Getting civiVoter and civiCanvasser back on the road

[tom0063]

Hey Peter,

Thanks for the speedy response.

One very, very desired piece is not just a paper-list calling facility, but a web-based "at home phone bank" which would allow volunteers to call, giving them one contact at a time which we have pre-selected, providing select profile information, and a facility for recording various aspects of the call (preferably this should be configurable by an administrator for each calling campaign).

I've just been elected to the central committee of our county Democratic party along with other CHANGE candidates.

We may be able to lobby some money locally or from the state party, also though our local budgets are ridiculously small. However time is critical.

When you say "our client has probably gone past the point of no return" for the NZ election to be held latest Nov 15, do you mean that these things (or similar) cannot be done by, say early September, when they would still be useful?

What would be the best-case scenario if funding were to be found tomorrow (which it won't) to develop this functionality (and then again along with the other phone-list and walk-list functionality)?

The competing options are a local low-web footprint jerry-rigged Access monstrosity or an expensive VoterBuilder technology (too expensive, the national campaign has it but the local party will never go for it).


Best regards,

Tom

[petednz]

Hey Tom - the sort of web-based phone canvassing system that you envisage was a core part of what was built by Rob for the Canadian Greens, using Drupal on civiCRM 1.6 (?).

Discussions with Rob in April suggested that we would be looking at 100 developer hours to get that out of the garage, derusted, and switched to low-octane (as in move it up to work with 2.0).

At the time Rob felt he would have a window in late May. We could not find the funding or allies to get it done. We may still have access to some funding for this (though I am about to commit some of it to a pre-civi workaround).

Rob has taken the time to show both me and Chris (our developer) around the old system. He noted at the time that he believed that 2.0 would make some of what was done in Drupal easier to achieve in civiCore. We did not fully explore what would be required to work it out of the joomla/drupal divide. And I would be happy to lead discussions about this with Lobo and his merry band of men/women.

I think the key steps we would have to take to get this on the road is
- get commitments for sufficient users to fully explore the requirements and the task ahead
- have the specs of the existing civiVoter and civiCanvasser fully clarified
- assess what is involved in getting this closer to core/standalone
- define and agree the specs of the 'coalition of the billing'
- nail down key deliverables and key dates
- see some cheques (not in my holiday account, no!)
- press the big green button.

As indicated I am interested in moving this out of the Drupal stable (though our project is using Drupal so won't be hurt if it stays there) - and it is good to have had one expression of interest from both a Joomla user and a standalone user.

I don't want/need to play a key role in pulling this together, and am sure there are others with much better project management skills than I who may be able to step in. But I was happy being the one guy in the room for a while till the party warmed up.

And yes, I think the NZ Greens would be interested in something that could be delivered within 2 months (eek) - and did I mention my 6-week holiday (oh alright 3-weeks trying not to work while in other countries and 3-weeks when I really, really am on holiday - maybe).

[JoeMurray]

Trying to restart this again regarding web-based phone canvassing. An old client has come asking about immediate availability for a by-election (highly unlikely this is possible), and more generally for a May 2009 general election. Will be pinging Rob Thorne directly.

[JoeMurray]

So if Rob Thorne estimated 100 hours for 2.0, let's round it up a bit for 2.1 and new features to be determined by coalition of the 'billing'. I'm guessing that the duration of a 100 project is 4 weeks minimum.

I'm not sure what Rob's rates are. But let's say there needs to be at least USD$15k for the project.

A client has come to me with a request for virtual call centre functionality for a by-election that may be called as early as mid-September (but more likely later). They want to use the by-election to test new approaches for a future general election. They want to know what to request from their campaign budget for this.

I'm willing to propose to my client next week that they put in $7.5k (whether they can pay for that and the glue to get things in and out of their main system is still to be determined). Who else is in a window where they can commit some money for a mid- to late-September launch?

(I'm now the really motivated party - hopefully others still have enthusiasm and budgets they can tap.)

If Rob is not available, perhaps Lobo can help us locate some engineering resources.

