Profitable Business Automation Starts with…

Business automation often starts with the wrong question: "What tool should we buy?" A more useful question is: "Which repetitive tasks are the most costly if continued manually?"

The difference may seem small, but the impact is significant. Businesses that buy tools first often end up with multiple subscriptions, disconnected workflows, and teams still performing old tasks in new ways. In contrast, businesses that start with the problem can choose technology more thoughtfully and measure whether the investment truly pays off.

Interest in technology is indeed increasing. The U.S. Chamber of Commerce report in 2025 notes that 58 percent of small businesses surveyed are using generative AI, up from 40 percent in 2024. However, OECD also shows that the use of AI in core business functions remains relatively low, especially among small and medium enterprises.

This means that the challenge for businesses is no longer just finding new technology. The challenge is finding sensible points of use.

Start with Repetitive and Easily Measurable Tasks

Processes suitable for automation typically have three characteristics: they are performed frequently, follow a relatively fixed pattern, and their outcomes can be clearly verified.

Examples include:

Transferring order data from forms to spreadsheets.
Sending payment reminders to customers.
Creating weekly sales report summaries.
Classifying incoming emails based on request types.
Sending notifications to the team after a transaction is successful.

Conversely, processes filled with exceptions, requiring negotiation, or involving important decisions should not be fully automated right away. Technology can help prepare information, but the final decision still requires human input.

Calculate the Costs of Manual Processes Before Choosing a Solution

Use simple calculations. For example, if a staff member spends 30 minutes each day checking payments and sending reminders, over 22 working days, the time spent amounts to 11 hours per month.

If the staff's time is estimated at Rp75,000 per hour, the cost of this process is approximately:

11 hours × Rp75,000 = Rp825,000 per month

This figure does not include the costs of errors, delays, and missed opportunities because staff are not working on higher-value tasks.

This calculation does not automatically mean that the business must purchase software costing Rp825,000 per month. There may be simpler solutions that are sufficient, such as email templates, spreadsheet filters, built-in integrations, or automation rules in existing applications.

The goal of this calculation is to provide a rational boundary. If a tool cannot significantly save time or reduce errors, its subscription cost needs to be questioned.

Differentiate Between Time Savings and Increased Output

Automation does not always generate money directly. There are two types of benefits that need to be separated.

1. Operational Savings

This occurs when the same task can be done in less time. For example, a report that previously took two hours can now be completed in 20 minutes.

2. Increased Business Capacity

This occurs when the time saved is used to produce something of greater value. Staff who are no longer busy copying data can handle more customers, improve documentation, or follow up on sales.

The second type of benefit is often more important but also harder to measure. Therefore, do not just ask, "How many minutes are saved?" Also ask, "What is that time used for?"

Use a Simple Scoring System to Choose the First Process

If there are many candidates, score each process from 1 to 5 based on four criteria:

Frequency: How often is the task performed?
Time: How long does the task take?
Error Risk: What is the impact if an error occurs?
Clarity of Rules: How easily can the process be translated into consistent steps?

Add up the scores. Processes with high scores become initial candidates, especially if they can be tested without changing the core business system.

For example, sending payment reminder emails may score high due to its frequency, clear rules, and easily verifiable outcomes. Meanwhile, automating product pricing may score low because it requires market context and human consideration.

Run Small Trials, Not Large Projects

Trials should have clear boundaries. Choose one workflow, one team, and a measurement period of two to four weeks. Record the conditions before automation:

How long does the process take to complete?
How many people are involved?
How often do errors occur?
How much work is pending?

After automation is implemented, measure the same factors again. Do not rely solely on subjective impressions like "it feels faster." Comparing before and after helps distinguish real improvements from mere enthusiasm for a new tool.

Good automation is not the most sophisticated, but rather what makes important processes faster, more consistent, and still easy to monitor.

Do Not Overlook Hidden Costs

Subscription prices are not the only costs. Also consider the time needed to set up integrations, train the team, fix failed workflows, and transfer data if at some point you want to stop using the service.

Also check several practical matters:

Is customer data stored and processed in a way acceptable to the business?
Are there usage limits that could cause costs to spike?
Is there a way to export data if the service is discontinued?
Who is responsible when automation fails?
Do all critical steps have records or notifications?

This is especially important if automation touches payments, customer data, financial documents, or external communications.

What Does This Mean for Small Businesses?

Small businesses do not need to mimic the systems of large companies. The benefits can actually come from one or two simple processes that have been consuming attention every day.

Steps that can be taken now include:

Document repetitive tasks for one week.
Choose processes with high frequency and clear rules.
Calculate the time and costs currently used.
Test the simplest solutions before purchasing a new platform.
Set concrete success metrics.
Document the processes and prepare manual checks for exceptions.

If trial results show consistent savings, then expansion can be considered. If not, the failure is still useful because the business can halt the experiment before costs and complexity escalate.

Ultimately, automation is not a race to collect applications. It is a way to transform wasted time into new capacity. Businesses that can calculate, test, and monitor their processes will find it easier to reap the benefits of technology without getting trapped in subscription costs and unproven productivity promises.

Sources & Further Reading

– Rio Yotto @rioyotto