# College Students & \$\$ Per Household

#### TexasRookie

##### New member
I am going to be opening a Pick-up/Delivery only location just steps from a large University in a town of about 18,300 households.

I have read in other threads that the National Restaurant Association statistics put the average dollars per household at \$17.50-17.85 per month for pizza.

Based on those numbers, it would suggest that \$320,250-326,655 is being spent in the community per month on pizza. I assume that the student population which is over 25K would skew that number upwards. I am trying to estimate numbers and did not know how to factor in the College Student Population.

Do any of you that operate in a College town have a guess as to how the students would change the \$ per household average?

For those of you who operate in a College town, what % drop do you see when school is out?

Last edited:
I’ve long ago lost the study/numbers, but when I look at a college market I add in a \$100/year above and beyond for each college student’s spending on pizza. Then grab the phonebook and make a list of all the independents and chain locations. Using the averages from the annual PMQ survey, I then calculate what the town supports according to the number of pizza shops.

This gives me a range which I assume the total market falls within. Does it really? shrugs

My campus store dips about 40% during the summer. Build a solid foundation on the locals, THEN go hard after the students so you can survive when they are gone. And remember, they are two different markets. You can be different things to the different groups (residents love our gourmet pizza, the students love our package deal).

Last edited:
I’ve long ago lost the study/numbers, but when I look at a college market I add in a \$100/year above and beyond for each college student’s spending on pizza. Then grab the phonebook and make a list of all the independents and chain locations. Using the averages from the annual PMQ survey, I then calculate what the town supports according to the number of pizza shops.

This gives me a range which I assume the total market falls within. Does it really? shrugs

My campus store dips about 40% during the summer. Build a solid foundation on the locals, THEN go hard after the students so you can survive when they are gone. And remember, they are two different markets. You can be different things to the different groups (residents love our gourmet pizza, the students love our package deal).
Perfectly stated. Couldn’t have said it better
CaptSammy

Last edited:
TexasRookie:
I have read in other threads that the National Restaurant Association statistics put the average dollars per household at \$17.50-17.85 per month for pizza.
This number gets thrown around way too much on this board. Average of the entire U.S.? That’s a sample diverse enough to make any statistical analysis like this worthless.

There’s nothing wrong with using this number for some rough ideas, but you might as well just make something up. If you’re going to use an average for actual business planning purposes I would highly recommend drilling down to your community level and doing some research of your own.

Thirty phone calls and you’ll have accurate data for your own area. Do the same with the college kids.

Last edited:
Piper:
This number gets thrown around way too much on this board. Average of the entire U.S.? That’s a sample diverse enough to make any statistical analysis like this worthless.

There’s nothing wrong with using this number for some rough ideas, but you might as well just make something up. If you’re going to use an average for actual business planning purposes I would highly recommend drilling down to your community level and doing some research of your own.

Thirty phone calls and you’ll have accurate data for your own area. Do the same with the college kids.
I agree that we can’t give too much credit to that average number. It’s value to me is for very preliminary rough estimations. Beyond that, I agree that it can be over generalized and cause problems when used in a specific marketplace. Something to remember is that the more members of the population involved in a data collection, the more likely it is to represent a small piece of that population.

Wow. Thirty calls will give me accurate data? No chance. Certainly not more accurate than the national average survey data you challenged to start with. With the town I am in, the sampling error using 30 calls would be outrageous. What if I end up ‘randomly’ calling 15 homes that cannot afford to eat out right now? Unlikely, but it can happen.

Last edited:
NicksPizza:
Wow. Thirty calls will give me accurate data? No chance.
Trained statisticians would disagree with you. A sample size of 30 is where the Central Limit Theorem applies; sample means are normally distributed for sample sizes of n>=30.

It’s a pretty basic tenet of descriptive statistics. Here’s some reading: http://people.hofstra.edu/Stefan_Waner/ … distr.html

So yes, you could get 15 straight people that can’t afford to eat out right now. The Central Limit Theorem indicates that the odds of that happening by random chance is infinitesimal if your sample size is greater than 30. Those results would much more imply that it might not be the best area to open a pizza place in.

Last edited:
Big Picture for amateur statistics is that yopu have to find someone who is estimating and figure out if you trust their math. If you don’t have an advanced math background, then you have to look at reputations or the people doing the studies. Then, you take a deep breath and a big fat leap of faith.

We can find another thread for us math geeks to trot out the Central Limit Theorem, standard deviations, t-tests, assumptions of normal distribution, and the fallacies of large and small numbers. I need to talk pizza and food theory today.

Last edited:
Piper:
NicksPizza:
Wow. Thirty calls will give me accurate data? No chance.
Trained statisticians would disagree with you. A sample size of 30 is where the Central Limit Theorem applies; sample means are normally distributed for sample sizes of n>=30.

