4 | | The following is a list of suggestions for getting more people to participate in your project. Much of this is based on [http://boinc.berkeley.edu/poll_results.php the results of the BOINC user survey]; study this yourself. |
5 | | |
6 | | Most of these suggestions involve writing prose to be read by the general public. If (like many scientists) you are not good at this or avoid doing it, find someone who is good at it. This could be one of your students, a friend of a friend, or a professional writer. |
7 | | |
8 | | English is the most widely-spoken language among BOINC participants, and you should probably use it as the main language for web materials and email. BOINC provides mechanisms for [TranslateIntro Web site translation]; it's generally easy to get volunteers to do this. |
| 4 | The following is a list of suggestions for getting more people to participate in your project. |
| 5 | Much of this is based on [http://boinc.berkeley.edu/poll/poll_results.php the results of the BOINC user survey]; |
| 6 | study this yourself. |
| 7 | |
| 8 | Most of these suggestions involve writing prose to be read by the general public. |
| 9 | Find a good writer. |
| 10 | This could be one of your students, a friend of a friend, or a professional writer. |
| 11 | |
| 12 | English is the most widely-spoken language among BOINC participants, |
| 13 | and you should probably use it as the main language for web materials and email. |
| 14 | BOINC provides mechanisms for [TranslateIntro Web site translation]; |
| 15 | it's generally easy to get volunteers to do this. |
13 | | * Present your project's credentials: the educational credentials of its leaders, its research track record, and the status of its institution. |
14 | | * Describe what your project is doing: its high-level scientific goals, its methods, the details of the computation being done using volunteers, and the (non-distributed) computations that precede and follow this. How will your research affect the lives of everyday people now and/or 50 years from now? |
15 | | * Who owns the intellectual property that arises from volunteer computations? Will it be released to the public? When, and under what terms? |
16 | | * Show all the scientific results of the computation so far, and any publications that arise from these results. ([http://depts.washington.edu/~bakerpg/publications.html Rosetta@home] and [http://folding.stanford.edu/papers.html Folding@home] provide good examples of this). Announce new results and publications on the News column. Make sure your News column is being properly published as an RSS feed. |
17 | | * Give some personal information about your team members: their names, background, interests, and preferably a photograph. This will 'humanize' your project in the eyes of potential participants. |
18 | | * Take an active role in your web site's message boards. Read them frequently, and respond quickly to any negative threads that arise. Make a periodic posting giving 'insider info' on your project. |
19 | | * Make sure your the web site has clear navigation, so that the above information is easy to find from the front page. Do a user study - show your web site to a strangers, ask them to browse it and/or to find particular information, and get their feedback (you may be surprised). |
20 | | * If possible, create a graphical identity (logo, color scheme, etc.) for your project. Your web site should project professionalism and inspire confidence and interest in prospective volunteers. |
| 20 | * Present your project's credentials: |
| 21 | the educational credentials of its leaders, its research track record, and the status of its institution. |
| 22 | * Describe what your project is doing: its high-level scientific goals, |
| 23 | its methods, the details of the computation being done using volunteers, |
| 24 | and the (non-distributed) computations that precede and follow this. |
| 25 | How will your research affect the lives of everyday people now and/or 50 years from now? |
| 26 | * Who owns the intellectual property that arises from volunteer computations? |
| 27 | Will it be released to the public? When, and under what terms? |
| 28 | * Show all the scientific results of the computation so far, |
| 29 | and any publications that arise from these results. |
| 30 | ([http://depts.washington.edu/~bakerpg/publications.html Rosetta@home] |
| 31 | and [http://folding.stanford.edu/papers.html Folding@home] provide good examples of this). |
| 32 | Announce new results and publications on the News column. |
| 33 | Make sure your News column is being properly published as an RSS feed. |
| 34 | * Give some personal information about your team members: |
| 35 | their names, background, interests, and preferably a photograph. |
| 36 | This will 'humanize' your project in the eyes of potential participants. |
| 37 | * Take an active role in your web site's message boards. |
| 38 | Read them frequently, and respond quickly to any negative threads that arise. |
| 39 | Make a periodic posting giving 'insider info' on your project. |
| 40 | * Make sure your the web site has clear navigation, |
| 41 | so that the above information is easy to find from the front page. |
| 42 | Do a user study - show your web site to a strangers, |
| 43 | ask them to browse it and/or to find particular information, |
