As you might already know we recently ran a poll to help us decide what the name of our future application should be. This itself was an eye-opening experience that yielded some unexpected results.
What we did
We started by crafting our prompt and creating a poll with the names we have shortlisted. You can see the result of that work here. I then tweeted about the poll and what were trying to do. I’ve also had a couple of friends ask their friends on Facebook to take the poll.
I am a recent Twitter convert and I am not by any stretch of imagination someone who has a lot of followers. There are 43 (as of now) to be exact. I asked for retweets explicitly and I managed to get a total of 1 retweet from one of my followers who in turn has 700+ followers. I also DM’ed 2 people (who (a) have a lot of followers, (b) are very influential in the community and (c) follow me) asking them to retweet. This yielded nothing. I also tweeted the message (or variants thereof) repeatedly to get good timezone and day of week coverage. In total I think the tweets alone might have had few hundreds of impressions. My estimate would be around 300. Facebook might have yielded another 100. I am actually being generous with both numbers.
This brings the total of impressions to 400 for which we got 82 page views of the poll itself. Of these, 35 people voted. You can compute corresponding “conversion” rates yourself.
Things we learned and/or confirmed
- Facebook was the top referrer, which is as expected since it’s usually easier to compel the people you know to do something. People you don’t know are more reluctant to respond.
- It is difficult enough to get people to click, but I was more surprised by how difficult it is to get people to vote. About 43% of people who ended up on the page voted. And we are not talking about a multipage survey. It’s just one simple question with 3 options – that’s it! Perhaps the problem is the message. Perhaps the problem is the medium. My intuition tells me it’s neither, it’s just people’s nature. There was no incentive for taking the poll. The results are de-personified and there is no immediate value in this for the voter.
- Don’t count on retweets. They will not come unless people you ask to retweet have a vested interest in your enterprise. Good will is rare. I am not sure about DMs. The sample (2) is far too small to comment.
- Grabbing attention is hard. The message needs to be repeated but not to the point of annoying your followers/readers. Further messages bring less and less traffic and cause more and more frustration. The less and less part was confirmed by the wonderful WordPress analytics.
This was a good experiment which exposed what were were trying to accomplish (because we had to explain the main idea behind the application, albeit briefly) and forced us to confront real people with all their unpredictability and inertia. Needless to say the overlap between this sample and who we think are userbase will be is unlikely to be very large, but it was still instructive in that we had to put together a public announcement and talk openly about what we are up to.
In the next post, I will tell you about the actual poll results and what surprised us there.