A Survey Won’t Do the Inventing for You

Overall, the first app research survey was unsuccessful. Why? I was testing the wrong thing.
In an effort to create a iPad app using customer feedback, I thought starting with customer ideas would be useful. A customer survey can help you test hypotheses for product ideas before doing too much work on them. However, a survey won’t do the hard work of generating product ideas.
The survey results show why.
My original question was: What iPhone or iPad app do you wish existed? Tweet back or answer a 3-min survey: [survey link]
Most of the app requests I got, while thoughtful, were either iPad versions of branded apps they already use or apps irrelevant to my interests and experience. See a sample.
Best answers:
- “the only ipad app i need to completely replace a laptop is a really good podcast recorder that uses a line-in.”
- “NJ transit time schedule application”
- “an app that allows downloads from amazon mp3”
- “a note taking app - simple yet attractive looking”
- “an iPhone messenger like BBM [BlackBerry Messenger]”
Overall, advertising on Twitter was a good way to generate traffic for a customer survey.
The final stats for my ad, which I ran on the 140 Proof Twitter network (PS I work there, so I’m totally biased):
- Impressions: 92k impressions and 10 Retweets (12k free impressions)
- 192 clicks (0.20% CTR)
- 41 completed surveys (21% goal conversion)
- 34 replies (14 of them useful answers)
Overall 75 people responded to the “What iPad app do you want?” question in the space of 2 days. However, as noted above, the survey didn’t yield useful insights.
There are at least two ways to improve on this process.
- Come up with 3 ideas relevant to my goals and abilities, and test those app concepts on an audience. This will give people something more concrete to respond to.
- Define a target audience (perhaps a group of people active within a certain industry, an existing user base of another app, etc), and interview/survey them to determine what they need.
#1 is easier, so I’ll start with that.