Back to The Times of Claw

Set Up a Team CRM with DenchClaw: Complete Guide

Set up a team CRM with DenchClaw. Configure shared workspaces, user access, pipeline views, and collaborative AI workflows for your entire team.

Mark Rachapoom
Mark Rachapoom
·7 min read
Set Up a Team CRM with DenchClaw: Complete Guide

Set Up a Team CRM with DenchClaw: Complete Guide

DenchClaw started as a personal productivity tool, but it scales to team use without the architecture overhead of traditional SaaS CRMs. This guide covers setting up DenchClaw for a team: shared data, individual access, collaborative workflows, and keeping things from turning into a mess.

The core difference from individual setup: in a team, multiple people write to the same DuckDB. You need to think about access, conventions, and ownership from the start.

Architecture Options for Teams#

Before setup, choose your deployment model:

Option 1: Shared Local Server (Small Teams, 2-10 people)#

One person runs DenchClaw on a Mac/Linux machine on your network. Everyone accesses it at http://192.168.x.x:3100 (or your local domain). Simple, fast, one database.

Option 2: VPS Deployment (Remote Teams)#

Deploy DenchClaw on a VPS (DigitalOcean, Linode, your cloud of choice). Team members connect via the web UI or messaging channels. Database lives on the server.

# On your VPS
npx denchclaw --host 0.0.0.0 --port 3100

Configure with your domain, add TLS (Nginx + Let's Encrypt), and you have a private team CRM accessible anywhere.

Option 3: Dench Cloud#

For teams who want the power without managing infrastructure, Dench Cloud (dench.com) hosts a managed DenchClaw instance. Contact the team for early access.

Step 1: Install and Configure the Shared Instance#

npx denchclaw

On your server, DenchClaw initializes the workspace at ~/.openclaw-dench/workspace/. This is your team's shared data directory.

Configure a team-friendly name:

Set up this DenchClaw instance for the [Company Name] sales team

Step 2: Define Team Objects#

For a sales team, start with these objects:

Create these CRM objects:
1. Contacts — Full Name, Email, Phone, Company (relation), Title, Status (enum: Lead/Prospect/Customer/Churned), Owner (text), Last Contacted, Tags, Notes
2. Companies — Company Name, Website, Industry, Size, ARR, Stage (enum: Prospect/Customer/Churned), Account Owner (text), Primary Contact (relation to Contacts)
3. Deals — Deal Name, Company (relation), Contact (relation), Value, Stage (enum: Discovery/Proposal/Negotiation/Closed Won/Closed Lost), Owner (text), Close Date, Probability, Notes
4. Activities — Type (enum: Call/Email/Meeting/Demo/Other), Date, Contact (relation), Deal (relation), Owner (text), Notes, Next Steps

The Owner field is crucial for teams. It ties every record to a team member.

Step 3: Create Owner-Based Views#

Every sales rep needs their own view of the pipeline:

Create a view "My Deals" on the Deals object that filters to Owner = [current user] and sorts by Close Date

Since DenchClaw doesn't yet have per-user login (check current status), use an Owner text convention:

Create views for each team member:
- "Alice's Pipeline" — Deals where Owner = "Alice"
- "Bob's Pipeline" — Deals where Owner = "Bob"
- "Alice's Contacts" — Contacts where Owner = "Alice"
- "Bob's Contacts" — Contacts where Owner = "Bob"

Also create shared views:

Create these shared views:
- "All Open Deals" — active deals, all owners, sorted by Value desc
- "Stale Deals" — deals with no stage change in 14 days
- "Hot Leads" — Contacts with Status=Lead and Last Contacted > 30 days ago
- "Closing This Month" — Deals with Close Date in current month

Step 4: Set Up Team Messaging Channels#

Each team member connects their own Telegram bot or WhatsApp to the same DenchClaw instance. This way, anyone can query the shared CRM from their phone.

For Telegram: Each person creates their own bot via BotFather and connects it in DenchClaw settings. They get personalized access to the shared database.

For Discord: Create a team server, add the DenchClaw bot, and configure channel access per team role.