It’s a pretty basic tenet of descriptive statistics. Here’s some reading: http://people.hofstra.edu/Stefan_Waner/ … distr.html

So yes, you could get 15 straight people that can’t afford to eat out right now. The Central Limit Theorem indicates that the odds of that happening by random chance is infinitesimal if your sample size is greater than 30. Those results would much more imply that it might not be the best area to open a pizza place in.
As you say the theory works on the basis of ‘random chance’ which is fine in the world of maths but in the real world we don’t have random chance so this kind of theory doesn’t work. Bearing in mind we talking about marketing to two distinct types of customers ‘students’ and ‘non students’ theres your random chance gone in a big ball of smoke.

So back to reality maybe if you find 30 pizza customers (but thats gonna take more than 30 calls) then you may be able to calculate the ‘average spend’ but as its been highlighted that there are two very distinct groups here you’d probably want to identify which group they fit in and try and get an average for each group - now thats waaaaaay more than 30 calls!!

Last edited:
Wizzle Wassell:
As you say the theory works on the basis of ‘random chance’ which is fine in the world of maths but in the real world we don’t have random chance so this kind of theory doesn’t work.
No, that patently incorrect and it doesn’t “work on the basis of random chance.” But, I’m not going to argue about it with either of you in this venue. I’m currently working on my Master’s Degree in statistics. I think I have a grasp on how to conduct a survey.
NicksPizza:
I need to talk pizza and food theory today.
Nobody was stopping you from talking about “food theory”, I don’t think. I guess I shouldn’t have responded when you disagreed with me by posting your opinion as a fact:
NicksPizza:
Wow. Thirty calls will give me accurate data? No chance.
I took that as you wanting to talk about statistics. I suppose once you post your opinion there is no need for anybody else to say anything. I shall refrain from this in the future and ensure your posts are the final say on all topics.

Last edited:
Last edited:
Piper:
Wizzle Wassell:
As you say the theory works on the basis of ‘random chance’ which is fine in the world of maths but in the real world we don’t have random chance so this kind of theory doesn’t work.
No, that patently incorrect and it doesn’t “work on the basis of random chance.” But, I’m not going to argue about it with either of you in this venue. I’m currently working on my Master’s Degree in statistics. I think I have a grasp on how to conduct a survey.

I took that as you wanting to talk about statistics. I suppose once you post your opinion there is no need for anybody else to say anything. I shall refrain from this in the future and ensure your posts are the final say on all topics.
You obviously understand statistics much better than most of us, but:

a) you don’t seem to understand the concept of discussion (you clearly, however, understand the concept of toys out of the pram!). If you make a claim you believe in then back it up don;t fold at the first challange.
b) if 30 was the idea sample the I suggest you set up in business do surveys as you’d save many companies millions of dollars calling millions of people each year - now they only need to call 30 per survey!! but you an I both know that is not the case
c) you may have a grasp how to run a survey but you are being challenged as to how that works in the real world!

enjoy!

Last edited:
TexasRookie . . . don’t mind us. We are easily distracted from thread topics. At least I am. Back to your original inquiries, The college may have a Department of Student Affiars or Student Activities that may have some insights into the numbers of pizzas they use for their events and maybe fraternaty expenses. Not exactly the best single-stop for data, but it could be useful for the business plan identifying customer sources and revenue streams.

I think it is Paul who operates in Ganesville, FL and can give you some of his experiences with the University. It may be like some of the resort towns in that they have an ‘artificial’ population come in for several months out of the year and have to plan for the lean months. I don’t have the data, but if he doesn’t find this thread I’ll send him a private message asking he get in touch with you if he has time.

Last edited:
TexasRookie:
For those of you who operate in a College town, what % drop do you see when school is out?
Planning a business based on averages (national averages in a local area) is only going to give you a very very rough idea and one I’d personally not pin much to myself.

One issue is how well you market to the college/student community.

In my nearest major college town I’ve an acquaintance who has a shop a guy who does twice as much when the students are present (or half as much when their not - depends if your an optimist on not!). A friend of mine a few hundred miles away has a similar sized college and a similar pizza shop but for what every reason can’t crack the student market. In the summer they both appear to do similar business. Either way if you can crack the student market I believe they tend to spend a little more than the typical customer.

Last edited:
If I take an guess at the sales of the other pizza shops here, I come up with pretty close to \$18 per month per address. The stores that focus mostly on students drop significantly in the summer, while the family oriented stores stay reasonably steady. I drop about 15% from my average week in the summertime and 30% from my best weeks. Our campus store probably drops 30% from its average week and the biggest weeks this fall will be more than double what some of the summer weeks were. I’ll take the consistency of sales throughout the year over the huge spikes that student oriented stores experience. Just much much easier to operate.

Last edited: