Automation· July 5, 2026

Business automation: where to start and what to automate first

One in four employees in Russia spends more than six hours a week on reports that nobody has automated; that is the finding of a 2025 survey by hh.ru, Russia’s largest job site. Almost one working day a week goes into moving numbers from place to place. Business automation does not start with buying a system. It starts with answering one question: which manual area costs you the most.

This guide maps out the whole path: how to tell it’s time, what to automate first, which route to choose (an off-the-shelf system, a no-code builder or custom development), where to bring in AI, and how not to end up among the 70% of digital transformations that, according to BCG, fail to reach their goals.

What does business automation deliver in practice?

In short: automation means an operation is entered once, at the moment it happens, and from there it flows into accounting, reports and analytics on its own. Everything employees carry by hand between spreadsheets, emails and systems starts happening without them, faster and without errors.

The effects are measurable. According to a study by SberAnalytics and Sber Business Soft (November 2025, a survey of 559 respondents), companies that have adopted automation report that routine tasks get done faster (45% of companies), the share of manual operations has gone down (37%), and there are fewer errors in documents and reports (36%). The areas automated most often are document workflow and request processing (70%) and accounting and financial records (55%).

For a business owner, this comes down to three things: employees do their actual work instead of moving data around; the numbers add up without reconciliations; and you see the state of the business in real time, not in last month’s report.

How do you know your business is ready for automation?

If answering “how much did we make this month” takes half an hour of reconciliations and three people, it’s time. That moment comes earlier than it seems: routine work grows quietly along with the business, and the company gets used to manual labor as the norm until it starts costing noticeable money.

Typical symptoms:

The same figure is entered by hand in two or more places, and it differs from place to place.
There is an employee whose job is half “consolidating” spreadsheets.
Requests live in managers’ personal chats, and nobody has the full picture of the customers.
One person leaving brings a process to a halt: only they know “how the file works.”
The report for the owner is out of date by the time it’s ready.

If you recognized yourself in two or more of these, you are already paying the “Excel tax”: in employee time, accounting errors and decisions made on outdated numbers. A detailed breakdown with seven signs is in the article when it’s time for a business to move off Excel.

Why can’t you automate chaos?

You automate a process that already works: this is the rule that a noticeable share of projects break against. If requests get lost, responsibilities are blurred and records are kept “from memory,” automation applied to chaos simply speeds up the chaos: the mess multiplies at machine speed, and people can no longer keep up with catching the errors.

“AI is a magnifying glass. If there’s a mess under it, you’ll see a very large and very detailed mess.” — Stas Cheprasov, entrepreneur, from a write-up on vc.ru about automating a small business.

That’s why step zero of any project is to take the process apart: where the data originates, who enters it, who needs it next. We sit down with the people who do the work by hand and record the real process, and this usually takes a day or two, not months of written procedures. The real process almost always differs from how the manager imagines it.

What to automate first: do the math in money

The first area to automate is the one with the highest cost of routine work. It is calculated with a formula: frequency of the operation × time per operation × employee’s rate + cost of errors. The formula takes emotion out of the choice: the area with the largest monthly losses wins. This turns the automation decision into an ordinary investment decision with a payback period.

An example calculation. Reconciling the bank with the books takes an accountant two days a month, which is 24 working days a year, a full month of work. Add the cost of errors: a discrepancy found a month later costs hours of investigation, and sometimes real money. An area like this pays back automation faster than any other.

The statistics confirm it: the areas automated first are most often document workflow and request processing, accounting and customer support, in other words, high-volume, repetitive processes that are expensive to run. Classic candidates for the first step:

Reconciliations: bank against the books, warehouse against sales, settlements with counterparties.
Moving data between systems: from email to a spreadsheet, from a spreadsheet into 1C, the accounting and ERP platform most Russian companies run on.
Standard documents: invoices, contracts and acceptance certificates filled in by copy-paste.
Initial customer inquiries: common questions, qualification, booking.

How to choose the area without getting buried in a “year-long project” is covered in a separate method: automating routine work: where to start so the first step pays off.

If you’d rather skip doing the math yourself, we do it at the first consultation: we go through your processes, find where money and time are being lost, and set a goal in money terms. That’s how every one of our custom development projects begins.

Off-the-shelf, no-code or custom development: how to choose your path

The path depends on two things: how standard your process is and what share of turnover you are prepared to spend on automation. Vendors of each option pull in their own direction: for an off-the-shelf vendor, custom work is something “almost nobody needs,” and for developers, no-code builders are “toys.” Here is a neutral frame:

Path	When it fits	Hidden cost
Ready-made system (“off-the-shelf”)	Standard processes: accounting, EDI (legally binding electronic document exchange between companies in Russia), standard CRM	The process bends to someone else’s logic; customizations held together with workarounds; a subscription forever
No-code builders	Testing hypotheses, simple internal tools	Rising platform prices, a ceiling on complexity; if the platform shuts down, you lose the product
Custom development	The process is non-standard and makes money; deep integrations and control over data are needed	More expensive and slower at the start; requires a competent contractor

Testing a hypothesis even before a no-code builder is often faster with vibe coding: AI writes a prototype from your description in an evening. The boundary is the same as with no-code: as soon as real data, payments and integrations appear, you need custom development.

A budget guideline: for a company with revenue in the hundreds of millions of rubles, automation costs an acceptable 1–2% of turnover. For a business with a turnover of 5–10 million rubles, the same costs become 10–20% of turnover, and that is critical: a “grown-up” system is a bad fit for such a company. It should start with a targeted solution.

The setup that works in practice is a hybrid: a standard core where the process is standard, and custom development where the process sets you apart from competitors. That’s the principle we build on: our catalog of ready-made solutions covers the quick start, and custom development covers what makes the business what it is.

Where do you need AI, and where is classic automation enough?

AI isn’t needed everywhere, and that’s fine. Rules, integrations and schedules cover the bulk of routine work more cheaply and predictably, without “hallucinations” or fine-tuning. A model makes sense where the input data is unstructured and the volume is high: free-form customer questions, emails, documents in arbitrary formats.

Task	What works
Approvals, document routing, notifications	Classic: “if–then” rules
Syncing systems, reports, auto-generated documents	Classic: integrations and templates
First-line support, sorting incoming requests	AI: free-form questions, qualification
Extracting data from emails and documents	AI: unstructured input
Decisions with a high cost of error (money, legal matters)	A person, for whom the system prepares the data

The economics of AI on the first line have already been calculated: according to a Sber study, AI assistants in support and call centers cut payroll costs by up to 30–40%. In our case with a property developer, the first-line bot took over the entire flow of initial inquiries: managers now work only with warm leads. When such an agent pays off and when it stays a toy is covered in the article on the first-line AI agent.

How much does automation cost, and how do you calculate payback?

There is no single price list for automation, and that’s fine. The price depends on the number of areas, the depth of integrations, and whether you are building a targeted solution or a full setup. Don’t trust fixed “all-inclusive” price tags: behind them is either a template that will bend your process or extra charges as the project goes.

The right frame is an investment one. You have calculated the cost of routine work for an area (the formula above); now compare it with the cost of the solution and you get the payback period. A project without a clear payback period isn’t worth starting, however high-tech it looks.

We’ve described our approach to estimates on the page about development costs: first an audit and a goal in money terms, then an estimate for the specific task. That way the client pays for a result they calculated themselves.

How not to lose data: Federal Law 152-FZ and turnover-based fines

Since May 2025, a personal data leak is no longer an abstract risk. A first leak costs a company a fine of 3–15 million rubles; a repeat leak brings a turnover-based fine of 1–3% of annual revenue, from 20 to 500 million. These are the current provisions of Russia’s Code of Administrative Offences after the 2025 amendments, and an automation project is exactly where data starts moving between systems.

For a project, this means three requirements for the contractor: data and hosting in Russia, a contract with an NDA and a clear division of responsibility, and role-based access, meaning who sees and changes what. Uploading your customer base to public foreign AI services “for a test” is a way to end up in the leak statistics.

We build setups on the client’s infrastructure or on servers in Russia, in compliance with 152-FZ, Russia’s personal data protection law. Sensitive data does not leave your perimeter without approval, and this is set out in the contract.

What an automation project looks like: the stages

A project that gets to results is structured the same way in almost any industry: first an audit and a goal in money terms, then architecture, then rollout in iterations and ongoing support. On paper this sounds obvious; in practice, the projects that reach results are the ones where none of the four stages was skipped. Our process looks like this:

Understand the task. We look at the processes as they actually run and find where money and time are lost. We set a goal in money terms; without it, the project can be neither evaluated nor accepted.
Design the system. Architecture for the industry and scale: what we build, what we integrate, and in what order. This is also where the fate of the “service zoo” is decided: what to keep and connect, and what to replace.
Build and roll out. The system is assembled in iterations: the first working area appears within weeks, not “everything at once at the end.” Data is migrated and the team is trained.
Support and develop. The business changes, and the setup is extended: new integrations, areas and automations as you grow.

