How long does CRM implementation take for a small business in Tashkent?

Guide: 4–8 weeks — audit 3–5 days, funnel setup 5–10 days, integrations 1–3 weeks, pilot 2 weeks. More channels, branches and custom logic extend the timeline, as does a process owner slow to approve stages.

Should we migrate the entire old client base?

No. Migrate active contacts and open deals — and if needed, active clients from recent months. Cold Excel lists often pollute CRM, create duplicates and hurt reports.

Which integration to start with in Uzbekistan?

Usually Telegram and the website form — the main lead entries. Add Instagram and telephony after those two are stable. That keeps the pilot from ballooning.

Why does CRM not stick with managers?

Usually no process owner, an over-complex funnel, no short playbook, no training on real leads, and notifications that never arrive. Simplify stages, fix time-to-first-reply, and show that the system removes chaos rather than adding bureaucracy.

What budget should we plan for implementation?

Soft Tashkent guide: box setup and training — from a few million UZS / about $300–600 one-off; custom and complex integrations — higher. Support is separate by change volume. Not a UZNEO price list.

Can we implement CRM without stopping sales?

Yes — via a 1–2 week parallel pilot: old tracking stays alive until every channel writes into CRM reliably. A hard full cutover in one day often breaks discipline more than it saves time.

How is a box product different from a turnkey CRM?

A box product is faster and cheaper when the process is typical. Turnkey fits when you have several lines, non-standard stage logic, or a deep bot and website link. System comparison is in the AmoCRM / Bitrix24 / custom CRM article.

CRM19 minJuly 10, 2026

CRM implementation: stages, timeline and typical mistakes

Robert MirovUpdated Jul 10, 2026

The owner opens the chats: leads from Instagram, Telegram, the website and phone calls live in different places. A manager replies “when they can”; deal history sits in someone’s head or in Excel. A month later nobody can say how many leads reached payment — or where they dropped. A CRM (Customer Relationship Management — a system to track clients and deals) is not a “pretty app”. It is a single source of truth: who came, from where, at which stage, who owns the lead, and what happens next. CRM implementation is a process change, not installing a license. Without a lead map, a short playbook and training, the system becomes an expensive address book. Below is a calm order for small and mid-size business in Tashkent and other cities in Uzbekistan: audit → one funnel → migration and integrations → roles → pilot → KPIs. Timelines for 1–5 managers, role and Before|After tables, checklists, soft budget ranges. No “payback in a week” promises — and no panic over typical mistakes; fix them one junction at a time.

01Why CRM if “Excel and chats are enough for now”

Spreadsheets and private chats work while volume is low and one person holds everything in their head. Add a second manager, night leads from Telegram, or ads across several channels — and chaos gets expensive faster than it looks.

Before	After
Lead in a private chat with no number or status	Card with source, stage and owner
“Who took it?” in voice notes and screenshots	Assignment rule + notification
Deal history in the manager’s head	Notes and stages in one card
End-of-month “gut feel” report	Funnel: reply time, conversion, loss reasons
Ads “seem to work”	Source and UTM from the first touch

CRM does not replace a salesperson on complex negotiations. It cuts losses at the entrance and gives the owner facts instead of feelings. If lead volume is low and one person replies within an hour, you can wait. If you already lose leads between channels, rollout pays back through saved deals — not “software magic”.

02Stage 1: audit and process map

Before choosing a CRM, document how a lead moves from first contact to payment. In a team session, answer honestly — not “how it should be”, but how it works today. • where leads come from (Telegram, Instagram, website, calls, referrals) • who handles first contact and how fast • which deal stages actually exist vs. only on paper • where clients, invoices and chat history live today • what happens on a loss: do you record a reason, or just “gone” Audit checklist (1–2 meetings): 1. Channel list and weekly lead volume (even a rough estimate). 2. Current average time to first reply — real, not “we try”. 3. Where leads get lost: chat, Excel, “in someone’s head”, between shifts. 4. Which fields managers need day to day vs. owner reports only. 5. Who decides on the funnel (process owner — the person who approves stages and KPIs). 6. Which integrations are critical from day one, and which can wait. Output — a 1–2 page process map. Without it you cannot configure the funnel honestly: the vendor will guess, the team will argue in week three. UZNEO often includes CRM consulting here. Audit takes 3–5 business days and does not require buying licenses.

03Stage 2: CRM choice and one funnel design

Based on the map, choose the system and design one primary funnel. Two funnels “for every case” on day one almost always hurt: managers mix statuses, reports fall apart.

Element	Guide	Why
Stages	4–7	More — managers skip statuses
Names	Team language	“Invoice sent”, not “MQL”
Required fields	Minimum	Otherwise cards stay empty
Lost deals	Status + reason	Otherwise analytics fail
Assignment	Rules or manual	Clear lead owner
Reply SLA	Hours, not “asap”	You can measure and improve

Practical rule: a stage changes only on an event — meeting booked, proposal sent, prepayment received, contract agreed. If a stage can be dragged with no reason, the funnel lies. For non-standard logic (multiple lines, B2B and retail in one flow), a turnkey CRM can beat fighting a box product. Compare AmoCRM, Bitrix24 and custom — separate article.

04Stage 3: data migration and integrations

Migrate only active contacts and open deals. Old “cold” Excel lists often pollute CRM: duplicates, dead phones, reports nobody trusts. Typical integrations for Uzbekistan: • Telegram bot → lead in CRM • website form → webhook (automatic data handoff) • Instagram Direct (intermediary or manual protocol at start) • IP telephony / click-to-call if you dial out • if needed — Google Sheets as a backup journal during the pilot Checklist before go-live: 1. A test lead from each channel lands in CRM within 1–2 minutes. 2. Source and owner are filled automatically or by a short rule. 3. Resubmitting the form does not spawn duplicates. 4. The manager sees a notification (bot / email / task). 5. Fallback exists: manual entry per playbook if the webhook fails. 6. Access rights: a manager cannot delete someone else’s deals “by mistake”.