[Donald Lobo]

So dave and i did chat with peter about this when we had our team meeting a few months ago. We do agree that having a CiviVoter component as a core component is a good thing for the platform and we'd like to see it happen.

Without a detailed spec, its a bit hard to estimate hours etc. But based on our experience etc, but based on pete's description of the functionality of systems he has worked on, we estimated 400 hours to build the CiviVoter component. In general we are fairly conservative with our hours, but it also means we support and upgrade it for the next few releases (at least)

lobo

[Dave Greenberg]

A few other thoughts on this...

* Our estimate reflects a pretty comprehensive component (as described by Peter Davis) - and is VERY ROUGH since we don't even have a high-level requirements outline.

* If we build this - it would be a CiviCRM "native" component, part of the distribution - rather than a Drupal module. Hence, we would not be re-using any existing code.

[JoeMurray]

So there are few steps here then, and my client likely won't see a release in the by-election timeframe they're hoping for if is called early.

I will start to put more detailed work into use case descriptions over the next several days in an appropriate place on the wiki.

I am hoping others will step up.

[Donald Lobo]

Joe:

You might also want to blog about this on the CiviCRM blog and expose it to a larger section of the community. That way we can potentially get a few people out of the woodwork to support and make this component a reality. You should have blogging rights on the civicrm site

lobo

[JoeMurray]

Thanks. Will start the spec then do a post.

Cheers,

[Donald Lobo]

There are two things that can happen in parallel:

1. Write a detailed description / spec of what the component should do

2. Recruiting a group of folks willing to fund it.

if we put the target budget at 25-40K (range is due to no detailed spec), we need 5 groups willing to contribute 5K - 8K USD. We can hope to rustle a few folks out of the woodwork by blogging about it and sending mail to folks in our respective networks etc

Coming up with an early release in mid-late sept will be hard but doable (if we restrict to a minimal feature set etc). This is good timing since we are starting to wrap up the 2.1 release

lobo

[xavier]

And here is the post:

http://civicrm.org/node/415

and the spec

http://wiki.civicrm.org/confluence/display/CRM/CiviPhonebank+-+Distributed+Phone+Bank+Management

A few comments:

- some features could be used for other activities than "phone calls". Eg: (assuming each call is a seperate activities, and each person called is a contact), the features are
1) Being able to create a campaign, that will generate activities based on criterias (eg location of the callee/caller, objective...)
2) Being able to group targets of the campaign (regular/smart groups ?) so you call them
3) Being able to group users (regular/smart civicrm groups) so they can get assignments of call activities
4) being able to let each user to request a new assignement of call (create automatically the activity, automatically associated with the right contact&campaign+recap display of the contact key infos)
5) let the user change the status of the activity (reached/voicemail/convinced...), and request a new assignment

If 4) Can be easily overrided,  then you can use it to plan door to door campaigns, or any other activity than phone calls too could be managed that way

On the specs, it's mentioned a follow up of the call (as an evolution). That's a feature that has been discussed in a different context for the case. Am I over generalising seeing a lot of common points between cases and campaigns (basically, grouping planned/happened activities between identified contacts/ representative of a group) ?

X+
- integration with voip/asterisk is pretty trivial at a basic level: adding a callto:" link on the phone number (click to dial)  and with the asterisk manager, you can get the information about the call being made (+duration)

[ericcumbee]

is there a copy of CiviVoter and CiviCanvasser source code that i could look at. i've been thinking about how i would build a large scale voter database, and am really interested in seeing what direction these modules are taking.

[petednz]

Hi Eric. You would need to chat to Rob at Torenware - but also note that Matt looks like he has started some work on a phonebank system on the drupal side of the fence - see this post http://forum.civicrm.org/index.php/topic,4628.0.html

What sort of project do you have on the go?

[Donald Lobo]

check: http://cvdemo.torenware.com/ for downloads

Note that this was built for CiviCRM v1.6, so its fairly old

lobo

[ericcumbee]

Thank You.

The fact that it is for a old version of civicrm really dosnt matter to me. I have no plans of putting it into use except for research purposes. i just want to tinker with it, and hopefully eventually build my own system from scratch.