ihateit & The Founder Institute — Part VI: 2 Days To Interview 600+ People
Here is what we’ve learnt on how to run successful and cheap polls using social media. Our initial approach was wrong and naive, so hopefully, you won’t make the same mistake.
Last Tuesday, our 1 minute pitch failed to convince our panel of mentors, and we almost got kicked out from FI. As a result, we received a special assignment, and we now have 2 days left to interview 600+ people about a potential new name for our company.
This assignment made us realised that we had actually never tested our name with our target segment. This was a great opportunity to use new tools to test assumptions. Our weapons of choices were paper survey + Facebook Survey to obtain qualitative results, and Twitter poll to obtain quantitative results. We ended up interviewing over a thousand people in just 2 days for a total cost of ~$250 (We could have got similar results for $150, but we wasted $100 on our first Twitter poll).
Our Final Workflow
Step 1. Brainstorm and find 10 potential new names. With the help of some friends and marketing colleagues, we came up with:
- Rattle
- Stormy
- ihateit
- Megafon
- Angie
- irate
- irk
- Madlips
- Lipson
- Vex
- ivent
Step 2. Manually interview 30+ people who do not know about our company and belong to our targeted segment (in our case teenagers aged between 15 and 17). The interview consisted in asking them about how they felt about the names based on our company’s mission statement, and asking them to choose their favourite.
Step 3. Run a Facebook Survey ($98) targeting our segment, and compare the result with the paper survey conducted in step 2.
Step 4. Select the top 4 names. Interesting enough, both the Facebook Survey and the paper survey in step 2. resulted in the same top 4:
- Vexd
- Rattle
- Madlips
- ihateit
Step 5. Run a Twitter poll targeting our specific segment for $50.
As you can see, the results were pretty clear: Vex won. That being said, because of some trademark and domain name issues, we finally decided to use Vexd instead of Vex.
This workflow seems pretty simple and logical, but it wasn’t for us. We did a made a lot of mistakes before hitting the nail on its head.
Lessons Learnt & What Not To Do
Don’t Bother With Organic Reach. Pay To Promote Your Poll.
Unless you have a lot of genuine followers on Twitter, or have a huge network on Facebook, you won’t be able to scale your poll efficiently. Even if you could, you wouldn’t be able to target your segment properly. We spent hours trying to identify influencers on our social media channels to help sharing our poll, but that ended up being a complete waste of time. Promoting your tweet is so cheap anyway that there is no real point in relying on organic reach.
Don’t Start With A Quantitative Poll. Start With A Qualitative One.
We started to use Twitter Poll too early. We should have used it only when we had gathered enough evidence through a smaller significant sample (usually around 50+) to prove that our top 4 candidates were legit. Start with a paper survey or 101 interviews. Alternatively, because Facebook Survey allows fine tuned surveys, you could easily promote your survey to get detailed answers on there.
Don’t Interview People Who Know About Your Company Upfront, and Who Don’t Belong To Your Targeted Segment.
That seems obvious, but unfortunately that’s exactly what we did first. We relied on the people who were the easiest to reach to decide which were the names we should consider. Those people were biased because they already knew our company. They were also not part of our segment (i.e. Snapchat gen.). We asked them to vote for their favourite name out of the 11 previously mentioned. Here were the top 4 based on their answers:
- Rattle
- Stormy
- ihateit
- Megafon
Later, we were lucky to be part of an event called Initiate-48 run by Generation Entrepreneur, where teenagers aged between 15 and 17 had 48 hours to build a pitch for their startup idea. After interviewing them, the result ended up being:
- Vex
- ihateit
- Madlips
- Rattle
Facebook & Twitter Poll Features Are Very Different.
We really like the Twitter poll feature because of its ease of use, setup, and price (free to setup, and cheap to promote). It’s a bit restrictive though (4 choices max., without mentioning the tweet length restriction), but that’s also what makes it engaging. If you’re goal is to run a simple yes/no or choose your favourite option type of poll, then Twitter is your best option. On the other hand, Facebook Survey allows richer polls, which would allow you to describe in greater depth the context of your poll, as well as providing a wider variety of survey types (e.g. single choices, multiple choices, …), but it costs $50 for a single survey, and $96 for unlimited survey for one year. That cost does not include the promotion of your survey.
Conclusion
Testing assumptions is one of the keystone of the Lean Startup paradigm, but we still haven’t read articles, or met people who could clearly nail the tools and skills to achieve this efficiently. I guess this is why we still call that “hacking”. Cross-functional teams made of different skills and background gather to discover best practices are still the best way to tackle this challenge. That’s really exciting! That means that we’re living in a period where we can all contribute to build that future pragmatic framework to deliver innovation faster.
Help us by clicking the ♡ button, and follow us on Twitter for more news.
Part V: Special Assignment ← P R E V I O U S
N E X T → Part VII: Legal & IP — The Essentials
