Sales Projection Spreadsheet Accuracy

OK, a little background. Like most new posters, I’ve been lurking this forum for a long time, reading and absorbing everything I can. While I realize there is some art to sales projections, I’m wondering if there can be more science behind it too.

So far, I’ve analyzed 90+ different locations in my area. I’ve done this two ways. First, I go to this site and get the demographic info for 1, 2, 3, and 5 mile radius.

Then, I go to this site, and draw a radius for that address for 1, 2, 3, and 5 mile. I start with 1 mile, then change the radius size to 2 miles and add, and so on and so forth. Once that is complete, I download the KML file, and open that file with Google Earth. I then type “pizza” into the search box and count how many DELCO competitors there are.

I input these two data sets into my spreadsheet, number of households, and number of pizza places into two different pages of the same spreadsheet. The first one takes the average of the four areas and ranks the location against all others. The second page takes the data from the first page and assumes you’ll get 100% fair market share of the 1 mile radius, subtracts the homes from 1 mile to assume you’ll get 50% of fair market share of ‘new’ households in the 2 mile radius, subtracts those from the 3 mile radius and assumes you’ll get 33% fair market share of the ‘new’ households in the 3 mile radius and 20% in the 5 mile.

I’ll admit those percentages are a little arbitrary, and realize that in different areas of the country, driving 5 miles for pizza is more acceptable than others.

My hope is that some of you will download my spreadsheet, input the data for your store, and reply with how accurate it was, in terms of a percentage as I could understand why people would be guarded about their specific sales numbers. Also, I’m hoping that will keep people a little more honest with regards to the spreadsheet results. I’ve run the numbers against a few known stores, but those were from years ago and the demographics could’ve changed significantly since the sales numbers we’ve obtained. The spreadsheet, specifically the V2 page, only slightly underestimated the sales of that particular store.

If this spreadsheet works well, it could save a lot of newbies a lot of time trying to guesstimate sales at a particular location.

Unfortunately, I can’t upload the spreadsheet, so if you PM me, I’ll send you a copy, if you’re interested.

I do not see distance from location parsed to miles as having much to do with what percentage of customers you would likely get. Also, dividing the market just a among pizza stores is simplistic. You compete against every food option out there not just pizza.

Sales would initially have more to do with visibility to traffic and $$ spent on marketing than on the measures you are proposing. Demographic studies can be useful for potential… not for forecasts.

I would suggest that you spend some time sitting outside competitors counting delivery vehicle trips. Make some assumptions on number of deliveries per trip and average ticket and you will get a better picture of what each competitor is doing. Look at what marketing they are doing, see what you think of the product.

I agree that mileage isn’t the most effective way to measure reach in terms of customers. In some locales, driving 5 miles for a pizza wouldn’t be anything to blink at, in my area, and many suburban areas, I couldn’t fathom driving five miles for a pizza.

I disagree slightly that you’re competing with every food joint out there, at least for terms of this, since I’m strictly counting money spent on pizza. If I wanted to create more work for myself, and include every restaurant and include every dollar spent on eating out, I certainly could, I just don’t find it helpful. Not many people sit around and say, “Wanna go out to eat? Sure, do you want pizza delivery or a 5 star steakhouse.”

I’m not necessarily trying to get a model that is 100% accurate, as that isn’t even remotely possible, however, it would be nice if there were a more concrete way for people to measure the potential of their stores (I do agree with that point). It was surprisingly accurate for the stores I have the numbers on, and I’m just curious if it rings true on a larger scale.

Or the TT can continue telling every newbie that comes along that they haven’t given enough information to help. This, if we can get enough data to figure out how accurate it is, could actually help people.

If accurate forecasting was a simple as finding a formula based on population, proximity and competition no-one would ever fail.

Other variables are far more important. It is not that newbies do not provide enough information; the challenge is that no amount of this kind of information is enough to answer whether that newbie is going to have good pizza (they ALL think they do), going to understand customer service, going to have a good location both for operations/capacity as well as visibility and customer access, going to produce and execute a competent marketing plan or even whether they have the financial stamina and the stomach to employ it to make it past the first year.