Before	After
Website lead in someone’s inbox	CRM lead with source “website”
Telegram in a manager’s private chat	Bot / rule → card + notification
Instagram “screenshot to the group”	Protocol: who enters it and within how many minutes
Resubmitted form = second identical lead	Dedup or an explicit link

More detail — CRM integration with Telegram, Instagram and website.

05Stage 4: training and a one-page playbook

Training is not a single webinar titled “look at the UI”. You need a one-page playbook you can pin at the desk: 1. New lead → first contact within X hours (or minutes if the flow is hot). 2. Every call / chat → note in the card the same day. 3. Stage change → only when a condition is met (invoice, meeting, prepay, loss with reason). 4. Deal close → win or loss reason from a short list. 5. If an integration is silent → manual entry + signal to the CRM admin. Run 2–3 short sessions: demo → practice on real leads → review after a week. Without sales automation (bots, reminders for “stale” leads), discipline drops fast: people return to familiar chats. A 20-page document nobody opens does not work. One page plus in-system reminders works more often.

06Project roles: who owns what

CRM is rolled out by a team, not “IT alone” and not “the vendor will figure it out”. Without a process owner the project often stalls at stage approval.

Role	Usually who	Responsibility
Process owner	Owner / sales director	Funnel and KPI decisions
CRM admin	Trained manager	Fields, rights, reports, small tweaks
Vendor	External team	Integrations, migration, launch support
Managers	Sales team	Daily use and feedback

Block 2–3 hours per week from the owner for the first 4 weeks: approve stages, review pilot leads, stop the “we need five more statuses” debate. Give the CRM admin a clear channel to the vendor — otherwise every field fix goes through the owner. Involve managers early: show that the system removes chaos, not “adds reporting for reporting’s sake”.

07Timeline, first-month KPIs and budget ranges

Guide for small business (1–5 managers):

Stage	Time	Output
Audit and map	3–5 business days	1–2 page process
Funnel and roles	5–10 days	Stages, fields, rights
Integrations	1–3 weeks	Test leads from channels
Pilot	2 weeks	Parallel tracking without gaps
Full cutover	usually 4–8 weeks	CRM = single source of truth

The timeline grows with more channels, custom logic, several branches, or a process owner who is slow to approve stages. First-month KPIs (review weekly): • share of leads contacted on time (SLA) • average time to first reply • conversion by funnel stage • share of deals with source filled • share of losses with a stated reason “Closed deals” alone, without reply time and source, is a weak KPI at launch: you may close fewer deals but understand the funnel better. Soft budget range (Tashkent guide, not a UZNEO price list): • box setup + training — from a few million UZS / roughly ~$300–600 one-off • custom / complex integrations — higher; depends on channels and logic • support — from ~$100–300/mo if you need tweaks and reports An over-complex funnel and no process owner are common reasons the system does not stick. Not “failed forever” — a signal to simplify and return to the process map.

08Parallel pilot, rollback plan and typical mistakes

For 1–2 weeks of pilot do not shut down old tracking completely. Run CRM parallel to a spreadsheet or chat until every lead lands reliably. A hard “switch on Monday” often creates panic and a quiet return to Excel. Rollback: integration breaks → manager enters the lead manually per playbook → tech fixes the webhook. Main rule — the lead is not lost. After two weeks with no misses, CRM becomes the single source of truth.

Before	After
Friday “we switched to CRM”, Monday chaos	1–2 weeks parallel, then cutover
Webhook down — lead vanishes	Manual entry + ticket to admin
Mistakes pile up for a month	Weekly pilot review
All channels at once	Telegram + website first, rest later

Typical mistakes checklist (calm to fix one by one): • no process owner; • 12+ stage funnel “like a holding company”; • Telegram not connected — leads in private chats; • migrating the entire old Excel “just in case”; • one training session and no playbook; • KPIs only “deal count”, no reply time or source; • rights set to “everyone can do everything” — cards get broken or deleted; • expecting CRM itself to teach the team how to sell. Fix one junction per week — that is how rollout sticks. You do not need to rewrite the whole system because of one Instagram gap.

09What to prepare before you start with a vendor

Most delays are not “the developer is slow” — they are an undescribed process and missing funnel decisions. Owner checklist before the brief: 1. Lead channels and approximate weekly volume. 2. Draft deal stages in your own words (4–7). 3. Who will be process owner and CRM admin. 4. Whether bot, website and telephony are needed from day one — or two entries are enough. 5. What to migrate from the old base: open deals only, or also active clients from the last N months. 6. Languages managers and clients use (matters for stage names and templates). 7. What “success in 30 days” looks like — e.g. 95% of leads in CRM, reply ≤ 2 hours in working hours. The clearer the brief, the shorter the timeline and the fewer rewrites in week five. UZNEO usually starts with the process map, not with “which logo is on the screen”.

Summary

Successful CRM rollout starts with the process, not the invoice logo. Run an audit, configure one clear funnel, connect Telegram and the website, train the team on a short playbook, run a parallel pilot, and watch KPIs from week one. Mistakes at the start are normal — pretending “the software will figure it out” is worse. Simplify stages, name a process owner, and fix integrations one at a time. UZNEO supports implementation — from bot development and sales automation to a full CRM system. Message us on Telegram for a timeline based on your lead volume and channels.