How I successfully tested and rejected two niche markets using online surveys

2012 April 13
by Emmanuel Goossaert

I recently got some interest for micro-ISV projects (micro independent software vendor). These are small products/startups that require minimal time and financial investment, and generate small but steady revenue. The idea is that I would not build it myself, I would outsource it so that I can focus on other things than code. The time investment I am aiming at is around 4-6 hours per week. I am now researching niche markets in an attempt to create my first micro-ISV.

The paths followed by successful micro-ISVs have similarities, as they all include testing the market before doing anything else. Testing can take various forms, it can be a sales website, a prototype, or even a shadow product (faking the product and doing the computer’s work by hand before spending time actually coding the product). I recently had a couple of ideas for micro-ISVs and I wanted to test them before building anything, instead of wasting time coding something nobody wants. So here is at the story of how I tested and rejected two niche markets using online surveys. Please note that I am not explaining how I designed the surveys, but just what was the process of using surveys as a way to test niche markets. I will cover survey design in a future article.

The first market: CSAs

I have friends in France who are part of CSAs or Community-Supported Agriculture organizations (“AMAP” in French). So basically, there are buyers and farmers, they meet every two weeks, and the buyers commit themselves to buy from those farmers directly. It’s like a farmer’s market, but with a social and solidarity aspect to it (it’s a bit more complex than that but this description will be enough for the point of this article). In the CSA of my friends, there are 60 buyers and 12 farmers. So the problem is that every two weeks, they have to handle the orders and accounting for many people. They are solving the orders with Excel spreadsheets that they send to each other by email, and they are solving the accounting by merging the spreadsheets into one big final spreadsheet. It is working okay, except that they are wasting a huge amount of time in communication and coordination with this spreadsheet solution. There are many of these CSAs in France, and I know at least one CSA that has a some logistics issues. My assumption was that other CSAs will have the same problems and the same needs, and maybe I could build something for them all. So now, how do I test this assumption?

Testing the CSA market using an online survey

Most people doing ISVs are building market testing websites, drive some traffic to them using AdWords, and look at how many people sign up for the newsletter or click on the disactived “buy now” button (I am not going to explain how to build market testing websites, there are plenty of blog posts out there covering this already). But I did not feel like my target market was the kind of people actively looking for CSA-oriented software. A quick check using Google Keyword Tools told me that there were 80 monthly search in France for “CSA software” (“logiciel AMAP” in French). It was going to be hard to test the market using a mini site.

I figured that the best way would be to just go out there and ask questions to these people. But I am currently based in Barcelona, and I have no time or money to travel to their houses and ask them questions. I could send them emails, but it would be time consuming to engage in conversions with people and to extract information from these conversations. Then I thought that I could simply use an online survey and that I could reach quite a bit of people doing that. Also, it would be very easy for me to organize the answers in order to extract information and take decisions with regard to the next steps. Where I got lucky is that these CSA guys are very well connected, and organized by administrative regions. I managed to find a few websites with PDF files containing email addresses of CSA groups. I manually copied 50 of these addresses into a spreadsheet to manage them with more ease, and I created an online survey using Google Forms. Finally, I wrote a quick email and I sent it with a link to the survey to these 50 email addresses. The survey contained questions regarding the organization of their groups, how they meet, how they do things, what problems they have and what they currently do to solve these problems.

I got 12 answers, that is to say 25%, which I consider to be an extremely successful result. I even engaged into direct email conversations with some of the CSAs, which was even more insightful. And so the conclusion of this survey was that all these CSAs do have a huge problem with orders and accounting, but they do not care enough about it to pay for a solution. The reasons are:

  1. Many of these CSAs are run by seniors (i.e. people in retirement), who have plenty of time on their hands, and find socially rewarding to feel they are doing something useful for their community. They are happy to waste ten hours per week on logistics and accounting tasks that they could do in ten minutes with the appropriate software.
  2. The philosophy behind the CSAs is to put the human at the center of things, and not the money. So they want to maximize the interactions they have with each others and with the farmers. Having a website managing the orders would remove part of the social interactions, and they will not allow that.

My conclusion was that even though CSAs do need software, it would be a waste of time and money for me to build it because of the philosophy of the people running them. My decision for the next step was to close the project and to not build anything for this market. However, in the emails I exchanged with some of the CSAs I tested, I got some insight regarding the farmers. It seems there could be something to build for them, so I thought it would be worth testing.

