How to connect Pipedrive to Slack for deals that have gone quiet

Pipedrive is organised around activities because its premise is that a salesperson should always know what to do next. That gives it one signal no other CRM produces as cleanly: a deal with no next activity scheduled is, by the tool's own definition, not being worked.

That is the thing worth building a Slack integration around. Not a feed of stage changes, which duplicates a board that is already open on somebody's second monitor, but a quiet nudge about the deals that have gone silent.

What we'll cover

• The one signal worth acting on
• Nudging without nagging
• Building it with CodeWords
• What belongs in the team channel
• Context that makes a call better
• What Pipedrive's own automation already does
• The handoff when a deal is won
• Forecast conversations, which the channel should not replace
• Making it survive
• Limits worth knowing about
• What to set up first
• Frequently asked questions

The one signal worth acting on

No next activity scheduled is the cleanest measure of a stalled deal available in any CRM, and Pipedrive computes it natively.

It needs no interpretation. A deal with nothing planned is not waiting on a decision; it is waiting on nobody.

It is hard to game without doing the work. A salesperson can move a stage to look busy; scheduling a call means having a call.

Combine it with value and stage so the nudge is proportionate. A large deal with nothing scheduled is more urgent than a small one.

Report coverage, not completion. The share of open deals with a next step is the number that tells a manager whether the pipeline is being worked, and it is better than counting activities completed.

The three other honest signals are related: expected close dates in the past, deals aged unusually long in one stage, and values that have never changed since creation. All are quiet, all accumulate, and none produces an event.

Nudging without nagging

The tone of these messages decides whether they work or get filtered.

Send to the owner, privately. A public list of whose deals are stalling is a different message and not a better one.

Once, not daily. A recurring reminder about the same deal is a notification people learn to dismiss, and dismissing becomes a habit that covers the useful ones too.

Group into one message. "Four deals have nothing scheduled" with the list is read; four separate messages are not.

Send it at the start of the day, when somebody can act, rather than at the end when they cannot.

Let them act from the message — schedule an activity, or mark the deal as lost, which is frequently the honest answer and the one that takes the longest to admit.

Report the total to the manager weekly, not the individual nudges, so the pattern is visible without anybody being singled out daily.

Building it with CodeWords

CodeWords connects to more than 3,000 integrations, and the connection is made once and reused.

1. Open CodeWords and start a new automation.
2. Describe what should happen in plain language to Cody, the automation builder: which deals, what condition, and who should be told.
3. Authorize the connection to Pipedrive with an API token for a dedicated integration user, and to your Slack workspace.
4. Describe the exceptions: a deal already won or lost, an owner who is away, a deal deliberately parked.
5. Test against one pipeline and a test channel before anything reaches the sales team.

You describe the outcome; Cody builds it, connects it, and deploys it. The free plan covers light use, with Pro at $39 per month and Business at $100 per month as usage grows; details are on the pricing page.

What belongs in the team channel

A short list, because a channel of deal activity is a worse version of the board.

Deals won. Genuinely worth announcing — it is good for the team and it tells the people downstream that work is coming.

Deals above a threshold changing stage, to leadership rather than to everybody, since large deals warrant attention routine ones do not.

Deals lost with the reason, to whoever should learn from it, which is usually product and marketing.

The weekly pipeline summary: total value by stage, what moved, and the coverage number.

Nothing per activity, per email, or per field edit. Those are the record of work, not events, and forwarding them buries the four above.

Context that makes a call better

Where a Slack integration does something the CRM cannot.

Product usage on a trial, so the salesperson knows whether the prospect has actually used the thing before calling to ask how they are finding it.

Open support tickets, so nobody makes an upsell call to an account with an unresolved escalation.

Payment status, for an existing customer, which lives in a different system and changes the conversation.

Recent activity from the company, such as a new signup from the same domain, which is a buying signal nobody surfaces.

Deliver it as a lookup somebody triggers, not as a push, since the useful moment is just before a call and the automation cannot know when that is.

What Pipedrive's own automation already does

Check before building, because some of this exists.

Workflow automations create activities, move deals, send emails, and can post to Slack directly.

Required fields per stage enforce data entry at transitions.

Email sync attaches correspondence to deals for users who enabled it.

Activity reminders already notify owners about scheduled activities.

If the requirement is "tell the channel when a deal is won", Pipedrive's own automation does it and is visible in the settings to whoever manages the sales process. Build custom for the stalled-deal analysis, for anything joining Pipedrive with product or support data, and for anything where the message needs to be actionable.

The handoff when a deal is won

The moment where a good sales process most often produces a poor customer experience.

