Business automation → CRM implementation

Twenty CRM and amoCRM implementation

We implement CRM for IT companies, studios and B2B services in Kazakhstan: Twenty CRM on your own servers or amoCRM for messenger-driven sales. Pipeline, fields, import, task tracker integration and an AI layer around sales.

Discuss your project

Overview

CRM implementation: the service in short

Gotomarket moves sales out of spreadsheets, email and WhatsApp into a CRM where every deal has a stage, an owner and a next step. Our main tool is Twenty CRM: an open source system that can run on a server in Kazakhstan with no per-user fee. For teams that sell mostly through messengers, we implement amoCRM. We connect the CRM with the website, the Plane or Jira task tracker and AI assistants, so a won deal becomes a project right away and client context is never lost.

Launch: 4–6 weeks
CRM: Twenty, amoCRM
Hosting: servers in Kazakhstan or cloud
Integrations: website, tracker, messengers, AI

In plain words

What it is and why it matters

What it is: We move sales out of spreadsheets, email and WhatsApp into one client management system (CRM), where every deal has a stage, an owner and a next step.
Who it is for: IT companies, studios and B2B services that lose leads between email, the website and messengers, and hand won deals over to development by hand.
What you get: A configured Twenty CRM on your server or amoCRM, migrated clients and deals, links to your website, messengers and development task system, management reports and working rules for sales managers.
Where to start: Tell us where your leads and deals live today and how many people sell. We will map your pipeline (the sales stages) and suggest which CRM fits.

Solution architecture

How we implement CRM with the 5D methodology

We first describe how the company actually sells and only then configure the system. We launch with one team and scale once the pipeline works.

Discover

Map the pipeline

We record lead sources, deal stages, roles, loss reasons and where client data lives today.

Design

Design the model

We choose Twenty or amoCRM and define objects, 5–7 stages, required fields, permissions and integrations.

Deploy

Configure and migrate

We roll out the CRM, import clients and deals, and connect the website, email, messengers and task tracker.

Drive

Run the pilot

The sales team works in the CRM for 2–3 weeks. We review deals with no next step and fine-tune the pipeline.

Dominate

Add AI and reporting

We build management dashboards, connect AI assistants and assign a system owner.

Unit economics and ROI

How a CRM pays off in an IT company

Without a CRM, leads get lost between email and messengers, and client history leaves with the sales manager. Leadership cannot see where deals stall or why they are lost.

A CRM shows stage conversion, time in stage and deals with no next step. Self-hosted Twenty removes the per-user fee, so cost does not grow with the team. The actual effect depends on lead flow and the discipline we establish during the pilot.

Lost leads: reduced
Client history: in one place
Twenty licenses: zero when self-hosted
Sales forecast: based on data

Result

What you get

The implementation includes not only a configured system but also the rules the sales team follows in it every day.

A pipeline and role map
Twenty CRM on your server or a configured amoCRM
Objects, stages, fields and permissions
Clients and deals imported from spreadsheets or an old CRM
Integrations with the website, email and messengers
A link to the Plane or Jira task tracker
Management pipeline dashboards
Operating rules, training and a system owner

Comparison of approaches

Twenty, amoCRM or a spreadsheet

The choice depends on how you sell, where data must be stored and who will maintain the system.

Approach	What happens	Limitation	When it fits
Self-hosted Twenty CRM	Open source CRM on your servers with custom objects, API and webhooks	Someone must own the server. Telephony and campaigns come via integrations	IT companies and B2B with long sales cycles and data requirements
Twenty Cloud	The same CRM hosted by the vendor	Data is stored outside Kazakhstan	Small teams without data residency requirements
amoCRM	Cloud CRM with strong WhatsApp, Telegram and Instagram channels	Cloud outside Kazakhstan, per-user pricing, less flexible data model	High inbound lead flow through messengers
Spreadsheet and chats	Deals in Google Sheets, conversations in WhatsApp	No history, stages, control or reports	A founder's first 10–20 clients

Limitations and risks

When a CRM implementation will not work

A CRM reinforces a sales process. Without one, the system becomes another spreadsheet nobody updates.

No pipeline owner

Without someone who reviews deals with no next step, the data goes stale within a month.

Too many fields

Twenty required fields on day one turn the CRM into bureaucracy. We start with five to seven.

Parallel records

If deals keep living in WhatsApp and a spreadsheet, the CRM never becomes the source of truth.

A good fit

When the service fits

Leads get lost between email, the website and messengers
The sales team has no shared pipeline or forecast
Client data must stay on servers in Kazakhstan
Won deals are handed to development by hand

Not a fit

When to start with a different step

Only the founder sells and there are fewer than twenty clients
There is no clear product and price yet
Nobody is ready to own the data in the CRM
You only need email campaigns, not deal management

Terms in plain words

CRM: a system that stores the history of clients, deals and agreements
Pipeline: the stages a deal goes through from lead to payment
Open source: software with publicly available source code that you can install on your own server
Webhook: an automatic notice to another system about an event, such as a won deal
AI layer: AI assistants connected to the CRM, knowledge base and work systems
Self-hosted: A setup where the system runs on your own server rather than in the vendor's cloud, so client data can stay in Kazakhstan.
Custom objects: Your own record types in the CRM beyond the standard ones, such as account, plan or project, so the system reflects how your product works.
Lead qualification: Checking whether an incoming lead fits your product and is ready to buy, so managers spend their time on real deals.