| 44 | and get their feedback (you may be surprised). |
| 45 | * If possible, create a graphical identity (logo, color scheme, etc.) for your project. |
| 46 | Your web site should project professionalism and inspire confidence |
| 47 | and interest in prospective volunteers. |
23 | | The world will not beat a path to your door. You need to work hard to spread the word about your project. |
24 | | |
25 | | * Get in the mass media (newspapers, magazines, radio, television) as much as possible. If your institution has a PR director or media spokesperson, contact them while you're developing your project, and again any time your project has major news. If no such person is available, call local media outlets yourself. |
26 | | * Exploit existing organizational relationships. If you work at a University, try to get your project running on the PCs in the teaching labs, and on the PCs of students, faculty and staff. If you have connections with organizations with PR capabilities (i.e., web sites or newsletters), enlist their support for your project, and get them to publicize it. A typical example: professional organization in your subject area. |
27 | | * The BOINC web site will generally announce new projects. Also, make sure your project is listed on [AccountManagers account managers] like !GridRepublic and BAM! |
28 | | * Use the web. Announce your project in forums like Slashdot, and on the message boards of the major cross-project teams like BOINC Synergy, Overclockers UK, Team Anandtech, etc. |
| 50 | To get volunteers, you must publicize your project. |
| 51 | |
| 52 | * Get in the mass media (newspapers, magazines, radio, television). |
| 53 | If your institution has a PR director or media spokesperson, |
| 54 | contact them while you're developing your project, |
| 55 | and again any time your project has major news. |
| 56 | If no such person is available, call local media outlets yourself. |
| 57 | * Exploit existing organizational relationships. |
| 58 | If you work at a University, try to get your project running on the PCs in the teaching labs, |
| 59 | and on the PCs of students, faculty and staff. |
| 60 | If you have connections with organizations with PR capabilities |
| 61 | (i.e., web sites or newsletters), |
| 62 | enlist their support for your project, and get them to publicize it. |
| 63 | A typical example: professional organization in your subject area. |
| 64 | * The BOINC web site will generally announce new projects. |
| 65 | Also, make sure your project is listed on [AccountManagers account managers] |
| 66 | like !GridRepublic and BAM! |
| 67 | * Use the web. |
| 68 | Announce your project in forums like Slashdot, |
| 69 | and on the message boards of the major cross-project teams like BOINC Synergy, |
| 70 | Overclockers UK, Team Anandtech, etc. |
34 | | * ''Newsletters''. These are periodic (perhaps every few months) and are sent to all participants. Typically you would use them to summarize your project's results, to discuss its future plans, to make announcements, etc. |
35 | | * ''Reminders''. These are sent to participants who seem to have stopped computing to your project, or who signed up but never got any credit. Typically they would be short messages, encouraging the participant to take a specific action. |
36 | | * ''Friend-to-friend''. These are sent by participants to their friends and family, to tell them about your project and urge them to join. The sender can add an optional message. |
37 | | |
38 | | Effective use of all types of email is critical to maintaining and growing your participant base. In the absence of any email, participation typically decreases by a few percent every month. BOINC supplies the framework, but you must write the actual emails, or modify BOINC's samples as needed for your project. |
| 76 | * ''Newsletters''. These are periodic (perhaps every few months) |
| 77 | and are sent to all participants. |
| 78 | Typically you would use them to summarize your project's results, |
| 79 | to discuss its future plans, to make announcements, etc. |
| 80 | * ''Reminders''. These are sent to participants who seem to |
| 81 | have stopped computing to your project, or who signed up but never got any credit. |
| 82 | Typically they would be short messages, |
| 83 | encouraging the participant to take a specific action. |
| 84 | * ''Friend-to-friend''. These are sent by participants to their friends and family, |
| 85 | to tell them about your project and urge them to join. |
| 86 | The sender can add an optional message. |
| 87 | |
| 88 | Effective use of all types of email is critical to maintaining and growing your participant base. |
| 89 | In the absence of any email, participation typically decreases by a few percent every month. |
| 90 | BOINC supplies the framework, but you must write the actual emails, |
| 91 | or modify BOINC's samples as needed for your project. |
42 | | * They let you send different emails to different 'classes' of participants. For example, you can send a different newsletter to participants who haven't computed for your project in a while. |
43 | | * They let you personalize emails, e.g. by inserting the participant's name or their total credit. |
44 | | * They provide a mechanism for inserting a secure 'opt-out' link. Note: You should ALWAYS include an 'opt-out' link at the bottom of emails (both HTML and text). It may be illegal for you to do a mass email without one. Make sure you test this link. |
