If you want to run two WhatsApps (one for sales and another for after-sales/support) within the Kommo CRM without duplicating leads, without losing history, and without fighting over “contact owner,” the solution is simple: a single contact + multiple pipelines (funnels) + entry rules + ownership governance. The mistake is trying to “separate” by creating two CRMs, two databases, or two contacts. This kills ROI, kills tracking, and causes internal conflict.
I am Luiz Otávio Gonçalves (mechanical engineer turned digital engineer) and, in practice, this design is one of the most effective at unlocking customer service scale with WhatsApp in Brazil — especially when the company grows and realizes that “selling” and “retaining/serving” are different operations, with different SLAs and metrics.
The real problem: 2 WhatsApps without architecture become 4 pains
When you use one WhatsApp for sales and another for service, without a method, four predictable problems arise:
- Duplicate contact: the same client enters sales and then contacts after-sales; someone creates another contact and you lose the history.
- Fragmented conversation: the team doesn’t know what was promised, what was negotiated, and what has already been resolved.
- Ownership dispute: the salesperson “holds” the contact, after-sales can’t act (or vice versa).
- Garbage report: you don’t know the real CAC, repurchase rate, response time, or funnel bottleneck.
The antidote is operational architecture: one CRM, one database, two funnels, two communication lines, and routing rules. And yes: it can be done cleanly in Kommo.
Who this is for (and who it’s not for)
Serves for companies that:
- Sell via WhatsApp and have Volume enough to justify separate after-sales (e.g., clinics with recurring services, real estate, courses, industry with technical after-sales, assistance, contracts).
- Need SLA (first response time, resolution time) and tracking of commercial promises.
- Want to retain and increase LTV without inflating the team (automation + process).
Is not useful (still) for those who:
- Have low volume and only one operator: separating WhatsApps will only add friction and cost.
- No minimum CRM discipline (registration, stages, activities). If the team “hates CRM,” fix the basics first.
- Wants “magic automation” to compensate for a bad product or bad service. CRM doesn’t fix broken operations.
The golden rule: 1 contact, 1 history, N journeys
The client is one. What changes are the journeys (sales, onboarding, support, renewal, referral). In Kommo, this means:
- A single contact (with normalized phone and standardized fields).
- Two pipelines (e.g., “Sales” and “After-sales/Retention”).
- Two WhatsApp numbers connected to Kommo (via official WhatsApp API/Cloud API with BSP — this prevents bans and allows scaling).
- An ownership policy: who owns the contact and who owns the deal in each funnel.
If you want to deepen what changes when you scale WhatsApp (cost and operation), I also recommend reading: Kommo + WhatsApp API: the real cost (in BRL) and what blocks ROI when you scale customer service.
Recommended architecture (the design I implement)
I’ll give you a practical model that works in 80% of scenarios. Adjust later by niche.
1) Two funnels (pipelines) with different goals and metrics
Pipeline 1: Sales (goal: close and receive)
- New lead
- Qualified
- Proposal sent
- Negotiation
- Won / Lost
Pipeline 2: After-sales / Retention (goal: activate, resolve, retain, expand)
- Welcome / Onboarding
- In progress (request)
- Pending client
- Resolved
- Repurchase / Upsell
Sales measures conversion rate and cycle. After-sales measures SLA and retention. Mixing everything in one funnel is the fastest way to destroy visibility.
2) An “owner” policy (no fights and no blocking service)
I separate it like this:
- Contact owner: usually held by an “Account” role (e.g., account manager) or can stay with the original salesperson. The important thing is that it’s a rule, not a discussion.
- Deal owner: changes by pipeline. In the sales funnel, it’s the salesperson/SDR. In the after-sales funnel, it’s the CS/support team.
This avoids the classic: “don’t touch that client, they’re mine.” The client belongs to no one. The client belongs to the company.
3) Two connected WhatsApps: what is fact and what is illusion
Fact: it’s possible to operate 2 numbers within Kommo and organize by queue/team.
Illusion: thinking two numbers alone solve the process. Without rules, it becomes double noise.
Real data: Kommo has a paid plan per user and is usually charged in US$ per user/month (with variation by plan and monthly/annual billing). To check updated prices and which license fits your scenario, I detailed here: How much does Kommo CRM cost? Prices and plans in 2026. And if your pain is choosing a license without falling into a trap, here: Kommo CRM licenses: Which is the best for your company?.
Now, the sensitive part: WhatsApp API (official) has its own costs (conversation/message according to current Meta and your BSP rules). I won’t guess numbers here because it varies by category and country and is updated. What I do is: I include it in the ROI spreadsheet the cost per conversation + cost per user + automation cost and I make this explicit in the project.
4) Routing: how to avoid duplication when the client changes WhatsApp
What prevents duplication is not “goodwill.” It’s system rules + registration standards.
Practical checklist I implement:
- Normalize phone (country code + area code + number) and block creation without standard.
- Phone search rule: when a new message arrives, Kommo needs to identify an existing contact before creating another.
- Deal creation by channel/pipeline: if the message came from the after-sales WhatsApp, create/activate deal in the after-sales pipeline (not sales).