For IT companies

CRM implementation for IT companies in Kazakhstan

For an IT company the CRM has to know not only contacts but the product model: accounts, plans, projects and the link to development. That is why our main tool is Twenty CRM with custom objects and an API, while amoCRM is the choice where sales run through messengers.

SaaS: accounts, plans, renewals

For SaaS we create custom objects in Twenty that match the product: account, plan, subscription, renewal date. Through the API and webhooks the product sends sign-ups and plan changes to the CRM, and the manager sees whose trial is ending or whose renewal is coming. New business and renewals are kept in separate pipelines so the revenue forecast does not get mixed up.

IT studio: from brief to project

Studios have a long sales cycle: brief, estimate, proposal, approval, contract. We set up stages with required fields such as scope, timeline and decision maker, and a link to Plane or Jira: a signed contract creates a project with a task template. The negotiation history stays in the client record, not in the manager's inbox.

B2B service: several decision makers

In B2B sales several people make the decision: the user, the department head, finance and the business owner. In Twenty we link the company to its contacts and their roles in the deal and record who has already joined meetings and who has not. This shows why a deal is stuck, and the client relationship is not lost when the sales manager changes.

Messenger sales: amoCRM

If the product sells at volume and leads come through WhatsApp, Telegram and Instagram, we implement amoCRM: we connect the channels and set up lead distribution between managers. Every conversation lands in a deal, and a returning client opens their existing history instead of creating a new record.

Implementation practice

How Twenty CRM and amoCRM implementation works in practice

CRM implementation starts with a pipeline map and a data model, not with interface settings. Below are the decisions we make on every project, and they determine whether the sales team still works in the system a month after launch.

Data model before any setup

First we describe the entities of your business: companies, contacts, deals, and for an IT product also accounts, plans, projects or licenses. In Twenty these become custom objects with relations, in amoCRM they become deal and contact fields. We define 5–7 pipeline stages with transition criteria and required fields at each stage. A typical mistake is twenty required fields at launch: managers start filling them in formally and the data loses meaning. We start with the minimum and add a field only when a report or a management decision depends on it. Field and stage names are written for managers, not developers.

Self-hosted Twenty on Kazakhstan servers

Twenty is an open source CRM under the AGPL-3.0 license, which we deploy with Docker on a Kazakhstan hosting provider's server or in your data center. This keeps the database in Kazakhstan, if that matters to your company, and keeps client data under your control. We configure the domain, SSL, backups, monitoring and the update procedure, and connect Google or Microsoft mail and calendar. Access is split by role: a sales manager sees their own deals, the head of sales sees the whole pipeline. The system owner gets a maintenance guide and a clear procedure for outages.

API and webhooks around the product

The main strength of Twenty for IT companies is its open API and webhooks. We connect the CRM to the website and forms, to the product itself, to the Plane or Jira tracker and to payment events such as a Kaspi payment. A won deal fires a webhook and creates a project with tasks, and closing the tasks updates the client status. For each integration we define which system owns each field, how errors are handled and who gets notified. Chains are built on n8n or with direct API calls, so the logic is visible and your developer can maintain it.

amoCRM for messenger sales

In Kazakhstan clients often write on WhatsApp rather than by email, and with a large flow of inbound leads amoCRM remains a practical choice. We connect WhatsApp, Telegram and Instagram, configure lead distribution rules and response time control. We are honest about the limits: data is stored in a cloud outside Kazakhstan, and the data model is less flexible than in Twenty, so complex links to the product are harder to build. If you need both messengers and local data storage, we consider Twenty with channel integrations.

Import, pilot and AI on top of the pipeline

We migrate clients and deals from spreadsheets and the old CRM via CSV or API: remove duplicates, map fields, move active deals and clients from the last two years in full and archive the rest. Then the sales team works in the CRM for 2–3 weeks in pilot mode while we regularly review deals without a next step. Parallel tracking in spreadsheets is switched off at once, otherwise the CRM never becomes the source of truth. Once the pipeline works, we add AI: lead qualification, conversation summaries and email drafts, with results written to the client record.

Reports and a management dashboard

We design the management dashboard around the questions leadership needs answered every week: how many leads came in and from which sources, conversion by stage, how many deals have no next step, why deals are lost and what flow of closed deals to expect. For each question we define fields and data entry rules, otherwise the report looks good but is wrong. For IT companies we also show the link to development: how many won deals are waiting for a project to start. In small companies the business owner makes the decisions, so the dashboard must be readable in a couple of minutes without anyone explaining it.

