Hmmm Dave, you mentioned that buying a decent CRM would be expensive. Anybody who would invest serious time into a flawed app is truly biting off an expensive mouthful. The GPO doesn't have much $$, and so cannot afford to fool around with time consuming almost good enoughs. I'm sorry, but civiCRM might be just tickety boo for light usage, but it is too slow and clunky for sustained efforts.
How do I distill 20 years of work experience for you? Start by always thinking of the users as people implementing a PROCESS. There are the processes of making sequential calls for canvassing, or systematically renewing memberships, or simply identifying issues, or an inhouse opinion poll. If you distill all the desired outcomes, and processes into a linear flow, then you are approaching the ideal world of process management. All of the processes for which the CRM will be utilised are extremely labour intensive. By making transition from task to task in a sequential process quicker, and error free you are improving productivity proportionally. In English, a 10% efficiency improvement in a linear process yields a 10% better return on invested effort. The difficulties you may be experiencing in setting up, and implementing the CRM pale into insignificance when you look at what the potential benefits are to a multitude of users.
Hey, maybe civiCRM has magically speeded up since last i attempted to use it as a hosted canvassing d-base. If it hasn't then it is a complete waste of your effort. The users would be better served in sending around spreadsheets. You will lose the major advantage of retaining all your legacy data in a central database if your hosted app isn't head and shoulders above the alternatives.
Now, if you want to get ambitious about creating a game changeing Canvassing app, go and have a look at the capabilities of Insidesales.com. They have heavy duty hard/software to manage autodialling, multiple open lines for each canvasser with automated digital message drops in the background. One example of game changeing is the fact that the canvasser clicks a button to drop a pre-recorded message on an answering machine, while automatically dialling the next number. Autodialling of phone numbers yields somewhere around 50% productivity improvement for the canvassers. Message drop is up between 100% -200% productivity improvement. a tool with only these two features will turn the efforts of 10 canvassers into something comparable to 25-35 volunteers with phone sets and lists. Now that's a game changer.
That's for the simple electoral tasks. Now consider what happens if you have a more complex outreach program, with progressively more valuable 'steps' in the process. I'm thinking of what should have been done years ago at the GPC, bringing outreach in house. If the task is to contact people to determine level  of support for the green party. basic outcome is identified green supporter. The process then attempts to sell a one time membership, then progresses to monthly donor, then volunteer, then EDA exec/candidate. These are sequential asks, with escalating values of outcomes. An efficiency improvement in any one step of the process will improve the numbers of every subsequent process by 10%. eg 10% more cold calls means 10% more members, 10% more monthly donors, 10% more volunteers, 10% more EDA execs. Every improvement in productivity yields enormous returns, whether that improvement comes from script improvements, quicker reply email templates( collateral materials), better training, more accurate and pre-qualified lists, or simply quicker and more accurate CRM.
If I have anything to contribute to this discussion, it's that you must not be stingy with the underlying application. We can leverage our volunteers way way better, and we are ALWAYS short on volunteers. We can hope to have thousands of volunteers nationally, (or Provincially), during an election, and doubling their output is an enormous, and close to free lunch.
Bluegreenblogger a.k.a. Matthew Day