Pivoting to test a second market: the farmers

After realizing the CSAs did not want software, I also wanted to test another market, the farmers themselves. There are quite a few farmers in France who are trying to sell their production outside of the supermarket and industrial world. What they do is that they try to sell directly to customers, to minimize the loss in revenue due to the middle man. They might use CSAs or other solutions, but the thing is, they will have logistics and accounting issues as well. So I did some research, and found quite a few software already oriented towards French farmers, which seemed to solve various problems. I designed another survey with, like for the CSAs, questions about how they do things, what problems they have and what they currently do to solve these problems.

Testing the farmers market using an online survey

I found a website with a list of 7,000 independent farmers. I built a crawler in Python using Scrapy, and I got all their email addresses in a spreadsheet in a matter of minutes. I sent the survey to 350 of them, and got 30 mailer daemon answers (wrong email addresses), and 20 answers to the survey. That means there were 8.5% of wrong email addresses in the initial 350, and 6.25% of the 320 valid addresses gave me an answer. I find this quite successful too. Not as successful at the 25% reply rate of the previous survey but still, anything above 2% would be good for me.

The information I extracted from this survey is the following: they are all using software already. The worst part is that this software is being provided to them for free by the French Ministry of Agriculture. It’s kind of hard to beat that. The real issues farmers have cannot be solved with software. For instance, many of them complained that it is expensive to ship fresh produce like meat and seafood. And sometimes they have the money but cannot find a transporter to do it. I also had others telling me that they would like to solve some technical aspects of their day-to-day job: one person told me she is spending quite some time milking goats, and that she would love to automate it. Unfortunately, I cannot solve this problem with Php 🙂

I could try to create a software for farmers, but I would need to be specific as there are many specializations (meat, wine, cheese, fruits, cereals, etc.), and I would have to fight against well established software (i.e. the Ministry of Agriculture itself) who have won their trust and are present at all the farmer fairs organized in every region in France. Finally, independent farmers tend to not have much money to spend, therefore the price at which I could sell the software would not be that high. So I have decided not to pursue with this market either, as I consider that the needs farmers have and the money they could spend were not interesting enough for me.


I have a very high level of satisfaction with the results I got from these two surveys. First of all, I found a creative solution to test niche markets. Marketers have been using surveys and focus groups for a long time to test more general products. However, I haven’t found anything on the Internet yet about surveys applied to niche market testing. Then, one of the good things is that online surveys are free with today’s tools, and require just few hours of work to set up. They do require that you find an easily available list of email addresses, but this may not be as hard as it seems.

I am also happy that I got to interact with people outside of the technical world, and got more insights as to how they perceive technology. In addition, I learned more about their domains, which is interesting from a business perspective.

Finally, even though I ended up not pursuing the markets I was testing, I am really, but really happy that I did not waste time building stuff nobody wants.

That is all for now. In a future article I will cover what I learned about how to design an online survey to evaluate a niche market.


3 Responses leave one →
  1. July 9, 2012

    This is a great idea. I’ve been looking for ways to test niches for little or no money, and this idea of using surveys certainly fits the bill.

    Did you run into any trouble emailing 320 out of the blue like that? My only worry with this approach is since they aren’t “opt-in” subscribers to some sort of list, you might get flagged as spam by the recipients or even your own email provider.

    • July 11, 2012

      Indeed, sending emails at such a scale can be problematic. I have no way to know whether or not my email has been flagged as spam, but given the high response rate, I would guess it did not. I sent emails by batch of 25 to lower chances of being flagged as a spammer. Also, I started by sending only a couple of batches, and waited for a few direct email answers from some of the recipients. Those answers were just saying: “Hi, I answered your survey”, and that’s enough for what I wanted: I know very little about spam filters, but my guess is that if someone answers an email, it must reduce the chances of the content of this email to be flagged as spam. Indeed, from a spam filter’s perspective, if someone answers, then the email must be somehow meaningful. From thereon, any subsequent use of that email has less chances to end up being flagged as spam by a spam filter. But well, this is just a personal theory, and I would need to dig a bit into some spam filter implementations to clarify all that.

      As for not being flagged as spam by the recipient, I think that the key aspect is for the email to look personal enough. By personal, I mean that as long as the email seems like it was written by a human, as a personal message directed to the recipient, then very few recipients will flag it as spam.

      Finally, I considered using some spam filter analysis, like the service that Mailchimp is offering. But since the only useful ones are paying services, and I didn’t want to spend much money on this small market test, I ended up not using any. If I needed to test niche markets more regularly or on a larger scale with this technique, I would definitely pay for some spam filter analysis software.

Trackbacks and Pingbacks

  1. Europe is not ready for drop-shipping | Code Capsule

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS