Skip to main content
Free 20-minute marketing audit
Limited spots available

Have a question?

We usually reply within 5 minutes.

WWhatsApp TTelegram

Custom ERP Development in Uzbekistan: When You Need It and How It Is Built

When a company in Uzbekistan needs a custom ERP, how it is built step by step, what drives the cost, and how it compares with 1C and packaged software.

Custom ERP Development in Uzbekistan: When You Need It and How It Is Built

A custom ERP is not packaged software but a management system built around your company's actual processes: sales, inventory, procurement, branches, finance and staff tasks run in one database under one logic. In Uzbekistan it typically makes sense when your processes are specific, when data is lost between several tools and spreadsheets, or when adapting a global or off-the-shelf product to local reality costs nearly as much as building something fit for purpose. The sound sequence is: discovery → process mapping → MVP modules → integrations → rollout → support. Skipping a stage usually means doing twice the work later.

This guide is written for foreign companies entering or already operating in Uzbekistan — subsidiaries, distributors, manufacturers and investors. It explains each stage in plain terms, the factors that drive cost, an honest comparison with off-the-shelf options (including 1C), and the most common risks.

What an ERP is, and why spreadsheets stop working

ERP (Enterprise Resource Planning) is a system where all significant business events — goods received, goods sold, payment collected, task completed — are recorded once and linked together.

Many local businesses, including partners and distributors you may acquire or work with, run on a familiar patchwork: sales in Excel or Google Sheets, inventory in a separate sheet, accounting in 1C, orders coming in through Telegram groups, payments visible in Click or Payme merchant dashboards. Each tool works on its own. The problem lies in the gaps between them: someone re-types data, someone forgets, someone keeps "their own" version. At month-end, management sees three different numbers — and headquarters sees a fourth.

An ERP closes those gaps: data is entered once and is consistent everywhere.

When a custom ERP is the right call

A custom ERP is a significant commitment, so the first question is not "how to build it" but "do we need it at all". A full self-assessment is in signs your business needs an ERP system. The typical triggers:

Also read Your Customer Calls Twice and Still Has to Spell Their Name: Connecting IP Telephony to ERP in Uzbekistan Technology · 10 min
  • Your processes are not standard. A distributor with agent-specific pricing, bonuses and return terms, for example, can only model that in packaged software through workarounds.
  • Many systems must be connected. POS, e-commerce, marketplaces, Telegram bots, local payment systems and the e-invoicing operator all need to feed one database.
  • You are scaling branches. A retail chain needs real-time, central visibility of each branch's stock, inter-branch transfers, stock counts and cash.
  • Customising a global or packaged product is getting out of hand. Every new requirement means another module, consultant or licence — and it still doesn't fit.
  • Headquarters cannot get reliable numbers without chasing people. This is closely tied to management accounting for owners (P&L, cash flow); without a clear management reporting model, even a good ERP delivers only part of its value.

If none of these apply, an off-the-shelf solution is probably the right choice for now — and that is a perfectly sound decision.

A pharmacy warehouse worker picks an order and confirms it with a scanner

Custom ERP, 1C and off-the-shelf software: a neutral comparison

There are three main routes, and each has its place. The full analysis is in custom ERP vs 1C vs off-the-shelf software; here is the short version:

  • Local statutory accounting and tax — 1C (configuration + customisation): Strong, adapted to local requirements; Off-the-shelf SaaS / global ERP: Often requires local add-ons or a separate system; Custom ERP: Usually via integration with 1C or local accounting software
  • Time to launch — 1C (configuration + customisation): Medium; Off-the-shelf SaaS / global ERP: Fast for SaaS; long for global ERP rollouts; Custom ERP: Phased, first module within months
  • Fit with specific processes — 1C (configuration + customisation): Possible, but cost grows with complexity; Off-the-shelf SaaS / global ERP: Limited to product capabilities; Custom ERP: High — built around the process
  • Availability of specialists locally — 1C (configuration + customisation): Wide; Off-the-shelf SaaS / global ERP: Vendor or partner dependent; Custom ERP: Depends on the building team and documentation quality
  • Local integrations (bots, marketplaces, payments) — 1C (configuration + customisation): Possible, often with extra work; Off-the-shelf SaaS / global ERP: Limited to existing connectors; Custom ERP: Designed for your needs
  • Ownership — 1C (configuration + customisation): Licence model; Off-the-shelf SaaS / global ERP: Subscription, data on vendor servers; Custom ERP: Code and database owned by your company (per contract)

When 1C is the right choice: if your core need is statutory accounting and tax reporting, your processes are close to typical, and you work with a strong 1C specialist. Local accountants know it, and it tracks local requirements. For many foreign-owned companies the most practical architecture is hybrid: 1C for statutory books, a global system for group reporting where relevant, and a custom operational layer for sales, inventory and branches — all synchronised via integrations.

When off-the-shelf fits: small scope, simple processes, and speed matters more than fit.

