By Luiz Otávio Gonçalves Published 2026-09-08

How Kommo CRM works in practice: from lead to closing (honest overview)

How Kommo CRM works in practice: from lead to closing It’s basically this: the lead enters through channels like WhatsApp, form, or Instagram, and automatically lands in a visual funnel, customer service happens in a Unified chat (with automations, Salesbot, and AI assisting), and the manager sees the pipeline and bottlenecks in real time. Now the honest part: Kommo is simple to understand and tricky to operate well. It doesn’t deliver a ready-made process — and that’s where 90% get stuck.

I am Luiz Otávio Gonçalves. I come from Mechanical Engineering and became a pioneer of Digital Engineering in Brazil, implementing CRM and automation focused on execution and ROI. What you’ll read here is the same “operational tour” I give during client onboarding: the system running, end to end, from lead to closing — without generic definitions of “what CRM is” and no talk of “plans and pricing.”

Who this article is for (and who it’s NOT for)

It’s for you if:

• You’ve heard of Kommo, maybe even opened an account, but didn’t understand the flow from lead to closing.
• You want to see Kommo as Launch: channels → pipeline → customer service → follow-up → management.
• You’re unsure if Kommo can handle your volume and customer service routine (especially WhatsApp).

It’s NOT for you if:

• You want a CRM that already comes with a ready-made sales process, a closed and predictable playbook. Kommo is not that.
• You want to “install and start using” without mapping steps, responsibilities, SLA, and cadences. Without process, you just create mess faster.
• You’re only looking for price. For that, I have specific pages: How much does Kommo CRM cost? Prices and plans in 2026 and Kommo CRM licenses: Which is the best for your company?.

The operational view: Kommo is a “conversation + pipeline system”

If I had to sum up Kommo in one operational sentence: it organizes conversations around deals within a visual funnel — and you automate what can be automated, without pretending everything can be automated.

The most common mistake is treating Kommo as “a place to register clients.” In practice, Kommo performs when you treat it as:

• Lead intake center (WhatsApp, Instagram, forms, etc.).
• Customer service center (unified chat, templates, distribution, SLAs).
• Process engine (funnel, rules, automations, tasks, follow-up).
• Management view (pipeline, conversion by stage, time in stage, bottlenecks).

Step 1 — How the lead enters Kommo (WhatsApp, form, Instagram)

Kommo was made to “converse with leads” at scale. So, the start of the flow is: capture + identification + routing.

Most common sources in practice:

• WhatsApp: the lead calls, clicks the link, responds to an ad, or comes from a QR Code. Kommo creates/updates the contact and immediately pulls the conversation into the chat.
• Forms: website, landing page, Typeform/Google Forms (usually via integration). When the form arrives, you create the lead in the funnel and trigger the cadence.
• Instagram: direct messages can become centralized conversations (depending on the connector/channel used). The point is: the seller isn’t hostage to a thousand tabs.

What I configure during onboarding to avoid chaos:

• Minimum fields: name, phone (with country/area code), source, product/service of interest, city/state (if necessary).
• Duplicate rules: WhatsApp tends to create duplicates if you don’t standardize phone numbers. Bad data becomes misleading reports.
• Source tags: “Organic IG,” “Meta Ads,” “Website,” “Referral.” Without this, you can’t measure ROI.

By the way, if you want to measure ROI of automation and CRM in a mature way (without guesswork), I explain here: How to measure the ROI of business automations.

Step 2 — How the lead lands in Kommo’s visual funnel (pipeline)

The lead entered, Kommo needs to answer two questions:

• Which funnel does this lead enter? (e.g., “New Leads,” “Renewal/Retention,” “Partnerships”).
• At which stage does it start? (e.g., “New,” “Awaiting Qualification,” “Scheduled,” “Proposal,” “Closed”).

The visual funnel is where Kommo becomes “easy to understand.” You see cards moving. But here lies the trap: stage is not decoration. Stage is an operational decision.

Example of a funnel that usually works (generic but realistic):

• New lead (entered, no one has touched it yet)
• In service (first conversation happening)
• Qualified (has pain, profile, and timing)
• Proposal sent
• Negotiation
• Won / Lost

What unlocks results: define entry and exit criteria per stage (e.g., only moves to “Qualified” if there’s a budget or confirmed decision-maker; only moves to “Proposal sent” if the proposal was actually sent and recorded).

Step 3 — How Salesbot works (and where it really helps)

Salesbot is Kommo’s “automatic attendant” to do the work that doesn’t need a human: screening, repeated questions, data collection, routing, immediate response outside business hours.

In practice, I use Salesbot for:

• Reception: “Hi, I’m the assistant. Do you want X, Y, or Z?”
• Quick qualification: name, city, goal, urgency, budget range (when it makes sense).
• Routing: sends to seller A/B according to product, region, queue, or schedule.
• Automatic follow-up: if the person disappears, you trigger reminders with rules (no dumb spam).

Where the Salesbot DOES NOT solve:

• Complex negotiation.
• Consultative sales (B2B with multiple decision-makers).
• Cases with many exceptions (“it depends”, “let's design”, “I need to evaluate”).

If you want follow-up that really closes (and not just annoys), read this: Automatic sales follow-up: what it is, why almost everyone does it wrong, and how to set it up.

