Knowing which processes to automate is not enough. You also need the right method to turn a good idea into a project that genuinely works in day-to-day operations.
Many automation projects start well and lose momentum after a few weeks. This is not because the idea was poor, but because the project lacked a method. Choose the wrong tool, leave the team out of the process, or fail to map the workflow properly, and automation can add complexity instead of removing it.
Here are the steps for managing an automation project from beginning to end without losing your way.
Step 1: Map the Process as It Works Today
Before discussing tools, you need to understand what actually happens in practice.
For each targeted process, list:
- who is involved at each stage;
- which information moves between people and in what format, such as email, Excel files, or paper;
- how long each stage takes;
- where bottlenecks and unnecessary back-and-forth occur.
This mapping often reveals surprises. A process that appeared simple may involve five people and three different tools. Conversely, a task considered complex may amount to three clicks repeated one hundred times each month.
Without this step, you risk automating a poor version of the process or automating a problem that should have been solved in another way.
Step 2: Define a Measurable Objective for Each Automation
“Saving time” is not an objective; it is an intention. An automation project needs measurable benchmarks so that it can be managed effectively.
Useful indicators include:
- average request processing time before and after automation;
- data-entry error rate;
- time between an order and its approval;
- number of manual follow-ups avoided each month.
Setting an objective from the outset also lets you determine whether the automation has delivered on its promises. Without an indicator, you cannot justify the investment or make informed adjustments.
Step 3: Choose Between an Existing Solution and Bespoke Development
This is often where projects become more complicated. You have two main options.
An existing solution, such as a CRM, ERP, or no-code workflow tool, is suitable when the process is standard and the tool already covers most requirements. It is generally faster to implement and less expensive initially.
Bespoke development becomes relevant when:
- the process is specific to your industry;
- several tools must be connected in ways that standard solutions do not support;
- the volume or complexity justifies a dedicated investment;
- existing solutions force you to distort your working methods to fit their limitations.
A good approach is to test an existing solution within a limited scope. If it reveals significant limitations, bespoke development becomes an evidence-based choice rather than a gamble.
Step 4: Design the Workflow Before Coding or Configuring Anything
A workflow is a series of rules: when a specific condition is met, a particular action is triggered.
At this stage, formalise:
- triggers, such as a new quotation, a due date, or a received email;
- automatic actions, such as sending an email, creating a task, or updating a status;
- special cases and exceptions, such as missing information or a rejected approval.
Ignoring exceptions is one of the most common mistakes. A workflow that performs perfectly in 90% of cases but becomes completely blocked in the remaining 10% will eventually create more frustration than the old manual process.
Step 5: Test Before Deploying at Scale
An automation project should never be deployed across the entire company at once. Begin with a limited scope: one team, one type of case, or a defined period.
This pilot phase allows you to:
- confirm that the workflow behaves as expected under real conditions;
- identify cases that were not anticipated during design;
- collect feedback from the people who use the tool every day.
A two- to three-week trial is generally enough to identify the main adjustments required before wider deployment.
Step 6: Help Teams Adopt the New Process
A technically perfect tool can fail if teams do not use it or use it incorrectly.
Several principles make a real difference:
- explain why the change is happening, not only how to use the tool;
- train teams using concrete situations from their daily work rather than generic examples;
- provide a point of contact for questions during the first few weeks;
- clarify what will remain unchanged as well as what will change.
Teams involved in mapping the process during step 1 generally adopt the tool more easily. They recognise their work in the solution and do not feel that something has been imposed from outside.
Step 7: Monitor Results and Improve Over Time
Automation is never fixed permanently. Requirements evolve, volumes change, and new situations emerge.
Schedule regular reviews to:
- compare the indicators defined in step 2 with actual results;
- identify new issues reported by teams;
- assess whether the process can be simplified further.
A successful automation project is not one that works perfectly on day one. It is one that improves progressively through real-world feedback.
How Much Time and Budget Should You Allow?
It depends entirely on the scope. A simple automation using an existing tool, such as automatic follow-ups, notifications, or task creation, can be operational within one or two weeks.
A project involving several business tools, system integrations, or bespoke development generally requires several weeks for mapping, design, development, and testing.
The budget varies in similar proportions. It is better to begin with a limited scope and demonstrate concrete results than to target a company-wide project from the outset.
When Should You Work with a Specialist?
Some projects can be managed internally, especially when they rely on tools that are already in place. Specialist support becomes valuable when:
- several tools need to be connected;
- the process spans several departments and requires an overall view;
- a standard solution does not cover the company's specific requirements.
At Monark IT, we help SMEs design and develop bespoke workflows, from initial process mapping to deployment and performance monitoring.
FAQ: The Steps in an Automation Project
Should You Automate an Entire Process at Once or Work in Stages?
Work in stages. Starting with a limited part of the process allows you to confirm that the automation works before extending it. This approach also makes adoption easier for teams.
How Long Does an Automation Project Usually Take?
It depends on the scope. A simple automation can be implemented within a few days. A project involving several tools or bespoke development will generally take several weeks.
How Do You Know Whether the Automation Was Successful?
Compare the indicators defined before the project, such as processing time, error rate, and delays, with the results achieved after implementation. Without a measurable objective at the outset, evaluating the outcome is difficult.
What Should You Do If Teams Do Not Adopt the New Tool?
Return to the reasons for the change and check whether the workflow genuinely reflects how teams work. The problem often comes from insufficient team involvement at the beginning, not from the tool itself.
Is It Better to Use an Existing Tool or Build a Bespoke Solution?
It depends on how specific the process is. An existing tool is suitable for standard requirements. Bespoke development becomes relevant when several systems need to be connected in ways that standard solutions cannot support.
How Do You Build Automation That Lasts?
Automating a business process is not simply a matter of choosing software. It is a project in its own right, requiring an understanding of current operations, clear objectives, and the involvement of the teams concerned.
Following these steps one by one helps avoid false starts and creates automations that remain useful over time instead of solutions that teams stop using after a few months.