A separate topic is connecting existing systems: 1C for the accountant, a CRM for sales, the bank, messengers. Replacing everything with a single “mega-system” is usually unnecessary; the right move is to connect what already works into a single setup. How this works technically is shown in the article on integrations.

How do you choose a business automation contractor?

Choose an automation contractor based on experience with similar processes, the ability to understand your business, and transparency about the work ahead. Ask them to show relevant cases, document requirements and acceptance criteria, name the team and the testing procedure, and explain what knowledge, documentation and support will remain after launch.

Experience and understanding of the process

Check for similar projects and make sure the contractor starts with questions about the business, not with a tool.

Requirements and solution validation

Document the features, integrations, scenarios, stages and acceptance criteria; ask for a demo or a prototype.

Handover and support

Agree in advance on documentation, training, access, team composition and how support will work after launch.

Why do 70% of transformations miss their goals, and how do you end up in the 30%?

According to BCG research, 70% of digital transformations fail to reach their stated goals, and automation projects are no exception. The reasons for failure repeat so consistently that you can check them as a list before you start. Here are the five most expensive mistakes we see most often:

Digitizing chaos. The process didn’t work by hand, and now it doesn’t work automatically, only faster.
Copying “what everyone else does.” Someone else’s sales funnels and enterprise systems in a small business produce an expensive system nobody uses.
Building a patchwork quilt. The systems aren’t connected, and there is no single source of data.
Not defining metrics before the start. There is nothing to prove the effect with: the project is declared a failure even where it worked.
Not appointing an owner. Workflows break, data drifts apart, and nobody notices until it’s too late.

“One customer gets created three times. Each system has its own fields, its own formats, its own logic.” — Dmitry Goroshko, from a write-up on vc.ru about how automation starts working against a business.

Every item on this list is dealt with at the audit and design stages, which is why we don’t start development without them.

The first 90 days: an action plan

In three months you can go from “everything by hand” to the first automated area with a calculated payback, without buying systems at the first step and without stopping work. The plan below is drawn from our process and assumes you carry it out alongside your normal workload:

Weeks 1–2: go through your processes asking “where is the same action repeated,” “where are errors most frequent,” “where is the wait longest.” Note down 3–5 candidates.
Weeks 2–3: calculate the cost of routine work for each candidate using the formula frequency × time × rate + cost of errors. Sort them by money.
Weeks 3–4: choose one lead area. Define the success metric before the start: hours, errors, speed, in other words, what exactly will change and by how much.
Month 2: launch the first area, with a targeted solution or custom development. The old process runs in parallel for a while: the numbers are cross-checked and trust in the system grows.
Month 3: compare the metric against the “before” baseline and calculate the actual payback. Only then choose the next area.

It’s easier to track the effect with numbers that update themselves: which metrics an owner should see daily is covered in the piece on the owner’s dashboard.

If you want to take this path with a team that has 300+ automation and AI projects behind it at companies in Russia, tell us about your task: we’ll go through your processes, calculate the cost of routine work and propose a plan with a goal in money terms.

Sources

Frequently asked questions

Is automation only for large companies?+

Not anymore. In a 2025 SberAnalytics study, small businesses made up 22% of companies automating their processes: in Sberbank’s assessment, automation is no longer the preserve of large players. For a small company there is one rule: start with targeted automation of the most expensive area, not with a “big system.”

How long does an automation project take?+

Targeted automation of a single area takes weeks. A full setup (an ERP ecosystem connecting accounting, sales and analytics) takes months, but we roll it out in iterations: the benefits appear as the project goes, not at the very end.

Can we start with a ready-made solution instead of custom development?+

Yes, if your process is a standard one. For a quick start we have a catalog of ready-made solutions: an inventory bot, a first-line AI agent, a financial calendar and others. Each one is adapted to your processes, and when the business runs into its limits, the setup is extended with custom development.

What happens when the process changes after rollout?+

The system has to change along with the business, which is why we support and develop the setup after handover: improvements, new integrations and new automations as you grow. A system without support becomes outdated in a year to a year and a half.