Automation surrounds us everywhere: electronic document management, accounting, government services website. Even when you call to order a pizza, an electronic woman greets you and asks you to press a button.
But automation is not needed everywhere and not always. Let’s consider situations when it is worth automating some of the processes in an enterprise and when it is better to trust a person.
A story about a steam boiler
In the 70s, an enterprise operated in the USA. There was a steam boiler on its territory, which was constantly overflowing. I had to press the button to release the pressure so that nothing exploded. A mechanical engineer was in charge of the process, and he received a salary for a full time. The company decided to save money: dismiss the person responsible for the button and automate the pressing.
It hampers by one fact: the overflow interval ranged from several hours to several days. Because of this, it was not possible to put a simple mechanism that would press a button on a timer.
The management considered two options: replacing the boiler with a new one or installing sensors and automating the pressure relief. The first demanded a lot of money, and the second did not give a guarantee. These were the 70s of the last century, and people did not dare to trust the sensors, as they considered them unreliable.
An employee of the enterprise suggested teaching the watchmen to press a button and reducing the mechanical engineer. And so they did – they gave the guards instructions, prescribed fines for the explosion of the boiler, and raised the salary by 10%. The company worked like this for another ten years. Then the boiler broke down due to old age.
Did the management of the enterprise do the right thing? Found the solution was on the periphery, but it paid off. The meaning of the story is different. Each situation is unique, and no one will give a ready-made solution. Even automation has strengths and weaknesses – more on this in the following sections.
Business problems
When it comes to business, we face two extremes: the processes put on the conveyor or the employees do everything by hand. It depends on the scope of the business and the size of the company.
On the one hand, research shows that the field of recruiting is still not automated. The recruiter chooses the candidates himself, looks at the resume himself, contacts the person himself.
On the other hand, there is Amazon, where automation first penetrated the warehouse, and in 2016 the company introduced a fully automated store in Seattle. The buyer selects the goods and the debit money from the Amazon account at the exit.
Big deal
Harder doesn’t mean better. In large companies working with large amounts of data, the process design overload. Employees have to work with complex systems, and most of the functions are not in demand. In this case, it is better to lay down a cascade of process goals from simple to complex in advance. The point is that employees smoothly immerse themselves in the situation and achieve goals faster.
Not everything should be in one window.
Let’s say you have an accountant. Accountant needs data, some of which are in cloud Excel format. Some in a database, and some in the form of paper documents. He generates reports in Excel.
In the case of paper documents, the accountant first downloads the data from SQL, then imports it into Excel, and then builds reports and gives it to the boss.
To avoid unnecessary operations between the database and Excel, you can report in the SQL database itself. It would be a good solution and an excellent example of automation. The wrong option is to transfer all reports to SQL so that everything is in one window, and this will create a lot of unnecessary data import work and will not add value.
Each tool has its uses and limitations. You don’t have to try to drag everything into one place.
Optimization of working time
The manager wants to optimize the work of his subordinates. It seems fitting and logical: the business reduces harmful operations, increases labor productivity, and saves money. However, there is no direct relationship between speed and economy.
The company has an administrator. He answers calls, processes mail, and greets clients. If you implement an automatic mail processing system, you will free up 10% of the time. If you have ten such employees, then by saving 10%, you can cut one of them or not hire a new one. If you reduce by 10% the working time of the only employee who has nothing else to do, then the rest of the time he will watch cats on the Internet. There will be no savings.
”Add this function only.”
The director of the department asks to implement only the registration of documents of a particular type. The Automator quickly creates this function and hands it over to the customer. When using the algorithm, interesting questions arise: who can and cannot register these documents, when should do it, and what time frame? As a result, identifying flaws and fixing them turned into a long and thorny path.
It turns out that you need to design the automation itself and the business process itself. It is necessary to highlight the goals of the process, the roles of employees, the timing of the request, and much more.
It is easier to achieve success if you implement features in stages. Each step will contain a set of functions required for the entire process to work.
When is automation needed?
This issue requires an integrated approach: it is worth weighing the pros and cons, calculating the benefits, and consulting with experts. But there are vital issues to look out for if you’re thinking about automation.
Pros for automation:
- The process repeats constantly. It could be a daily report or email notifications. Suppose you have a separate employee who uploads contacts to the database for half a day. In that case, it should automatically upload, and the employee should find the best use, especially if there are several such workers.
- Elimination of the human factor. Copying data to a backup server, setting access rights, and accounting is easy to automate.
- There is dissatisfaction with any internal services in the company on the part of business units.
- You want to transfer proven practices from other companies.
- New ideas for business units don’t implement due to the lack of process automation.
Against automation:
- Price. Consider the cost of the application, its implementation, and support. It may be cheaper to hire an employee or pay an existing one.
- Processes are constantly changing. It happens with startups at the very beginning of the journey. If you don’t have established procedures, you risk automating chaos.
In general, there are three basic principles of automation:
- Automation must be beneficial. It means that automation brings profit to the company – for example, automatic approval of loan applications in a bank. The bank automates the process, saves employees time, and makes money on loans issued.
- Automation shouldn’t be harmful. It means that after the implementation of the system, the process should be simplified, not complicated. If additional points of failure appear after automation, this is a reason to think.
- Automation should cut costs. If there is no economic benefit, then there is no point in automation.
- If in your case, all three principles perform, then this is good and useful automation.
Conclusion
Automation of processes, and especially IT processes, is a good thing. It allows you to cut costs, optimize your work, and get rid of errors. But you don’t need to automate everything. If you do the work once a year, it will be cheaper to write instructions and give them to employees.
If the implementation of automation is a settled issue, consider the risks of do-it-yourself design. An experienced company will customize the system for your business processes and implement additional modules upon request.
