am afraid you have just burnt through $750 in cold hard cash. Gonski, straight from your profit! To put that into perspective, if we are a 10% Net Profit business, that $750 is the equivalent of $7,500 in revenue (and let’s be honest the industry IS NOT a 10% NP industry). 

Ouch!! There goes your bonus, off toward the horizon, never to be seen… BUT wait.. What if..we rostered Friday to Thursday?

That leaves how we figure out the target labour % in the first place. Well and here’s the interesting thing; it depends on your Cost of Goods!

It depends. It depends on your business alone.

Ivan Brewer