Back to Blog
Automation

Automation Check: How to Identify Processes with Real Potential

December 28, 2025
9 min read
Jonas Höttler

Automation Check: The Difference Between Theory and Practice

"We can automate that!" – This phrase comes up in meetings more often than ever. And it's often true. The problem: Not every automation makes sense. Some cost more than they save.

Why Half of All Automation Projects Fail

The statistics are sobering: Around 50% of automation initiatives don't achieve planned results. The most common reasons:

1. Wrong Process Selected

  • Too complex, too many exceptions
  • Executed too rarely
  • Too many manual decisions required

2. Wrong Expectations

  • Overestimated time savings
  • Underestimated implementation costs
  • Ignored maintenance efforts

3. Lack of Preparation

  • Process not standardized
  • No clear rules defined
  • Poor data quality

The 5 Criteria for Automation Potential

Criterion 1: Repetition Frequency

Rule of thumb: The more often a process runs, the higher the potential.

FrequencyPotentialExample
Multiple times dailyVery highInvoice receipt verification
DailyHighDaily reports creation
WeeklyMediumInventory updates
MonthlyLowMonth-end closing
AnnuallyVery lowYear-end closing

But note: Even rare processes can be worthwhile if they're very time-consuming.

Criterion 2: Rule-Based Nature

Automation needs clear rules. Ask yourself:

  • Are there clear if-then relationships?
  • Can decision criteria be documented?
  • How often are there exceptions?

Automatable:

  • "If invoice amount > €1000, then approval by department head"
  • "If inventory < minimum quantity, then trigger order"

Difficult to automate:

  • "If the customer is important..." (What does "important" mean?)
  • "Decide by feeling..." (Which feeling?)

Criterion 3: Digital Availability

The process must be digitally representable:

Easy to automate:

  • Data exists in systems
  • Interfaces (APIs) available
  • Standard formats (CSV, XML, JSON)

Hard to automate:

  • Handwritten documents
  • Phone calls as input
  • Decisions based on gut feeling

Pro Tip: Use our Digital Maturity Assessment to determine how digital your processes already are.

Criterion 4: Error Susceptibility

Manual processes with high error rates are ideal candidates:

Typical error sources:

  • Typos in data entry
  • Forgotten steps
  • Inconsistent execution
  • Copy-paste errors

Rule of thumb: With an error rate > 2%, automation almost always pays off.

Criterion 5: Time Effort and Costs

Calculate the true costs:

Annual process costs =
  (Processing time × Hourly rate × Frequency) +
  Error costs +
  Opportunity costs

Calculate it specifically: Our Process Cost Analyzer helps you determine true costs including hidden costs.

The Automation Scoring Model

Rate each process on a scale of 1-5:

Criterion1 (low)5 (high)
FrequencyAnnuallyHourly
Rule-basedMany exceptionsClear rules
Digitization levelPaper-basedFully digital
Error susceptibility< 1%> 10%
Time effort< 10 min/month> 20 hrs/month

Interpretation:

  • Score 20-25: Automate immediately
  • Score 15-19: Good candidate, examine in detail
  • Score 10-14: Possible, but calculate ROI precisely
  • Score < 10: Probably not worthwhile

Which Automation Technology Fits?

Different approaches suit different processes:

RPA (Robotic Process Automation)

Suitable for:

  • Existing systems without APIs
  • Data transfer between legacy systems
  • Rule-based screen work

Examples: UiPath, Automation Anywhere, Power Automate

Workflow Automation

Suitable for:

  • Multi-stage approval processes
  • Task distribution and tracking
  • Document workflows

Examples: n8n, Make (Integromat), Zapier

AI-Powered Automation

Suitable for:

  • Document processing (OCR + NLP)
  • Email classification
  • Data extraction from unstructured sources

Examples: Claude/GPT APIs, Google Document AI, AWS Textract

Low-Code/No-Code Platforms

Suitable for:

  • Quick prototypes
  • Simple integrations
  • Citizen development

Examples: Retool, Bubble, Airtable Automations

The Most Common Automation Mistakes

Mistake 1: Starting with the Most Complex Process

Don't start with the biggest, but with the simplest process. Quick wins create momentum.

Mistake 2: Aiming for 100% Automation

80% automated with human review is often better than 100% with error risk.

Mistake 3: Ignoring Maintenance

Every automation needs maintenance. Plan 15-20% of initial effort per year for upkeep.

Mistake 4: Not Involving Employees

The people who execute the process know it best. Involve them.

Your Automation Roadmap

Phase 1: Inventory (1-2 weeks)

  • Create list of all repetitive processes
  • Collect rough time estimates
  • Identify pain points

Phase 2: Scoring (1 week)

Phase 3: Business Case (1-2 weeks)

Phase 4: Pilot (4-8 weeks)

  • Automate one process
  • Measure, learn, optimize
  • Document for scaling

Free Automation Check

Want to know if your process is suitable for automation?

Our tool checks:

  • Automatability based on the 5 criteria
  • Suitable technology recommendation
  • Estimated savings potential
  • Next steps

Check Automation Potential Now

Conclusion: Systematics Beats Gut Feeling

The question isn't "Can we automate this?" – usually yes. The question is: "Should we automate this?" And only systematic analysis answers that question.

With the right criteria and a clear scoring model, you avoid expensive mistakes and focus on automations that truly create value.


Identified processes with high automation potential and ready to get started? Our AI Adoption Audit helps you with strategic implementation – from tool selection to implementation.

#Workflow Automation#Business Process Automation#Process Optimization#RPA#Automation

Have a similar project?

Let's talk about how I can help you.

Get in touch