The more often you do it, the better you get at it
Automation is hot. DevOps, CI/CD, Infrastructure as Code, it sometimes feels like manual work has become taboo. But is automation always the smartest choice?
Cindy Beernink, Database Consultant at OptimaData, often sees teams taking their drive for automation a step too far. “Sometimes building a script takes more time than simply solving the problem manually,” she explains. In this blog, she shares when scripting truly adds value and when you’re better off just rolling up your sleeves and getting it done.
Scripts can deliver tremendous time savings. Take provisioning a new virtual environment, for example. If you need to configure dozens of servers with identical settings, automation is invaluable. It reduces human error, enforces best practices, and allows you to rule out certain root causes immediately when issues arise.
“If you know multiple environments were built in exactly the same way, the risk of inconsistencies is virtually zero,” Cindy explains. “That means when an incident is reported, you can immediately focus on deeper root causes. That saves stress, time, and money.”
Teams often try to build technical masterpieces instead of keeping scripts clear and maintainable. Cindy’s approach is straightforward:
“When I transfer knowledge, I first let someone perform the steps manually two or three times. That way, colleagues truly understand what’s happening. Only when tasks start repeating does scripting make sense. And if you do automate, make sure the script provides clear, readable logging so it’s obvious what actions are being performed.”
“In our work as specialized consultants, our involvement is often short-term or one-off. Our real value lies in knowledge transfer, enabling customers or teams to move forward independently, with a deeper understanding than before.”
Not every situation calls for the same approach. Cindy points to different types of databases as an example:
“A data warehouse database is used very differently from a third-party application database. They require different configurations. Standardization is great but flexibility remains essential.”
This becomes even more apparent in day-to-day management.
“We often deal with troubleshooting, such as issues that only occur in production. For recurring problems, scripts can help. But you’ll frequently find that you need slightly different information each time. You can keep modifying scripts, but sometimes it’s simply faster and more efficient to run ad-hoc queries or use the available management tools.”
Cindy highlights another advantage of working manually:
“Charts and visualizations often make issues much clearer for the customer. They help people understand what’s happening and why action is needed.”
How do you decide whether something is worth automating? Cindy uses a practical checklist:
Market pressure to automate everything is strong. DevOps, CI/CD, and Infrastructure as Code are no longer buzzwords, they’re the norm.
“Continuous improvement and automation are key elements of the DevOps lifecycle,” Cindy says. “But implementing improvements and automating the resulting changes requires proper analysis first. Scripts can help with problem analysis, but sometimes it’s faster to manually review settings and logs before you start automating. Once you clearly understand the actions you want to perform, you can translate the solution into a script and automate it in line with DevOps principles.”
Some activities simply aren’t well suited for automation.
“Troubleshooting is a good example,” Cindy explains. “Often, you gain more by running targeted queries.”
There are situations where you want to work step by step, stay in control, and intervene when necessary, especially in business-critical processes where continuity is at stake. In those cases, manual control is indispensable.
Cindy’s advice?
“Ask yourself whether automation truly adds value. Does it save time? Does it help your colleagues? Scripts are powerful but they’re not always the answer. Sometimes automation costs more time than it delivers. I encourage teams to think more consciously before deciding to script something.”
Struggling with the question of when to automate and when not to? At OptimaData, we’re happy to think along with you. We don’t just deliver technical solutions; we also ensure knowledge transfer so your team can move forward independently. Feel free to get in touch.