Step 4 — How AI agents come into play (no fantasy)

When I talk about AI in Kommo, I don't sell miracles. I implement AI where there is Volume, standard and minimum data. A good AI agent is not the one that “seems smart.” It's the one that reduces response time, increases qualification rate and takes repetitive tasks away from humans.

Practical applications (that I see working):

• Pre-service: answer frequent questions and collect information before the seller steps in.
• Classification: mark interest, temperature, reason for contact, origin (when well configured).
• Seller's assistant: suggest next message, summarize long conversation, draft proposal.

Real prerequisite: you need a minimal process and a minimally organized base (fields, stages, tags). AI on top of mess only automates mess.

If you want a down-to-earth view of what automates and what still needs a human, here it is: AI Agent and AI SDR for sales: what really automates (and what still needs humans).

Step 5 — How the seller works in the unified chat (the heart of Kommo)

Kommo's “trick” is the unified chat: WhatsApp, Instagram, and other channels can all land in one place, connected to the contact and the pipeline deal.

Real sales flow inside the chat:

• The lead arrives and enters “New lead.”
• The attendant/seller takes over (manual or automatic via rule).
• He qualifies (right questions, not interrogation).
• He moves the stage card when the criterion is met (this is operational data).
• He schedules the next step (task, reminder, follow-up, meeting).
• He records the reason for loss when he loses (without this, you don't improve the process).

What I standardize to increase conversion without “training by motivation”:

• Message templates by stage (opening, qualification, proposal, return follow-up).
• Internal SLA: maximum time for first response and maximum follow-up time.
• Next mandatory action: no one closes the chat without scheduling the next step (task or stage).

Step 6 — How the manager reads the pipeline (and sees bottlenecks without guesswork)

Management in Kommo is not “watching cards move.” Management is looking at conversion rate, time in stage, reasons for loss and team capacity.

What I expect from the manager in routine:

• How many leads came in by source (and which source converts).
• Which stage is becoming a “cemetery” (many leads stuck).
• Average time to first response (especially WhatsApp).
• Standardized reasons for loss (to attack the root, not the symptom).

When this becomes routine, reporting stops being bureaucracy and becomes decision-making. If you want to automate reports and truly save time (without losing trust in the data), I go deeper here: Report automation: how to gain strategic time.

The Kommo “map” in practice (from lead to closing)

Here is a simple, operational map I use to align the team and remove noise:

The thesis (no fluff): Kommo is simple to understand and tricky to operate well

If you read this far, you already got my main point: Kommo doesn't give you process. It gives you:

• a great visual funnel,
• a very strong unified chat,
• automation (Salesbot) and space for AI,
• and integrations for you to connect the rest.

What it DOES NOT give you is: well-designed stages, criteria, SLA, cadence, data governance, management rituals, and message standards. This is design. And it's exactly where 90% of operations get stuck: the tool may be “installed,” but the process hasn't been built.

A real fact (so you don't decide blindly): Kommo's price in US$

I won't make up numbers. Kommo is sold by user subscription and, in the international public table, plans appear at US$ (recurring charge). In practice, the amount in R$ will vary by exchange rate, taxes, and payment method. If you want to see updated prices and an honest reading of total cost (licenses + implementation + integrations), go straight to: How much does Kommo CRM cost? Prices and plans in 2026.

Quick checklist: “first steps” to get Kommo running in 7 days

If you want to start from zero and have something usable (without paralyzing perfectionism), follow this order:

• Day 1: define 1 funnel (just one) and 6–8 stages with exit criteria.
• Day 2: connect 1 main channel (usually WhatsApp) and validate contact/lead creation.
• Day 3: create minimum fields + origin tags.
• Day 4: message templates by stage + first response SLA.
• Day 5: basic automation: create a follow-up task if there is no response within X hours.
• Day 6: standardized loss reasons + pipeline update routine.
• Day 7: review with the team: what got stuck, what became a bottleneck, what will be adjusted.

Want to do this faster and with less rework? That's where guided implementation comes in. Because you buy the tool; you build the operation.

If you want me to implement this flow with you (from lead to closing), with automation and data governance, my path is simple: request a project.

FAQ — real questions about how Kommo works in practice

1) Is Kommo suitable for WhatsApp-centered operations?
Yes, and it's one of the scenarios where it is most used. But the result depends on the process: distribution, SLA, cadence, and standardization. Without that, it just becomes “WhatsApp with labels.”

2) Do I need to have a perfect funnel before starting?
No. You need a funnel usable with minimum criteria. The perfect funnel comes later, with data (time in stage, conversion, loss reasons).

3) Does Salesbot replace the salesperson?
No. Salesbot speeds up screening, collects data, and ensures quick responses. A salesperson is still necessary to properly qualify, negotiate, and close — especially in consultative sales.

4) Why does my Kommo “not deliver results” even with leads coming in?
Almost always it's a lack of process: stages without criteria, no one owns the lead, inconsistent follow-up, dirty data, and no management routine. The tool alone doesn't fix this.

5) What is more important: automation or team training?
Both, but in the right order: first process and standard (what to do, when, and why), then automation to reduce friction, and training for consistent execution. Method > improvisation.