CRM for Outreach

[Wituareard]

Joep already has a testing instance of Twenty with some email-derived data in it - good for validating what it can do now we are at version > 1.

As Nils notes, we would eventually expect to use a self-hosted instance

[Wituareard]

@anthonybailey ^

[anthonybailey]

We wanted a place to discuss CRM choices hence this issue.

We may well end up with e.g. a Twenty fork and at that point we'd want a new repository under Pause AI org in GitHub, but for now we have this.

I've dumped an amalgam of previous CRM plans and history below.

[anthonybailey]

Joep from https://discord.com/channels/1100491867675709580/1213591270992379935/1387330295128129601:

Current CRM situation (for outreach / lobby)

We have a hubspot instance that I know me, Chris, Diego, William baird have used / are using . But the whole CRM chapter needs work.

Some problems:

• Kinda hard to use / set up. Requires browser extension that integrates with gmail.
• Email integration contains very sensitive data, so you can't share these with anyone. Only a small group of trusted folks.
• CRMS in general contain sensitive data (e.g. the emails of potential targets are sensitive)
• Hubspot is very expensive when > 5 users are in there

[anthonybailey]

Anthony et al in https://discord.com/channels/1100491867675709580/1406730286259245158/1406794815659573451

We lack a good CRM solution (most assume concentrated seats, we need any member to be able to contribute, and good email integration.)

A year back Twenty was rising but not ready. How is it now?

Nils:

twenty v1 was released 2 months ago

[anthonybailey]

Maximiliano et al discuss CRMs for volunteer management rather than outreach here:

https://discord.com/channels/1100491867675709580/1324358376196669460/1324358376196669460

We basically use AirTable for this and it is working OK in practice, but we would swap something else in if appropriate. This issue is supposed mainly to be out an Outreach CRM though.

Dumping my own comments from Discord...

---

We have several distinct CRM needs. Outreach team (Diego) needs to track external parties we engage (influencers, experts, policy folk.)

Sean Pan is keen on more automated assistance in such a system, and Max Winga expressed interest in wiring it up but does have a proper job working at Conjecture!

We also want to track actual and potential movement membership and activity metrics (at both global and regional chapter levels.) Ella is your point person here. Onboarding should be a subset of that given flow.

---

Absolutely no need for these to be either the same or different CRM systems.

We've played with HubSpot. Main issues have been that the business model assumes individually licensed seats in an org each using the CRM a lot. But, a volunteer movement has a lot of folk who'll mail just a few folk (representative, etc.) and tracking that in the same system isn't typically well supported. We also found it less trivial than expected to populate by backfill forwarding and cc-ing ongoing email outreach to hook addresses.

I will note we also looked seriously at https://www.twenty.com/ as something to build on, if our combined needs for LLM-assisted-automation and wide user base require enough bespoke work. (Edit: more detail in following GitHub comment.)

---

Because we were mostly discussing requirements, the implementation Maximiliano proposed remains mostly unexamined.

I would like to call out that last time we discussed CRM requirements was more in the context of outreach, where we came to understand that the license regime for typical CRM products isn't well suited to a movement composed of a large number of volunteer members, any of whom might be writing a few contacts. This was one of the main reasons we were considering building rather than buying.

For managing members in the context of current growth plans then especially if we become less funding constrained it is much less clear that off-the-shelf solutions won't be fit for purpose. Buying and integrating standard systems can be much better than building from scratch.

(Aside: even the CRM idiom is somewhat weaker in the current context. How much are our audience of potential and active volunteers like customers of the "product" that the central Pause AI org is "selling"? Well, some - it can be a useful lens - but hiring / staff management is another appropriate perspective and the set of systems that help do that are much larger than CRM per se.)

---

And a similar overview of CRM for volunteers by Will in https://discord.com/channels/1100491867675709580/1317270365201891419/1319439241314439209

Recommendations regarding action priorities and sequence:

• CRM needs a well-defined scope and purpose that is immediately useful and the work involved is proportionate to how long we expect it to hold up. Primary suggested scope = managing volunteers in setting up local chapters. Thus, what we are going to be asking of volunteers in the next 3-6 months, especially with regards to setting up local chapters, should drive the CRM design.