Running up a scenario based on population, average income, typical pizza spending etc etc has as much potential to create unjustified confidence as it does to give a good answer or kill an idea actually would have worked.

Not if someone understands statistics… With enough data, the results could be relatively predictable. The problem is your idea of predictable and mine are probably a little bit different. It’s no big deal, really, just trying to do an experiment and maybe help people.

You said that this has been fairly accurate for the stores you have figures for, so you must have done some calculations. Where did you get the total “market share” figure for your calculations?

“Not if someone understands statistics… With enough data, the results could be relatively predictable. The problem is your idea of predictable and mine are probably a little bit different.”

Nonsense. The kind of statistics you are referring to are not rocket science and I daresay you have no knowledge of my abilities in that regard. I have been using statistics in business applications for going on three decades and even incorporate it when I teach as a guest lecturer in the local college business courses.

If you are trying to forecast sales for a national chain with a known product, experienced management, reliable execution and an established marketing track record combined with professional store design and location selection I would agree with you. My point is that, for new indy operations, these variables have far greater impact than the ones your statistics address… therefor trying to create this model for indy shops is a good tool for comparing the potential of different markets and locations but not for forecasting actual sales of those shops.

A tool like that is a good discipline for comparing the merits of different locations a new business might be considering. I would be very reluctant to propose that the results produced would be a forecast that should be relied upon.

Just some examples of variables that could throw the numbers way off from a statistical model:

  1. Good visibility/access vs poor visibility/access: This alone could move sales by 20%.

  2. Great marketing and high investment vs mediocre marketing and low investment: This variable could change sales by more than 100% and makes all this discussion moot.

  3. Competent store design and build-out: How many new stores limit the peak sales they can do because they fail to understand the peak volume they need to achieve the average volume they seek.

  4. Reliable execution and consistent product quality combined with a marketable menu/price program: Every new business owner thinks that they have exceptional product. Most don’t. By definition “exceptional” means everyone does not have it. Beyond that, many new businesses do not immediately grasp what it takes to deliver the same experience over and over again.

National chains are good at all these things which is why they can use demographic statistics effectively to predict results. They know that if they lease a space with certain specific traits, deliver the product and service per standards and spend “X” dollars on marketing the demographics will take care of the rest. On the other hand we all know of indy operators that kill it in places that should not work due to extraordinary product and hard earned customer loyalty and we have all seen indy stores that failed despite seemingly good market demographic prospects.

Use demographic info and incorporate the presence of competitors by all means to compare the merits of locations but do not be enticed to believe the numbers they produce represent forecasts.

A great indy store can actually increase the market for pizza; Exceptional product and customer service has a gravitational force.

Also, at some point having a lot of competition is a good thing. It means that there is a lot of business available in the category… then you are just in a bear race. (I do not have to be faster than the bear… I just have to be faster than you) I would rather open in a location that has 6-8 stores doing reasonably well than in a location that has only one competitor regardless of how they are doing.

  1. I would listen carefully to those on this site. If you listen to others who have done the “trial and error” it will cut down on your frustration.

  2. I love projections and forecasting - because there is no wrong answer!! But do not fall in LOVE with them - they are only designed to give you a benchmark.

  3. Demographic statistics only look good in government reports. It has little to do with how successful a business (in this case a pizza business ) will do. I know a guy that has been at the same, not so great, location. But it’s great for that neighborhood. And after 15 years in the same location the sales have increase from 2008-2012 an average of 12% a year. Not because the demographics of the neighbor have increased, but because of what bodegahwy says - “reliable execution and consistent product”

  4. One final thought is that stats look good on paper or on your screen. You can make them dance and flash for you. But I have yet to see a stat for how many unhappy customers you made because you think waiting 15 min for you lunch is acceptable or that your numbers say you only 2 delivery drivers.

I agree with bodegahwy maybe some actual legwork is needed. Go check out others in the area and actually see how they are doing.

wow these responses are spot on…stats, business plans, etc are just a piece of the puzzle but with out product & execution they are meaningless