For Slack: DenchClaw supports Slack channels. Add it to your team Slack workspace.

Step 5: Define Update Conventions#

Without conventions, a shared CRM becomes useless fast. Establish these rules:

Ownership: Every contact and deal must have an Owner. Unowned records are reviewed weekly.

Status updates: Stage changes must include a note explaining why.

Follow-up dates: Every active deal must have a Next Action date within 14 days.

Deduplication: Before adding a contact, search first: Is there already a contact for Sarah Chen at Acme Corp?

Document these conventions in a DenchClaw document:

Create a document called "CRM Conventions" in the team workspace with our data entry rules

Step 6: Set Up Team Reporting#

Create a weekly pipeline review report:

Every Monday at 8am, generate a team pipeline report showing:
- Total pipeline by owner
- Deals closing this week
- Deals that haven't moved in 14+ days
- New contacts added last week
Send the report to the #sales Slack channel

Create a monthly revenue report:

On the first of each month, generate a revenue summary for last month:
- Total closed won by owner
- Win rate by owner
- Average deal cycle time
- Pipeline added vs. pipeline closed
Post to #sales-leadership

Step 7: Collaborative Deal Management#

For deals that involve multiple team members:

Deal rooms: Each deal can have an entry document that serves as a deal room — all notes, call logs, proposal drafts, and correspondence in one place.

Create a deal room for the Acme Corp deal. Include a notes section, open questions, next steps, and attach the proposal draft.

Handoffs: When deals change owners:

Handoff the Acme Corp deal from Alice to Bob. Add a handoff note with context.

DenchClaw updates the Owner field and logs a handoff note to the deal document.

Cross-functional context: Add non-sales objects for customer success or support:

Create a Support Tickets object linked to Companies. Track ticket volume per account.

Now sales can see support ticket volume before a renewal conversation.

Step 8: Access Control#

DenchClaw's current access model is trust-based — anyone with access to the instance can read and write everything. For teams that need stricter separation:

  1. Separate workspaces per team: Keep sales and customer success in different DenchClaw instances
  2. Read-only views: Create view-only apps for stakeholders who don't need to edit
  3. Audit logging: DenchClaw logs all AI operations — check ~/.openclaw-dench/workspace/logs/ for activity

Full RBAC (role-based access control) is on the DenchClaw roadmap.

Common Team Setup Mistakes#

Mistake 1: No owner field. Every record with no owner is a liability. Enforce ownership from day one.

Mistake 2: Too many stages. More than 7 pipeline stages creates confusion and maintenance overhead. Pick the minimum viable stages.

Mistake 3: Not deduplicating on import. Two reps import the same contact from different sources. Run deduplication after every import.

Mistake 4: No conventions document. Three people will enter company size three different ways. Write it down.

For the full DenchClaw feature overview, see what DenchClaw is. For a single-person setup, see the zero to CRM guide.

Frequently Asked Questions#

How many people can use the same DenchClaw instance?#

DenchClaw has been used by teams of 2-15 people on a shared instance. Performance scales with your server hardware. The DuckDB database handles concurrent reads well; concurrent writes are serialized.

Does DenchClaw have user accounts with separate logins?#

Not yet — this is on the roadmap. Current team setups use an Owner field convention to separate records.

Can each team member have their own AI context while sharing data?#

Yes. Each team member connects to DenchClaw via their own messaging channel (Telegram bot, etc.). Their conversation context is private; the data they read and write is shared.

What happens if two people update the same record at the same time?#

DuckDB handles concurrent writes via its transaction model. The last write wins. For collaborative editing on entry documents (markdown files), this can cause conflicts — similar to editing the same file in a shared folder.

Is DenchClaw GDPR-compliant for team use?#

Since all data lives on your own infrastructure, GDPR compliance is your responsibility. DenchClaw doesn't process or store data on external servers (unless using Dench Cloud). Consult your legal team for specific requirements.

Ready to try DenchClaw? Install in one command: npx denchclaw. Full setup guide →

Mark Rachapoom

Written by

Mark Rachapoom

Building the future of AI CRM software.

Continue reading

DENCH

© 2026 DenchHQ · San Francisco, CA