45 | | |
46 | | The scripts requires that you use [http://phpmailer.sourceforge.net/ PHPMailer], a PHP function for sending mail that is more full-featured than the one built into PHP. Download it, put it in `html/inc`, and set the `USE_PHPMAILER`, `PHPMAILER_HOST`, and `PHPMAILER_MAILER` variables in [WebConfig your project.inc file]. |
47 | | |
48 | | All of the tools let you send multipart HTML/text messages. We recommend that you use this feature - and HTML message can include your logo and/or institutional insignia, can include hyperlinks, and can look more attractive. |
| 95 | * They let you send different emails to different 'classes' of participants. |
| 96 | For example, you can send a different newsletter to participants |
| 97 | who haven't computed for your project in a while. |
| 98 | * They let you personalize emails, e.g. by inserting the participant's name or their total credit. |
| 99 | * They provide a mechanism for inserting a secure 'opt-out' link. |
| 100 | Note: You should ALWAYS include an 'opt-out' link at the bottom of emails (both HTML and text). |
| 101 | It may be illegal for you to do a mass email without one. Make sure you test this link. |
| 102 | |
| 103 | The scripts requires that you use [http://phpmailer.sourceforge.net/ PHPMailer], |
| 104 | a PHP function for sending mail that is more full-featured than the one built into PHP. |
| 105 | Download it, put it in `html/inc`, and set the `USE_PHPMAILER`, `PHPMAILER_HOST`, |
| 106 | and `PHPMAILER_MAILER` variables in [WebConfig your project.inc file]. |
| 107 | |
| 108 | All of the tools let you send multipart HTML/text messages. |
| 109 | We recommend that you use this feature - and HTML message |
| 110 | can include your logo and/or institutional insignia, can include hyperlinks, |
| 111 | and can look more attractive. |
52 | | * Create a directory (mass_email, reminder_email, or ffmail) in your `html/ops/` directory. In that directory, create separate files for the text body template, HTML body template, and subject line to be sent to each class of participants. NOTE: the HTML files are optional; if you leave them out, text-only emails will be sent. |
53 | | * Run the script in testing mode (see below) to ensure that the emails are as you intend. |
54 | | * Once testing is complete, run the script in production mode. Typically, the newsletter script is run from the command line. The reminder script is typically run as a [ProjectTasks periodic task], every 24 hours or so. |
55 | | |
56 | | The newsletter and reminder scripts use the recent-average credit (expavg_credit) field in the user table. To make sure this value is accurate, run [ProjectTasks update_stats] manually if you're not running it as a periodic task. |
| 115 | * Create a directory (mass_email, reminder_email, or ffmail) in your `html/ops/` directory. |
| 116 | In that directory, create separate files for the text body template, |
| 117 | HTML body template, and subject line to be sent to each class of participants. |
| 118 | NOTE: the HTML files are optional; if you leave them out, text-only emails will be sent. |
| 119 | * Run the script in testing mode (see below) to ensure that the emails are as you intend. |
| 120 | * Once testing is complete, run the script in production mode. |
| 121 | Typically, the newsletter script is run from the command line. |
| 122 | The reminder script is typically run as a [ProjectTasks periodic task], every 24 hours or so. |
| 123 | |
| 124 | The newsletter and reminder scripts use the recent-average credit |
| 125 | (expavg_credit) field in the user table. |
| 126 | To make sure this value is accurate, run [ProjectTasks update_stats] |
| 127 | manually if you're not running it as a periodic task. |
204 | | * Failed: the account was created at least 14 days ago, has zero total credit, and hasn't received a reminder email in 30 days. These people typically either had a technical glitch, or their hardware and/or preferences didn't allow sending them work, or the application crashed on their host. The reminder email should direct them to a web page that helps them fix these problems. |
205 | | * Lapsed: the user has positive total credit, hasn't done a scheduler RPC in the past 60 days, and hasn't been sent a reminder email in the past 30 days. They probably stopped running BOINC or detached this project. The reminder email should gently prod them to start running BOINC and attach to this project again. |
| 288 | * Failed: the account was created at least 14 days ago, has zero total credit, |
| 289 | and hasn't received a reminder email in 30 days. |
| 290 | These people typically either had a technical glitch, |
| 291 | or their hardware and/or preferences didn't allow sending them work, |
| 292 | or the application crashed on their host. |
| 293 | The reminder email should direct them to a web page that helps them fix these problems. |
| 294 | * Lapsed: the user has positive total credit, |
| 295 | hasn't done a scheduler RPC in the past 60 days, |
| 296 | and hasn't been sent a reminder email in the past 30 days. |
| 297 | They probably stopped running BOINC or detached this project. |
| 298 | The reminder email should gently prod them to start running BOINC |
| 299 | and attach to this project again. |