Step by step

CRM implementation in an IT company: five steps

Below is how a Twenty CRM or amoCRM implementation runs in practice: who takes part at each step and what the company keeps after it.

Sales review on real deals

We meet the owner and the head of sales and go through recent deals: where the client came from, how many touchpoints there were, where the deal stalled and why it was won or lost. We look at spreadsheets, email and WhatsApp conversations. The output is the actual pipeline map, a list of lead sources and the reasons for losses, not an ideal diagram from a sales textbook.

CRM choice and data model

Based on the review we choose Twenty or amoCRM and design the model: objects, stages, required fields, roles and permissions. For SaaS we add accounts and plans, for studios projects and estimates. The decision is agreed with the owner and the future system owner. The artifact is a data model document and an integration list naming the master system for each field.

Deployment and data import

For Twenty we deploy the system with Docker on a server in Kazakhstan, for amoCRM we set up the account and channels. We clean and import clients and deals and connect email, the website, messengers and the task tracker. On the client side an administrator or developer takes part if the product needs to be linked to the CRM via API. The output is a working CRM with real data and an import report.

Sales team pilot

The sales team works only in the CRM for 2–3 weeks. Once a week, together with the head of sales, we review deals without a next step, empty fields and awkward stages, and adjust the system right away. Parallel spreadsheets are closed on the first day of the pilot. The step ends with a corrected pipeline, first conversion data and a decision on adding AI and the other teams.

AI, reports and handover

We connect AI scenarios: lead qualification, conversation summaries, email drafts. We build the management dashboard, train managers by role and hand the working rules over to the system owner. For self-hosted Twenty we hand over backup and update instructions and agree on the support format. The artifact is a CRM that works without our daily involvement. A month after launch we check how the pipeline is being filled and simplify stages if needed.

Ways to work

CRM implementation formats in Kazakhstan

Implementing a CRM for one sales team usually takes 4–6 weeks including the pilot. Complex integrations with the website, telephony and the task tracker add 2–3 weeks.

Format	What is included	Timeline	Outcome
Twenty CRM on a Kazakhstan server	Docker deployment, custom objects and pipeline, import, basic integrations, pilot, working rules	4–6 weeks	A CRM on your servers with client data stored in Kazakhstan
amoCRM for messengers	Pipeline and fields, WhatsApp, Telegram and Instagram channels, import, manager training	4–6 weeks	Every messenger conversation linked to a deal
CRM integrations with tracker and product	Link to Plane or Jira, the website, the product and payment events via API and webhooks	Adds 2–3 weeks to the rollout	A won deal immediately becomes a project in the tracker
AI layer on top of the CRM	Lead qualification, conversation summaries, email drafts, knowledge base search	Defined after the diagnostic	AI results written to the client record

We choose the right format during a free basic diagnostic session, once the pipeline, lead sources and required integrations are clear.

Frequently asked questions

CRM implementation: questions

Which CRM do you implement?

Our main focus is Twenty CRM, an open source system for IT companies and B2B services. For teams that sell mostly through WhatsApp and other messengers, we implement amoCRM. We make the choice during the diagnostic, based on your sales process and data requirements.

What is Twenty CRM?

Twenty is an open source CRM under the AGPL-3.0 license. You can run it on your own server with Docker or use the vendor's cloud. It has custom objects and fields, workflows, an API and webhooks, and email and calendar integration with Google and Microsoft.

Twenty or amoCRM: which one to choose?

Twenty suits IT companies and B2B with long sales cycles, a flexible data model and a need to keep data in Kazakhstan. amoCRM is stronger where there is a high inbound flow from messengers. If in doubt, we start with a pipeline map and the choice becomes obvious.

Can CRM data stay in Kazakhstan?

Yes, with self-hosted Twenty on a server at a Kazakhstan hosting provider or in your own data center. The database then stays in the country and under your control. amoCRM and the cloud version of Twenty store data outside Kazakhstan.

How much does a CRM cost?

Self-hosted Twenty is free; you pay only for the server and maintenance. Twenty Cloud and amoCRM are billed per user per month, and amoCRM is sold in Kazakhstan through partners in tenge. Implementation cost depends on the import and integration scope and is fixed after the diagnostic.

How do you move clients from spreadsheets or an old CRM?

We clean and map the data, then import companies, contacts and deals via CSV or API. Active deals and clients from the last two years move first, the rest is archived. The result is checked with the sales team before launch.

Can the CRM be connected to a task tracker?

Yes. Through the API and webhooks, a won deal in Twenty or amoCRM creates a project and tasks in Plane or Jira, and closing tasks updates the status in the CRM. Sales and development see the same client context.

Where does AI come in?

Once the pipeline works, we connect AI assistants: website lead qualification, conversation summaries, email drafts and knowledge base search. AI works on top of the CRM and writes results into the client record instead of living in a separate chat.

How long does implementation take?

For one sales team, usually 4–6 weeks including the pilot. Complex integrations with the website, telephony and task tracker add 2–3 weeks.