We recently announced our new pricing and plans, but how did we get those numbers? In this post we’ll break down MeetSpace’s scientific approach to pricing by looking at how to do a perceived value study and how to generate a Van Westendorp price sensitivity chart from the data using Google Sheets.
Price your product ... with science!Tweet
But first, a huge shout out to Price Intelligently. Their Blog was an excellent source of information and their team was also happy to answer our questions about the process. I highly recommend reading through their material and their email course.
Let’s start with Dutch economist Peter van Westendorp and his Price Sensitivity Meter. From Wikipedia we have this amazing chart that looks like it’s chock full of data (because it is):
There’s a lot going on here! We have six different lines with four intersections and also some kind of “Acceptable Price Range”. Let’s break it down.
Red Line (too cheap): This is the percentage of people surveyed that thought that the price was so cheap that they would question the quality of it. It starts at 100% at $0 (but not always, just for this data!) and descends across the chart and bottoms out. This line tells us how many people would probably not purchase the product because it was just too sketchy! Would you feed your family a $1 large pepperoni pizza? For this line, the lower the percentage of people the better, because that means less people are sketched out.
Purple Line (too expensive): This is the percentage of people surveyed that thought the price was so expensive they wouldn’t even consider the product, no matter how awesome it was. They just can’t afford it. This is the Lamborghini line. Yeah, I’d like a Lamborghini but I would not even consider it when shopping for my next car. For this line as well, we want it to be as low as possible.
Red x Purple: This intersection is where the maximum number of people think that the price is not too cheap and not too expensive. This is the simplest answer on the chart: the most number of people would buy your product here.
Now, we could get a little fancy and multiply the percent versus the price: 80% of people would pay $700, but if you move right along the graph to the $1000 mark, you can see about 75% of people would pay $1000. 0.80 * $700 = $560 paid for each user who would buy your product (but their decision depends on price). 0.75 * $1000 = $750 per person who would buy your product. Shifting to the right actually got us more money!
However, this is not the only factor. Perhaps by having just 5% more people, your virality amplifies that enough that you end up with a bigger market, and hence a higher total number of people considering your product. Or, consider if your product was sold entirely via sales, selling to less people at a higher price point may make you more money (because you’re looking at Lifetime Value vs Cost to Acquire a Customer right? LTV vs CAC?). It all depends on you, but we’re going to look at the intersection of Red x Purple.
Orange x Blue: Orange is percent who thought your product was cheap (or “a good value for the money”) and blue is expensive (or “getting expensive but I would still consider it”). This is the “indifference price point” because at this point the most people aren’t bothered by the price (but they would all consider it). So this is nice to know because the people between this point and the Red x Purple point are the ones who are wringing their hands trying to figure out if it’s worth it.
There’s more to dig into with this chart, but those are the basics and that’s plenty for us to use to pick our pricing.
OK, so that’s a great chart, but how do we get that data? The key is to conduct a Value Study. It’s really a Price Sensitivity Survey, but “price” and “survey” are dirty words that turn people off from doing surveys. But “value” and “study” are both cooler sounding and make you feel more important and considered when you’re filling it out.
The very first thing in your survey should be a description of your product and its features. Keep this as short as possible. We said:
We’re building a new team-based video conferencing product and we’d love your help determining its value. This should take you about two minutes. Thanks!
Please consider the following video conferencing product:
- Fast and smooth video chat that doesn’t overheat your computer
- HD audio with very low delay
- HD 1080p screensharing optimized for code, documents, and presentations
- Permanent accounts for your team
- Permanent meeting rooms with dedicated URLs
- Guest access control
- End-to-end encryption of all audio and video data
And that’s it! Remember: surveys that take more than 2-3 minutes see a harsh dropoff in respondants.
To conduct a Van Westendorp Price Sensitivity Survey (cough I mean value study) you only need to ask four questions:
- At what price would you begin to think the product is too expensive to even consider?
- At what price would you begin to think the product is getting expensive, but you still might consider it?