When custom ERP fits: when your processes are part of your competitive advantage and bending them to a product would cost you.

How a custom ERP is built: 6 stages

Each stage is covered in detail in ERP implementation stages. Here is why each matters and where projects typically go wrong.

1. Discovery: defining the real problem

Discovery is research before any code is written. An analyst interviews management, department heads and front-line staff, and studies existing spreadsheets, reports, Telegram groups and documents.

Outputs:

  • a description of the current state (what data lives where, who enters it, who uses it);
  • a list of key pain points and their business impact;
  • project boundaries: what comes first and what is deferred.

The most common mistake is one meeting with management followed by "let's build". Actual processes often differ from what leadership — and especially a remote headquarters — believes: the warehouse receives goods in a different order, cashiers apply discounts their own way. You only see this from inside the process.

2. Process mapping

Each core process — order to cash, receipt to sale — is mapped step by step: who does what, which document is created, where approval is required.

The map does two things. It exposes unnecessary steps; sometimes half the problem is solved by changing the process, not by software. And it becomes the basis for the specification — see writing an ERP technical specification.

Rule of thumb: automate a bad process and you get a bad process that runs faster.

3. MVP: the first working modules

The MVP (minimum viable product) is the first version that solves the most painful process. Instead of building everything at once, one or two modules go live, staff start using them, and the next step is planned from real feedback.

Typical MVPs by business type:

  • Retail chain — Typical MVP modules: Catalogue, branch stock, sales, transfers; Next phase: Stock counts, staff KPIs, loyalty
  • Distributor — Typical MVP modules: Orders, agents, pricing policy, receivables; Next phase: Routes, delivery, bonus calculation
  • Manufacturing — Typical MVP modules: Raw-material stock, bills of materials, production orders; Next phase: Costing, planning, quality control
  • Service company — Typical MVP modules: Clients, orders/projects, tasks, payments; Next phase: Resource planning, profitability analysis

Industry specifics: ERP for retail chains, ERP for manufacturing and warehouse inventory automation.

4. Integrations

In Uzbekistan an ERP almost never works alone. Typical integrations:

  • Local accounting (most often 1C) — exchange of documents and balances.
  • Electronic invoicing operators (systems such as Didox or Faktura.uz) — creating, sending and tracking e-invoices, which are a routine part of B2B operations locally.
  • Local payment systems (Click, Payme and others) — automatic payment matching.
  • Telegram bots — Telegram is a primary business messenger in Uzbekistan, so order intake, staff tasks and daily management reports often run through it. See ERP + Telegram bot.
  • CRM and marketing — the chain from lead to sale. See ERP and CRM integration.
  • Marketplaces, POS, e-commerce — stock and order synchronisation.
  • Group systems — consolidation and reporting to headquarters, if applicable.

Each integration is a mini-project: what the external API supports, what happens on failure, and which side is the "source of truth" when data diverges must be agreed upfront. This is exactly where systems built without experience break most often.

5. Rollout

Working software is half the job. The other half is getting people to use it. Rollout includes:

  • cleaning and migrating legacy data (products, customers, balances, receivables);
  • role-based training, in the language your staff actually work in (Uzbek and/or Russian);
  • a parallel-run period, with old and new systems side by side and figures reconciled;
  • closing the "way back" — otherwise staff drift back to Excel.

6. Support and evolution

The business changes — a new branch, new products, new regulatory requirements, an external API update. The ERP has to change with it. A support plan should cover backups, monitoring, response times and how change requests are estimated and prioritised.

Data security deserves separate attention: who sees what, how access is revoked when someone leaves, where the database is hosted and how it is backed up. If you have group-level data policies, align them with local hosting and personal-data requirements early. See ERP data security.

A developer and a production manager map the process on a wall

What drives the cost of a custom ERP

A reliable price can only be given after discovery. A "fixed price" quoted before processes are studied usually means either the processes weren't understood, or change requests will follow. Cost drivers:

  • Number and complexity of modules — Effect on cost: Each module means its own logic, screens and reports
  • Process specificity — Effect on cost: Many exceptions and conditional rules increase scope
  • Number of integrations — Effect on cost: Each external system needs its own development and testing
  • Users and branches — Effect on cost: Roles, permissions, load, infrastructure
  • State of legacy data — Effect on cost: Cleaning messy spreadsheets takes real time
  • Mobile app or bot — Effect on cost: Additional interface and testing
  • Reporting and analytics — Effect on cost: Complex management and group reports are designed separately
  • Multilingual interface — Effect on cost: Uzbek, Russian and English UI and documents add scope
  • Support level — Effect on cost: Response time and development capacity

A detailed breakdown is in ERP cost factors. How to evaluate the spend as an investment and when it pays back: ERP return on investment.

One more point: the cheapest proposal is often the most expensive in the end. When discovery and process mapping are "saved", the costs come back as rewrites, extra integration work and the effort of regaining staff trust.

