Find Operational Friction Before You Automate

Find Operational Friction Before You Automate

Automation can remove repetitive work, but it can also make a weak process fail faster. Before adding a new tool, workflow, or AI capability, it helps to understand where the work is actually getting stuck.

Operational friction is the avoidable effort that accumulates around otherwise necessary work. It shows up as repeated questions, manual copying, unclear handoffs, missing information, status chasing, rework, and routine decisions that keep returning to the same person.

The goal is not to label every manual task as inefficient. Some work is naturally hands-on, judgment-heavy, or customer-specific. The useful question is simpler:

Start with the work, not the tool

A common mistake is to begin with a technology and ask where it could be used. A better sequence is:

• Identify the work that creates recurring friction.
• Understand why the friction happens.
• Decide what kind of improvement would remove the cause.
• Choose technology only when it fits that improvement.

This sequence matters because similar symptoms can have very different causes. A slow customer-response process might need a clearer ownership rule, a better intake form, saved response templates, a CRM cleanup, or an automated follow-up workflow. The right answer depends on the underlying pattern.

Five friction patterns worth looking for

1. Repeated handoffs

Work moves between people, teams, inboxes, spreadsheets, or systems, but nobody is completely sure who owns the next step.

Typical signs include:

• customers asking for status because ownership is unclear;
• work sitting between stages;
• repeated internal messages asking, "Who has this?";
• information being re-entered after every handoff.

The first improvement may be an ownership rule, a handoff checklist, or a shared status view rather than automation.

2. Duplicate entry and copying

The same information is entered into more than one place, copied from email into a spreadsheet, or moved manually between systems.

This is a strong improvement signal because the work is usually repetitive and rules-based. Before automating it, confirm which system should be authoritative and whether the duplicated fields are actually necessary.

3. Waiting and approval bottlenecks

Routine work pauses while someone waits for information, approval, a decision, or access to a tool.

Look for:

• owner or manager approvals on ordinary decisions;
• jobs that cannot proceed because required information was not collected earlier;
• repeated follow-up to obtain the same kinds of inputs;
• employees waiting because decision limits are unclear.

Sometimes the best fix is not faster approval. It is defining which decisions no longer require escalation.

4. Rework and preventable correction

Work is completed, then repeated because something was missing, misunderstood, inconsistent, or entered incorrectly.

Rework often points to a process-definition problem. Checklists, templates, validation steps, clearer intake requirements, or better handoff information may solve the problem before more sophisticated technology is needed.

5. Poor visibility

People spend time asking for status, assembling reports manually, or making decisions without reliable information.

When the data already exists but is scattered, a reporting or dashboard improvement may create more value than automating the underlying workflow. Visibility can also reveal where a later automation project would have the highest impact.

Match the improvement to the cause

Once you identify a friction pattern, classify the likely first move.

Process clarification

Use this when ownership, sequence, decision rules, or exception handling are unclear.

Examples:

• define who owns the next step;
• establish an escalation rule;
• remove unnecessary approvals;
• clarify what information is required before work begins.

SOP, checklist, or template

Use this when the work is repeatable but inconsistent.

Examples:

• intake checklist;
• standard customer response;
• job-preparation checklist;
• quote or proposal template;
• end-of-day review procedure.

Reporting or visibility

Use this when the problem is not knowing what is happening.

Examples:

• shared status view;
• simple operations dashboard;
• standardized weekly report;
• clearer source-of-truth rules.

Tooling cleanup or integration

Use this when existing tools create duplicate work, are underconfigured, or do not share information cleanly.

Examples:

• remove a duplicate spreadsheet;
• configure existing software more completely;
• connect two systems that already have stable inputs and outputs;
• consolidate overlapping tools.

Automation or AI assistance

Use this when the process is already reasonably clear and the work is repetitive, rules-based, or language-heavy.

Examples include:

• routing structured requests;
• creating first drafts from known inputs;
• summarizing notes;
• extracting information from repeatable documents;
• generating reminders;
• moving validated information between systems.

Automation is strongest when it supports a stable process. If people still disagree about how the work should happen, clarify the process first.

A practical 30-minute friction review

Choose one recurring workflow and review it with the people who actually perform the work.

Ask:

• What starts this workflow?
• What information must be available before it can move forward?
• Who owns each stage?
• Where does the work wait?
• Where is information copied or re-entered?
• Where do people ask the same questions repeatedly?
• Where does rework occur?
• What requires owner or manager involvement?
• What status information is difficult to see?
• Which single improvement would make this workflow noticeably easier?

Do not try to redesign everything in one meeting. The objective is to identify one or two patterns worth investigating further.

What not to automate yet

Be cautious about automating a workflow when:

• the process changes depending on who performs it;
• required inputs are frequently missing;
• decision rules are not understood;
• exceptions are more common than the normal path;
• nobody agrees which system is authoritative;
• the business is still changing the underlying service or operating model.

In these cases, automation can hide the real problem or create a more complicated one.

A useful first outcome

A good friction review does not need to produce a major transformation plan. It should produce a clear statement such as:

That is enough to decide what to investigate next.

The accompanying Operational Friction Review Checklist provides a compact version of this process that can be used with a team or during an internal workflow review.