- At what price would you begin to think the product is so inexpensive that you would question the quality and not consider it?
- At what price would you think the product is a great deal for the money?
In addition, MeetSpace asked two more questions:
- How many people are in your team’s largest online meeting?
- How likely would you be to sign up for this product? (1-7 scale)
The first is a demographics question for MeetSpace, because we were going to charge per user per month. So we knew our plans would depend on how many people were on your team. This would let us know how the price would change based on team size. But it turned out it didn’t, so we won’t talk about it. But it was a good thing to know!
Likelihood to buy is a common addition to this survey for two reasons:
- If your likelihood to buy is low (below a 5 average) you may want to re-evaluate the product you’ve created. Perhaps conduct a Feature Tradeoff Analysis
- If you end up with a lot of people who put a 0, you may want to eliminate their responses to clean up your data
We won’t be using either for determining our target price in this article, but you may want to ask them at this time.
We’ve created a new Value Study in Google Forms to use in the rest of this post. It considers the value of a bagel shop. Check it out, and fill it out because we’re going to use live data for the rest of this post:
Take the bagel survey and learn how to do value-based pricingTweet
Mmmm, now I really want a bagel!
OK, it’s time to crunch some numbers. Here’s the live data from the form (I changed the labels to be easier to read):
Now, we’re going to need to use a few functions to get Van Westendorp data out of the raw data. We’re going to make another sheet called “analysis” and it looks like this:
Here’s a breakdown of each column:
- Buckets: manually fill this out. Each number is a bucket that we’ll use on the Van Westendorp x axis. I’ve chosen $0.50 increments from $0 to $20.
- Too Exp Frequency: First cell is the formula:
FREQUENCY(responses!B:B, A:A). This says “count how many things in column B of responses (Too Expensive) fit into each Bucket”.
- Too Exp Cumulative: First cell is the formula:
B2/COUNTA(responses!B:B)and the second cell is
B3/COUNTA(responses!B:B)+C2. Then, drag the second cell down to the last bucket to propagate the sum. This is saying that each cell is the percent of total people that thought that this price was too expensive. Its the people that thought this price was exactly too expensive, plus all the people that thought a cheaper price was too expensive.
- Exp Frequency: same as Too Exp Frequency:
- Exp Cumulative: same as Too Exp Cumulative:
- Too Cheap Frequency: same as Too Exp Frequency:
- Too Cheap Cumulative: different! now that we’re talking about cheapness, we have to flip the formula:
F2/COUNTA(responses!D:D)+G3. Now we say “everyone that thinks this price is too cheap plus everyone that thought a more expensive price was too cheap”.
- Deal Frequency: same as Too Cheap Frequency:
- Deal Cumulative: same as Too Cheap Cumulative:
At this point, we can use the column A as a label (price) then use the columns C, E, G, and I as our data. We’ll make a simple line chart. For the data source, we put:
analysis!A:A, analysis!C:C, analysis!E:E, analysis!G:G, analysis!I:I. Then we do the following (in advanced edit):
- Chart Types: Check “Use row 1 as headers”
- Chart Types: Check “Use column A as labels”
- Customization: Title and label axes and add legend
- Customization: Horizontal axis min 0 max 10 with 10 major lines ($1 each)
- Customization: Vertical axis min 0 max 1 with 10 major lines (10% each)
And that’s it! Here’s the live chart from the data from the survey so far:
At the time that I wrote this, I only had seven data points, but already there was an intersection at $5.50! By the time you read this, the data could be fairly different, but think about what you would decide to charge for the combo.
Now that you know how to read a chart like this, here’s MeetSpace’s own chart:
I bet you can guess what price we picked!
Fill out the form, send this post to friends, and lets see how much people think bagels are worth! Also, if the “big city” question turns out to be interesting, I’ll update this post with results.
I just learned how to do value-based pricing withTweet
Van Westendorp sensitivity analysis!
Have questions? Get in touch on Twitter with @meetspaceapp.