Key risks and how to reduce them

ERP projects fail more often for management and organisational reasons than for technical ones. The full list: why ERP projects fail. The most typical:

  • Vague goals — What it looks like: "Let's automate everything"; Mitigation: Discovery, measurable goals, clear scope
  • Big-bang launch — What it looks like: 12 modules at once, none finished; Mitigation: MVP and phased rollout
  • No local owner — What it looks like: Developer questions go unanswered; Mitigation: Named internal owner, weekly cadence
  • HQ–local gap — What it looks like: Requirements set remotely don't match local reality; Mitigation: Involve local managers and staff from discovery onward
  • Staff resistance — What it looks like: The system exists, but everyone uses Excel; Mitigation: Early involvement, training, closing the old path
  • Single-developer dependency — What it looks like: One person leaves, nobody understands the system; Mitigation: Documentation; code and database owned by the company
  • Neglected integrations — What it looks like: Bot or payment links break regularly; Mitigation: Error handling, monitoring, logs
  • Security postponed — What it looks like: Everyone sees everything, no backups; Mitigation: Roles, access logs, automated backups

Almost all of these risks are addressed at the start — in discovery, architecture and the contract. That is why a custom ERP should be treated not as a "coding service" but as a project combining business analysis, architecture and development.

Preparing for the project

Much can be done before the first meeting with a vendor. It shortens discovery and sharpens the outcome.

  1. List the pain points. Where did you lose money, time or customers in the last three months? For example: "we learned too late that a branch ran out of stock", "an agent's debt surfaced only at month-end", "one order was entered three times".
  2. Inventory your data sources. Which spreadsheets, Telegram groups, dashboards and applications are used, and who maintains them.
  3. Appoint a local project owner. Someone who knows the processes, has decision authority and can commit several hours a week.
  4. Define success. "Management sees sales and stock for every branch on one screen each morning" is measurable. "Everything should be automatic" is not.
  5. Split the budget in two. Build; then rollout and first-year support. The second part is the one most often forgotten.

Vendor selection checklist

A custom ERP is a long-term relationship. When reviewing proposals, get answers to these questions:

  • ☐ Is there a discovery and process-mapping stage before coding?
  • ☐ Is the project phased, with a clear date for the first working version?
  • ☐ What kinds of systems (sales, inventory, branches, local integrations) has the team built before, and can they show them?
  • ☐ Who owns the source code, database and server access — is it in the contract?
  • ☐ Will the system be documented well enough for another team to take over?
  • ☐ How are backups, roles and access rights organised?
  • ☐ What happens when an integration fails — is there an alert, is data preserved?
  • ☐ What are the post-launch support terms and response times?
  • ☐ Can the vendor work comfortably in Uzbek, Russian and English with your local team and headquarters?
  • ☐ Will the vendor tell you "this doesn't need automating" or "1C is enough here"?

The last question matters most. The partner who says "you don't need this" when it's true saves you more money than one who says yes to everything.

Conclusion

A custom ERP is not a software purchase; it is a redesign of how the business runs. Built well, it gives management accurate real-time numbers, makes branches manageable and frees the company from depending on one person's memory. Built badly, it costs more than money and time — it erodes staff trust in automation and makes the next attempt much harder.

So sequence matters: understand the process, map it, start with the most painful area, and expand in phases. Walking that path with a team that has hands-on experience with ERP, local integrations and business processes is what makes such a project controllable and predictable.

Frequently asked questions

Can we simply roll out our global ERP in our Uzbek subsidiary?

Sometimes, yes — especially for consolidation and group reporting. The gaps usually appear in local requirements: statutory accounting and tax reporting, electronic invoicing through local operators, local payment systems and the way local teams actually work. A common setup is the global ERP for group reporting, local accounting software for statutory books, and a local operational layer that connects them.

How long does custom ERP development take?

It depends on the number of modules, integrations and how well processes are already defined. With a phased approach, the first working version (MVP) goes live within months and further modules are added in stages. Trying to launch everything at once is one of the biggest risks.

Is 1C a reasonable choice for a foreign company in Uzbekistan?

For statutory accounting and tax reporting it is often the pragmatic choice: it is widely used locally, adapted to local requirements, and accountants know it. For operational processes that are specific to your business, many companies combine 1C with a custom system connected through integration.

Who owns the code and the data?

This must be stated explicitly in the contract: source code, database and server access belong to your company, and backups are stored where you have access. Also clarify data hosting location and access rights for headquarters before the project starts.

What ongoing costs remain after go-live?

Hosting, backups, bug fixes, changes driven by new business or regulatory requirements, and maintenance of integrations when external services change their APIs. Budget for these from the start.

Not sure what should be automated?
Let's map your processes together and see which parts are best served by off-the-shelf software and where a custom solution actually pays off.
Discuss a process map →

Let's work out where to start

Free consultation with a DORA specialist

Message on Telegram

Let's discuss your project

Leave your number and a specialist will get in touch.

Send a request via Telegram

Your details are used only to contact you.