Tell the delivery or onboarding team in their channel, not in the sales channel where they are not looking.

Carry what was promised. The commitments made during the deal live in notes and emails, not in fields, and they are the single most useful thing to move across.

Record what was actually sold, including any non-standard terms, since that is where delivery and billing disagreements begin.

Name the salesperson, because the handoff will generate questions and knowing who to ask saves a day.

Close the loop back when onboarding completes or stalls, so the salesperson hears it from a colleague rather than from the customer.

Forecast conversations, which the channel should not replace

A weekly pipeline summary is useful and it is not a forecast, and conflating the two causes trouble.

A summary reports what the CRM says. A forecast is a judgement about what will close, and the gap between them is the entire job of a sales manager.

Report the inputs, not the conclusion. Value by stage, deals with close dates this period, and the coverage number — then let the conversation happen.

Flag the close dates that have passed rather than quietly rolling them forward, since that is where optimism accumulates.

Show the change since last week, because a stage distribution on its own looks the same every week and tells nobody anything.

Do not post individual forecasts to a shared channel. A public per-person number turns a planning conversation into a performance one.

Keep it to five lines. A summary that reads like a report gets skipped, and the whole point is that somebody glances at it before the meeting rather than during it.

Making it survive

Make it idempotent, keyed on the deal and the condition, so a deal is not nudged repeatedly for the same reason.

Resolve stage identifiers per pipeline, since they differ and a hard-coded one works on exactly one pipeline.

Handle the away owner. A nudge to somebody on holiday should hold or escalate rather than disappearing.

Respect deals deliberately parked, using a field or a label the team agrees, or the nudge becomes noise for the deals that are legitimately waiting.

Report the outcome. Nudges sent, deals covered, and anything skipped.

Limits worth knowing about

Rate limits apply and depend on your plan, with a token budget that bulk work exhausts quickly.

Search is not immediately consistent, so use the identifier returned when you created a record rather than searching for it.

Custom field identifiers are opaque strings, which is what makes them stable and makes queries hard to read.

Pagination applies to listings, and a report built from the first page understates everything.

Some features are plan-dependent, including parts of the automation and reporting capabilities.

What to set up first

The stalled deal nudge: open deals with no next activity scheduled, grouped into one private message to the owner at the start of the day, with the option to schedule something or mark the deal lost from the message. It is the one signal Pipedrive produces that nothing else surfaces and that a manager cannot easily assemble.

Two habits make the difference. Group them into one message rather than sending one per deal, since four separate notifications get dismissed as a batch. And include marking the deal lost as an option, because a pipeline full of deals nobody will ever close is the thing the nudge is really about.

Frequently asked questions

What number belongs at the top of the weekly summary?

The share of open deals with a next activity scheduled. One figure, compared against last week, that tells a manager whether the pipeline is being worked — which a stage distribution looks identical every week and never answers.

Should stalled deal alerts go to a channel?

No — privately to the owner, grouped into one message, once. A public list of whose deals are stalling is a different message from a quiet nudge, and it produces defensive behaviour rather than scheduled calls.

Should the nudge offer to move the deal?

It should offer to schedule an activity or mark the deal lost, not to change stage. Those two are the actions that resolve a stalled deal honestly; moving a stage from a chat message just makes the board look tidier.

What should the team channel receive?

Wins, large deals moving, losses with reasons, and a weekly summary. Not activity, email, or field-edit events — those are the record of work rather than news, and they bury the four that matter.

Can I see product usage before a sales call?

Yes, as a lookup somebody triggers rather than a push, since the useful moment is just before a call and the automation cannot know when that is. Joining usage, support, and payment status is the thing Pipedrive alone cannot do.

Why does the alert cover one pipeline and not another?

Stage identifiers are per pipeline, so a rule written against one pipeline's stages matches nothing in another. Nothing errors, so the only thing that surfaces it is reporting how many deals each run actually examined.

Should wins be announced to everybody?

Announce them in the team channel, and send the handoff detail to the onboarding team's channel separately. They are two different messages for two different audiences, and combining them means the delivery team skims past the part they needed.

Should the weekly summary be a forecast?

No. Report what the CRM says — value by stage, close dates this period, coverage, and the change since last week — and leave the judgement about what will actually close to the conversation. Conflating the two makes the summary something people argue with rather than read.

What is the most revealing line in a pipeline summary?

Deals whose expected close date has already passed. It is where optimism accumulates, it is invisible on a board, and rolling those dates forward quietly is the most common way a forecast stays wrong.

Related reading

• Pipedrive and CodeWords
• Connecting HubSpot to Slack
• CodeWords integrations and templates