Automation — What It Does and What It Doesn't

Automation in the workplace gets discussed in two registers: the alarming one ("AI will replace your job") and the dismissive one ("just a bit of efficiency"). Neither is accurate for most people in most roles right now.

The reality is more specific and more useful: automation replaces repetitive, rule-based tasks within jobs — not jobs themselves, at least not yet, and not uniformly. Understanding what can be automated, how to think about automating it, and what cannot be automated gives you a practical advantage regardless of your role or industry.

The Logic Behind Automation

Almost all workplace automation follows the same underlying logic: if this condition is met, then take this action.

• If a new lead fills out a contact form, then add them to the CRM and send a welcome email;
• If a project task is marked complete, then notify the next person in the workflow;
• If a file is added to this shared folder, then convert it and move it to the archive;
• If today is the 25th of the month, then generate and send the usage report. This "if-then" structure is why automation tools are sometimes called workflow automation or trigger-action tools. The tools themselves — Zapier, Make (formerly Integromat), Microsoft Power Automate, and similar platforms — require no coding. They connect existing apps and define the trigger-action pairs through a visual interface.

What's Worth Automating

The best candidates for automation share three properties: they're repetitive (the same task done more than weekly), they're rule-based (the same inputs always produce the same action), and they're low-judgment (no human decision is required at each instance).

Common workplace tasks that meet these criteria:

• Routing incoming emails or form submissions to the right person or system;
• Generating regular reports from a data source and distributing them on schedule;
• Sending reminder notifications when deadlines approach in a project management tool;
• Syncing data between two systems that don't natively connect (a CRM and a billing system, for example);
• Archiving or renaming files according to a consistent convention. Tasks that don't meet the criteria — ones that require judgment, context, relationship awareness, or creative problem-solving — aren't good automation candidates, at least not with current tools.

What Automation Doesn't Fix

Automation amplifies whatever system it's built on. If the underlying process is broken, automating it makes the breakage faster and harder to notice.

Before automating anything, it's worth asking: is this task worth doing at all? Sometimes the right answer is to eliminate the task, not to automate it. A weekly report nobody reads that gets generated manually is inefficient. A weekly report nobody reads that gets generated automatically is invisible inefficiency — harder to notice and therefore harder to stop.

The second trap: automation creates dependencies. When a workflow relies on an automated connection between two systems, someone needs to own the maintenance of that automation — notice when it breaks, update it when either connected system changes, and document it so others can manage it. Automation that nobody understands is technical debt with a timer.

Building the Habit

The most practical entry point into workplace automation for most people isn't a complex multi-step workflow. It's identifying one task you do more than three times per week that follows the same steps every time — and asking whether a tool already exists to do it automatically.

Most organizations' existing software has automation features that go entirely unused because nobody explored them. Calendar tools, email clients, project management platforms, CRMs — most have built-in automation capabilities that require no additional tools. The ROI on one hour of exploration is often significant.