top of page

How To Forecast Sales For Your New Restaurant

Creating a sales forecast for a restaurant that doesn’t exist yet can be hard as there are many variables and assumptions to consider. But if the sales forecast leads the rationalisation of whether you are likely be profitable and therefore should go ahead with the venture, it’s obviously essential to do this as well as you can. Here's a step-by-step guide I’ve been using to support my restaurant consultancy clients over the years through the sales forecasting process subsequent to them having an outline concept and menu:

Market Research:

Conduct thorough research on the local market. Understand the demographics, competition, and consumer preferences in the area where the restaurant will be located. Make sure you do so for all meal periods and for different days of the week (eg Breakfast, Lunch, Afternoon, Dinner, Late night). Identify competitive benchmarks (competitor set). You will also need to consider seasonal volatility. More on this later…

Target Audience:

Define your target audiences and customer profiles based on your restaurant's concept and location. Try to get a sense for their spending power and types of use cases.

Seek proffesional advice:

As you will discover as you read on, there are many assumptions to make and trying to do this in isolation without any prior experience is only going to lead to those assumptions becoming guesses. It’s important to look at other trading businesses similar to what you are trying to do and bring in some expertise to give some further validation.

Tighten up, revise or create your menu:

Now either check on the positioning and pricing of the menu you have or start to write it at this point taking into account what you have learned in your research.

Covers or Transaction numbers:

Covers are customer numbers. Transactions are the number of payments. The Covers approach is more traditionally used by restaurants and Transactions by fast casual or grab and go style operations.

We are interested here in creating 3 scenarios. Low, medium and high case sales forecasts. We will start by laying down a base assumption (med case) which will be a conservative guess of the number of covers/ transactions for a given meal period on a given day of the week. I encourage my clients to consider this to be in an average month, ie not December, January or an obviously elevated or challenging month for example.

An example of this process would be to create a sales spreadsheet. The first piece of information would be the days of the week put into columns and the rows representing the meal periods you have decided upon. It’s a good idea to create all meal period options available at this point even though you may not consider them relevant. You might change your mind on this later and can easily amend the spreadsheet to accommodate this and evaluate the impact of those sales against the respective costs.

Now plot the estimated medium case covers/ transactions by meal period. Next do the same for the low and high cases. I would suggest 10%+ and - but it’s important to structure the spreadsheet so you can manipulate that percentage later and evaluate the impact to new variables. Now tally up each day to generate your total covers/ transactions by day and total covers/transactions for the week, and for each case (low, medium, high).

Discovering Average Spends/ Transaction Values:

Average spends are the average amount spent by a single customer, typically in a traditional restaurant setting. It is the total of all sales (ex vat) divided by the number of covers for the period in question. The average transaction value is total sales (ex vat) divided by the number of transactions for the period in question.

To forecast these average spends/ transactions means looking closely at the proposed menu. By this point you will have a menu that has been considered for the concept positioning, the location and market research. The menu will usually be broken into sections such as starters, mains, desserts. Or maybe small/ large plates or even a simple two price tasting menu for example. You’ll also need to do this in relation to how you have laid out the sections of your drinks and wine list. For each of these sections you want to look at the low, medium and high price points. You will now need to evaluate through your market research what would be the likely uptake of a given section of the menu as a percentage of the whole. For example you might say that 95% of people will likely order a main course during the dinner meal period but only 50% will order desserts. Same for wine or drinks. Perhaps 50% of people order wine and of those the likely uptake will be 2 glasses or half a bottle. You will also need to make assumptions about the percentage uptake of each of the price points (low, medium, high). For example you might say that more people are likely to order more inexpensive items at lunch compared to dinner, hence the average spend will be higher at dinner. This process is easier for working out average spends as you are considering how an average individual customer will likely order for the menu however it can be applied to average transaction value also but you will have to consider the likelihood of individuals order for other people as is often the case. You may consider, for example, that every second transaction on average will order two main courses, in which case the average percentage uptake of mains would be 150%.

This can be the most painstaking part of sales forecasting as there could be many variables to consider so it’s good to solicit the help and validation based on your competative set if doing this for the first time so that your assumptions can be questioned and further validated, especially with someone who has the experience and trading data of a similar offering. Sometimes when there are too many assumptions it will make sense to simplify or rely on more judgement however the more science and specificity you can go into here, the more it encourages you to really think hard about how your guests will be exploring your offering and indeed whether or not some things on your menu are worth doing at all.

Again, to do this you will need to create a spreadsheet that maps the sections of the menu. The low, medium, high case price points and the percentage uptake of each, again by meal period. You may even consider that it would be useful to plot this out for different days if you consider customers being likely to order in different ways on different days of the week.

Again, plot out all of these assumptions of average spends/ transactions by section of the menu, and for each meal period. Add up all sections amounts for each meal period and you have the total average spend/ transaction by meal period by day of the week (unless you’ve considered variation potential dependent on the day).

Once the covers/ transactions have been completed you can turn your attention to the average spend. Complete the average spend/ transactions assumptions in the same way by plotting them in to the sales spreadsheet by day, meal period using the medium case (base case) row. Next apply a 10% reduction and increase to generate your low and high cases.

Tallying up:

Now most of the hard work is done the next step is to multiple the covers/ transactions by the average spend/ average transaction value to achieve the total sales (net of vat) for each meal period and day of the week. Then tally the sales by meal period and you have the total sales by day. Tally the sales by day and you have the sales for the week for your each of your low, medium and high cases.

A note of caution:

Remember all of this effort has been about getting more deeply into considering where your sales will come from. But it’s all assumption. There are many variables at play here but you have built a tool to go back and adjust things to see what the impact might be. So for example, what if we wanted to be more competitive and lower our main course pricing, what would be the overall effect. Now you can easily do that and see what happens to your overall sales. But we must always remember these are assumptions and that’s why it’s important to look at the low case in particular if we want to be prepared for the evaluation of risk.

Seasonal volatility:

The last section of the forecast is to take your average weekly sales and multiply them by the number of weeks in a given month (averaging 4.33 weeks). We now have an average sale for an average month for our low, medium, high cases. The final step is to again revert to your market research and put forward some assumptions about the variance of sales from that average to the other months of the year. Ie perhaps December is 10% higher than November. January 30% lower than December. Don’t forget to consider numbers of days in each month and potential closures for holidays. Start with what you think will be an average month and then applying those percentage increases/ decreases to each subsequent month.

Once you’ve plotted that out you can tally up each case for each month and you have your total annual sales forecast for the given year on a low, medium and high case.

Remember, sales forecasting for a new restaurant involves a lot of assumptions and educated guesses. It's important to be as realistic as possible and prepare for different scenarios. As the restaurant starts operating, real data will come in, and you'll be able to refine your forecast for greater accuracy.

Of course the next phase in deciding whether to go ahead with you restaurant venture is to work out the costs attributed to your sales forecast cases. That’s a post for another day but I will leave you by saying that if your sales assumptions are done with this level of consideration and your respective costs are too high so that your business doesn’t stack up in the low case profitably, it’s time to put your energy elsewhere as the risk is too high and in doing so you might just have saved yourself years of unnecessary pain and loss.

I run 1-1 and group mentoring sessions on all financial forecasting for restaurants so feel free to get in touch if you are looking for some support and/ or validation. Good luck!


bottom of page