- If there is already an open deal in the same pipeline, don’t create another; just reopen, move stage, or create activity.
This requires carefully configuring automations to avoid “deal spam.” If you want automation and follow-up without annoying leads, also see: Automatic sales follow-up: what it is, why almost everyone does it wrong, and how to set it up.
5) Minimal automation that delivers ROI (no overengineering)
When I talk about ROI, I’m not talking about “looking nice.” I mean reducing time and increasing revenue with control.
In the Sales funnel, automations that almost always pay off:
- First response SLA: if a lead comes in and no one responds within X minutes, alert + redistribution.
- Follow-up by stage: proposal sent without response in 24h/48h? create task and suggest message.
- Funnel hygiene: lead stalled for X days returns to a stage or falls into a reactivation queue.
In the post-sale funnel, automations that almost always pay off:
- Welcome + onboarding (with checklists and data collection).
- Customer pending reminder (without relying on the operator to remember).
- NPS/CSAT and trigger referral request when resolved.
- Repurchase: date-based triggers (D+30, D+60, D+90) based on purchase/contract type.
Repurchase and retention are where the cash flow appears. I detail this strategy here: Kommo CRM for retention and repurchase: how to turn post-sale into revenue (without expanding the team).
6) Decision table: 1 funnel vs 2 funnels (objective criteria)
| Criterion | Only 1 funnel | 2 funnels (sales + post-sale) |
|---|---|---|
| Volume of customer service | Low, small team | Medium/high, needs queue and SLA |
| Repurchase/recurrence | Almost nonexistent | Important for LTV and margin |
| Support complexity | Simple, few doubts | Technical, deadlines, pending issues, tickets |
| Risk of commercial promises | Low | High (contracts, delivery, SLA) |
| Required metrics | Conversion and cycle are enough | Needs SLA + resolution + churn + upsell |
7) The “trick”: when post-sale should see the sales funnel (and when not)
Total transparency seems nice, but sometimes it’s noise. My practical rule:
- Post-sale should see what was sold, deadlines, conditions, and key messages (to avoid making different promises).
- Post-sale does not need to operate the sales funnel. Each team in its own funnel, with its goals.
- Sales should see onboarding status and satisfaction, because this becomes social proof and upsell.
This is governance: who can see, who can edit, who can move stages. Without this, the CRM becomes “no man’s land.”
8) How to measure ROI of this design (without guesswork)
If you run 2 WhatsApps and 2 funnels, you start measuring what was previously invisible:
- 1st response time by channel (sales vs post-sale).
- Resolution rate and average resolution time.
- Repurchase/upsell revenue originated by post-sale (without depending on the salesperson).
- Reduction of rework (less “send it again,” less “what was agreed?”).
Automation and CRM ROI must be a closed account: time saved + incremental revenue – costs. If you want the method to measure this properly, without self-deception, here: How to measure the ROI of business automations.
Mistakes I see blocking 2 WhatsApps in Kommo (and how to fix them)
- Error 1: creating “Sales Contact” and “Post-sale Contact”. Fix: contact is unique; use tags/fields and separate deals.
- Error 2: post-sale pipeline with sales stages. Fix: stages must reflect real operation (SLA, pending issues, resolution).
- Error 3: automation that creates a deal every time a message arrives. Fix: check if an open deal already exists; if so, just update/reactivate.
- Error 4: no ownership policy. Fix: define contact owner and deal owner by pipeline.
- Error 5: wanting “AI” before process. Fix: first standardize stages, SLA, and fields; then apply AI where it adds value.
Conclusion: two WhatsApps is not “more tools,” it’s more process
When you implement two WhatsApps in Kommo the right way, you build a pipeline: sales closes, post-sale activates and retains, and the history stays intact. This reduces internal friction, increases satisfaction, and opens space for repurchase with method. Without architecture, you just double the noise.
If you want me to design and implement this in your operation (from funnel to routing, automations, and metrics), the next step is straightforward: request a project.
FAQ
1) Can you use two WhatsApp numbers in Kommo at the same time?
Yes, you can. The critical point is not “connecting,” it is routing each conversation to the right pipeline and prevent duplication of contact/deal.
2) How to avoid duplicate contacts when the client contacts sales WhatsApp and then support WhatsApp?
Standardize phone (country code + area code + number), enforce this pattern in registration, and set rules for the system to search for existing contact before creating a new one. The correct design is one contact, multiple deals per journey.
3) Is it better to have a single funnel with post-sale stages after “Won”?
Almost never. Mixing sales and post-sale in one funnel destroys conversion tracking and doesn’t create SLA metrics. The most efficient is two pipelines with separate goals and reports.
4) Who should own the contact: sales or post-sale?
Depends on your model, but it must be a rule. I usually separate: Contact with an “account” owner (or salesperson) and deal with the pipeline owner (sales in the sales pipeline; CS/support in the post-sale pipeline).
5) Does this work for clinics, real estate, and B2B industry?
It works, as long as there is volume and real need for post-sale/retention. In small operations, separating WhatsApps can be overkill and reduce productivity instead of increasing it.