FeverBee SPRINT
Psychology of Communities
San Francisco, Nov 11 - 12

http://sprint.feverbee.com
Hi everyone,

We now have just 79 days until FeverBee SPRINT. Tickets prices will be rising again at the end of this week, buy tickets before the end of the week to get the best rates.

In this e-mail I want to talk about the key activity behind complete community optimisation; testing.
 
You can’t optimize anything you’re not willing to test.
 
Testing begins with a hypothesis. You benchmark the current efforts. You try something different. You use the results to guide future actions.
 
For example, a possible test might be:
 
“If someone makes a complaint, does offering them a choice of solutions lead to fewer dissatisfied responses than resolving the complaint?
 
Most tests will show no real impact. That’s the nature of tests. Sometimes it just doesn’t matter if notifications go out in your name or from the community, for example.

Some tests will have a very marginal impact. They can be statistical errors or false positives (or negatives). It’s the 25%+ tests we care about.
 
The 25%+ tests are the game-changers.
 
Believe me, there are a LOT of 25%+ opportunities out there.
 
Here are 10 tests you could run before the year is up.
  • Testing digests vs. e-mail notifications. Measure statistical average number of weekly posts over 3 months, turn on/off digest notifications, check the increase vs. decrease.
  • Testing telling your newcomers to describe what they’re working on, their biggest challenge, or their future goals. Measure how many newcomers make a contribution today. Change the call to action to participate in a discussion related to one of the above. Measure how many participate.
  • Test mentioning a member’s name in your response. When replying to a member, mention their name in the post (or tag them if you can). Test whether this increases the likelihood of them participating again by 25%.
  • Test long-form content vs. short-form content. Measure the number of clicks, shares, and comments to a detailed, long, article, and short, simple, tips. See whether either increases activity. Test posting 5 times a week, three times per week, and once per week too. 
  • Test different discussion topics. Run a survey of your community members. This should be a maximum of five questions. Ask members what products/software/hardware they use. Ask what they spend most of their time doing. Ask them to rank their biggest problems and their future hopes / ambitions. Test discussions on content on these topics.
  • Test different prompts to participate. Test what is more effective between appeals to participate for the benefit of the community, to increase their own status, to help other members, or simply to ‘respond’ to an existing discussion. Measure how many click the link and response. This works for your e-mail subject lines too. 
  • Test taking a contrary view vs. agreement view. Measure the average number of responses to a discussion. Now for two weeks, add a contrary/different perspective to every response you give in a community. Then in the latter two weeks agree with existing views / call for further information. Does either increase the average number of responses by 25%? 
  • Test relationship-building techniques. Measure which e-mails lead to members positively replying to your efforts to build relationships. Does flattery work best? Asking an open-ended question? Or making a statement the recipient is likely to agree with?
  • Test your own personality. Should you be serious? Lighthearted? Kind/generous? Measure the average number of responses to your posts with different personality types, whether the sentiment is positive/negative, and whether there the quality of responses is higher. 
  • Test the length of webinars (and topics). The length of a webinar drastically influences the level of engagement. Measure whether people are dropping out and test pure tactical webinars with deeper, theoretical, webinars. 
What works for one group may not work for the next. 
 
A few rules for these tests:
  • Only run one test at a time. For obvious reasons. Don’t contaminate the results.
  • Measure over the long-term. Sometimes the novelty attracts more activity/participation in the short-term. Check whether the results remain 25% higher. . 
Good luck.

Richard Millington
@RichMillington
www.feverbee.com
+44 (0)7763 831931
Copyright © 2015 FeverBee Limited, All rights reserved.


unsubscribe from this list    update subscription preferences