In my last post I outlined a financial model for opening a wine store.
In this post I’m going to do some local market research. This will allow me to estimate the most important and sensitive elements of opening this particular wine store. It’s an easy and cheap, but statistically valid, way of doing research.
In the last post we identified these as the most sensitive financial factors:
- Store Penetration in Y3 and Y2
- Image Seeker Segment Size, Market Share, and Volume purchased
- Enthusiast Segment Size and Market Share
- Luxury Enthusiast Segment Size
- Conversion rate from internet marketing campaigns
See the previous post(s) for a description of these factors.
An easy and cheap way to conduct valid research
… without spending $100,000 with a research company.
I’ll be using a mailbox drop with a letter (no envelope) from me asking people to go online to complete a survey.
Yep, it’s almost that simple.
Except you’re probably not that confident with market research so here’s how to do it.
I want to find out how many bottles they buy, from where, and what their likely price range is.
Let’s say we ask these questions:
1. In the LAST MONTH where and how many bottles of wine have you purchased?
Please enter bottle numbers in an many options as apply below.
- Store A
- Store B
- Store C
- Internet Wine Store
- Vineyard Direct
- Liquor Chain A (outside of town)
2. Would you basically describe your views on wine as:
Please select one option below
- There are so many wines on the shelves – it’s confusing since I can’t always tell from the label how the wine is going to taste. I like to drink wine, but I don’t know what kind to buy.
- I have some knowledge about wine and am actively seeking to further my wine knowledge. I look to wine writers and other experts for guidance. Buying a wine my friends don’t like is very disappointing.
- I enjoy wines from established wineries. I feel like they have perfected the art of winemaking. I like to be offered a wide variety of well-known wine brands.
- I shop in a variety of stores each week to find the best deals so I like to be offered a variety of wine specials and discounts. I am a heavy user of coupons and rebates and know what is on sale even before I walk into the store.
- I don’t know much about wine, I just know what I like to drink. I usually buy the same brand, preferably a domestic wine. To me, wine is an everyday beverage, so I don’t think much about it.
- I love to spend time browsing in the wine section. I’m a confident wine buyer. I like to refer to wine publications and I’m influenced by wine ratings and reviews. I appreciate and understand sophisticated wine information.
3. How much do you spend on a bottle of wine for:
- After work, everyday wine
- Weekend BBQ with friends
- Formal Dinner at your place with some good friends
- A special occasion – your loved one has a big birthday
I’ll give them these price point options in columns along the top:
n/a, Less than $4, $5-10 $10-15 $15+
Actually I’ll be asking a few more questions than that and am happy to share them but I wanted to simplify this post ;).
Once I have enough results I’ll redo my financial model estimates (see in the previous post).
If the survey suggests there are:
- a large group of “Image Seekers”
- who are buying lots of wine from a Liquor Chain Store,
- at high prices
… then I think a premium wine store will work.
If they are:
- mainly “Satisfied Sippers”
- who are buying little wine
- at cheap prices
… then the store is doomed.
No doubt it will be somewhere in between.
I’m not going to cover Internet marketing conversion rates with this survey (the last factor in the list at the top), I have another of way of assessing that which I’ll share in another post.
The Actual Survey Software
Sign up for a free account at Survey Gizmo. They provide a super easy way to create a survey on the internet (not your computer), they publish it via their servers, and provide reports you can use.
I’ve created a video of how easy it is below:
Music: Phil Collins, Papa was a Rolling Stone
You get up to 250 responses when using a free account. Which is about how many I need for a statistically valid sample.
Which reminds me to go through the math part.
The Math – arrgh!
Now I hate math and statistics as much as you. But there are some great tools and methods to keep this simple.
First let’s make sure we get a representative sample of the population.
Use this online calculator from Custom Insight. The best way to explain this is by using an example.
I want to see how many people have wine cellars in the local area. This is an indication that they are a high price/ high volume wine drinker.
If the 50% of the population has a wine cellar then I’m ecstatic. If 5% has a wine cellar then I may be in trouble. In fact I don’t care if 20% or 26% of the population has a wine cellar, only that it is about that figure. I’m after big differences here so I’m willing to tolerate rough figures:
- My tolerance for error is high and I’ll choose 6% in the field: How much error are you willing to tolerate?
- How many people are in your population?: 25000 as per my previous post.
- Confidence refers to how likely the figure is going to be correct. I’m happy with a 90% Confidence level.
- Clicking Calculate tells me I need 188 respondents.
So if I did this survey 100 times, then 90 of those times the survey would show results that are within a 6% range (between say 20 and 26% as in the example above). However 10 of those times the figures would be outside of that range. Which is pretty loose but I’m a businessman not an academic, I need the results fast and cheaply.
How many surveys to send out?
I believe that I’ll get a 20% response rate because I’m a local person asking sweetly 🙂 , not some faceless corporate. To increase response I could offer a prize but I don’t feel I need to, yet. So for the next field in the calculator I’ll put in 20%.
What is your estimated response rate?: 20%
The calculator tells me I need to send the survey to 940 people.
Making sure the sample is not biased
I could just get a student to deliver 940 surveys to the first 940 letter boxes he comes across. The problem is the sample will be biased towards those streets. They may be in a poor part of town, or a rich part. They may be closer to the shop, have lots of vacant houses… all sorts of things could skew the sample.
So here’s what I’ll do.
Let’s say I know there are 7000 households: 7000 households divided by 940 delivered surveys = 7.
So I’ll ask him and his friends to go to every street and post a survey request in every 7th letter box. At no time are they to decide whether the house is worthwhile or not as that will ruin the randomness of the sample.
The Survey Request
I’m just going to print out or photocopy the following:
Hello, my name is [your name], and I’m researching whether I should open a fine wine store in [town].
It would be great if I could, but I just don’t have enough information to make that decision so I thought I’d ask you :).
Your answers will be immensely helpful to me in either opening a great wine store that suits you or … helping me avoid making a financial mistake!
I’ll ask a few questions about your preferred wines, your recent purchases and what factors influenced these. It should only take about 5 minutes.
It is an online survey, so please use your internet browser to go here:
[website address from survey gizmo]
Your answers will be anonymous, and are for my own use to create the best store I possibly can.
Please only complete this survey if you are 21 years or older.
[your full name]
[full contact details]
The Online Survey is here: [website address from survey gizmo]
Sometimes everything just goes wrong, so its best to have a trial run yourself first.
Print out 100, deliver yourself and see what happens. You may need to adjust the questions or offer an incentive. By finding out with a small group first you check your process.
You can monitor your surveys day by day as well as run reports and download them.
So that’s how I’m trying to estimate the most important and sensitive elements of opening this particular wine store.
Now, given I’m 20 pounds overweight, can I talk myself into getting on my mountain bike and delivering many of these myself… or I could just stay home and open a bottle of wine while someone else does it. Hmmmn.
What’s your thoughts (not the overweight bit)?