# How to tell whether an automation is worth building

> Not every task deserves automating. Compare time saved, error reduction and maintenance cost with a simple calculation before you build.

Web page: https://oravol.com/insights/is-an-automation-worth-building

## About this article

- **Published:** 4 October 2026
- **Topic:** AI & Automation
- **Publisher:** ORAVOL

## Key takeaways

- Estimate time saved per month, then compare with build and running costs.
- Count reduced errors and faster response, not just hours saved.
- Simplify or remove a process before automating it.

## Article

Automation is satisfying to build, but some automations cost more to create and maintain than they ever save. A quick evaluation before building keeps effort where it pays off.

### Step 1: describe the task

Write down the task, how often it happens, who does it and how long it takes each time. Include the time spent fixing mistakes and chasing missing information.

### Step 2: estimate the time saved

**Monthly time saved = frequency per month × time per occurrence × share that can be automated.**

If a task happens 200 times a month, takes five minutes and four out of five occurrences can be automated, that's about 13 hours saved per month.

### Step 3: put a value on it

Multiply hours saved by the cost of the people doing the work. Then add benefits that don't show up as hours:

- **Fewer errors:** wrong invoices, missed follow-ups, data entry mistakes
- **Faster response:** quicker replies can win more enquiries
- **Scalability:** handling growth without hiring for admin
- **Staff satisfaction:** less repetitive work

### Step 4: estimate the costs

- Build time, internal or external
- Tool subscriptions and usage fees
- Maintenance: updates when connected apps change, fixes and monitoring
- Handling exceptions the automation can't process

### Step 5: compare

If the monthly value comfortably exceeds the monthly running cost, and the build cost pays back within a reasonable period, it's a good candidate.

### Red flags

- The process changes often or isn't well defined yet.
- Most cases are exceptions that need judgement.
- The task happens rarely.
- The tools involved have no reliable integration options.

In these cases, document and standardise the process first.

### Simplify before you automate

Automating a messy process makes it faster but not better. Remove unnecessary steps, approvals and duplicate data entry first; sometimes that alone solves the problem.

### Measure after launch

Track actual time saved, error rates and exceptions for a few months. Use the results to decide what to automate next.

Our [AI automation](https://oravol.com/services/ai-automation-chatbots) service starts with this assessment, so we build automations that pay for themselves.

## Questions

### What is a reasonable payback period for an automation?

Many businesses look for payback within months for small automations. Larger systems can justify longer periods if they enable growth.

### What hidden costs should we include?

Subscriptions, usage-based fees, maintenance when connected tools change, monitoring and the time spent handling exceptions.