• Volunteer mobilization is action priority 1, but it requires having a clear plan for what we will be asking them to do and to paint at least a rough picture of the path to victory. This means clarifying our policy stance on the international level as well as principles to drive policies we support or oppose on national and local levels--but it may be OK if this is a little vague for now (and so maybe is already done). It also means having a plan for escalating commitment and political pressure. This can also be filled in over time, but the minimum threshold is to inspire confidence.

• There is a sequential action path along the line of CRM -> volunteer mobilization -> expanding national chapters -> political targeting -> national/local policies -> international treaty, but planning works backwards.

• Fundraising, gaining public support of experts and major figures, and mainstream media coverage are all great and there is no reason to shoot for the moon wherever it is low effort, but I expect that these will mostly happen on their own as a positive feedback loop after we achieve some success by other means.

• I don't draw any arrows from fundraising because I take it as assumed that fundraising accelerates everything. Given what we have done with volunteers, I am not sure if anything is strictly bottlenecked by money.

• Improving messaging will facilitate everything else in a practical sense, but is heavily context-dependent and so is informationally downstream of everything else.

[anthonybailey]

Our oldest attempt is an Airtable form one can separately fill out to describe an outreach activity: https://airtable.com/appWPTGqZmUcs3NWu/pagoraIwzlYm2QwjD/form

It has been used somewhat, but not much.

127 comment thread from Discord on CRMs (for lobbying / outreach): https://discord.com/channels/1100491867675709580/1213591270992379935

Zoho was considered.

Chris Gerrby wrote a lot of ideas, including

---

I wrote in #volunteers, I'll paste here as well:

**What is the current demand for a proper CRM right now? **

Where the project is so far:

• Me and Louis Bergman have made a soft decision on CiviCRM Spark.
  • This solution, even thought pretty simple by comparison to other ones, requires a decent amount of time to learn and to set up.
• "Here is a comparison of different CRMs solutions*: https://docs.google.com/spreadsheets/d/1pIPKHQvOjg9mA-DJUdC2DpeyIZjweofGw9E-GhNE2Uk/edit?gid=0#gid=0
  • lists Blackbaud, Salesforce Nonprofit Cloud, Bonterra / Salsa CMR, CiviCRM, SuiteCRM, NationBuilder "Action Network", Bitrix24, Hubspot for nonprofits, Smartico, Pipedrive, Airtable, Odoo, Twenty, Solidarity, Zoho Bigin

---

Pato found https://activisthandbook.org/tools/crm#crms-for-activists which (among other choices) led us to investigate Twenty.

There's a downstream thread on the Twenty Discord: https://discord.com/channels/1130383047699738754/1130383048173682821/1276446670317551718

I would summarize outcomes as: looks promising, but currently lacked features we wanted, especially re integration through email (which we want for fwd-based backfill and to avoid every Pause AI member needing to use yet another system: instead they can just bcc an address.)

Joep:

Idea: Add a bcc@pauseai.info group to gmail. The goal is to forward emails there that may be relevant to PauseAI. It's basically a way to aggregate CRM stuff.

This hopefully feels far less bothersome than uploading to the outreach form.

At some point in the future, we may simply forward all of these mails to the real CRM!

We considered serious investment into Twenty and contributing implementation of those features, but demurred and left it.

We were excited that HubSpot appeared to offer decent email integration, and the small Outreach team ended up using HubSpot seats, but we were kind of sold a non-existent feature. Outreach from that period is captured some in the AirTable form, some in HubSpot, but also random Google Sheets, and people's memories. Great work by the team, but capture hasn't been great, and now the team is being restarted by Tom we want to do better.

And Twenty now seems to have the features we need! Yay.

[anthonybailey]

We will also discuss Muster with the US chapter since they aim to buy that: https://discord.com/channels/1286529161510387722/1398750647196848148

[joepio]

I've setup a Twenty instance (trial) and invited Anthony.

[anthonybailey]

Muster turned out to be US specific.

The product Pause AI US have used the most in practice is the activist stack https://www.littlegreenlight.com/

It looks pretty great! But, alas, again intended for US-only movements, not global ones.