Every enterprise has more automation opportunities than budget. The question isn't whether to automate — it's where to start. And the answer matters more than most organisations realise.
The wrong first project creates expensive scepticism. The team invests months, delivers modest results, and the organisation concludes that "automation doesn't work for us." The right first project creates momentum — proven value, executive confidence, and organizational appetite for more.
The Selection Framework
After deploying dozens of automation projects across APAC, we've developed a framework for identifying the right starting point. It evaluates four dimensions:
1. Volume × Repetition
The best automation candidates are processes that happen frequently and follow similar patterns each time. An invoice processing workflow that handles 500+ invoices monthly is a better candidate than a strategic planning process that happens quarterly.
The math is simple: automation value = (time saved per instance) × (frequency). A process that saves 15 minutes per instance but runs 1,000 times monthly delivers more value than one that saves 2 hours per instance but runs 10 times monthly.
2. Rule Clarity
Not every process can be articulated as rules. Some genuinely require human judgment, contextual awareness, or creative problem-solving. The best automation candidates have clear decision logic — even if it's complex.
A three-way invoice match (PO, goods receipt, invoice) has clear rules. A credit risk assessment for a new market entry does not. Start with the former.
This doesn't mean the process has to be simple. Complex rules are fine — multiple validation checks, multi-jurisdiction compliance requirements, exception routing based on ten different criteria. The key is that the rules can be defined, even if they're intricate.
3. Pain Visibility
Choose a process whose inefficiency is widely felt. If the CFO mentions invoice processing delays in every quarterly review, that's your first project. If the compliance team stays late every month-end to compile reports, that's your first project.
Visible pain creates executive sponsorship, user cooperation, and celebration when results arrive. Automating a back-office process that nobody complains about — even if it's technically a good candidate — generates no organisational momentum.
4. Integration Simplicity
Your first project should not require deep integrations with twelve systems. It should connect to one or two existing systems, preferably ones with APIs or standard data exports.
Integration complexity is the number one cause of automation project delays. For your first project, you want to demonstrate value quickly — not spend six months building connectors.
The Scoring Matrix
Rate each candidate on a 1-5 scale across all four dimensions. Multiply Volume × Repetition by Rule Clarity by Pain Visibility by Integration Simplicity. The highest score is your best starting point.
We've seen this matrix cut through months of committee debate in ten minutes. It replaces subjective arguments with a structured assessment that everyone can agree on — or at least disagree constructively about.
Common First Project Mistakes
Going too big: "Let's automate the entire procure-to-pay cycle" is not a first project. "Let's automate invoice data extraction and three-way matching" is.
Going too safe: Automating a trivial process to "test the waters" produces trivial results. Choose something meaningful enough that the results will be noticed.
Choosing politics over value: Sometimes the best automation candidate is in a department that's politically difficult. Don't let organisational dynamics override the analysis. If the finance team's process is the best candidate but the supply chain team is more willing, start with finance anyway — the value will overcome the politics.
Ignoring the humans: The people currently doing the manual work need to be involved in the project. They know the edge cases, the workarounds, and the things the process documents don't capture. And they need to understand that automation isn't about replacing them — it's about freeing them for higher-value work.
After the First Win
Once your first project delivers measurable results, three things happen:
- Executive confidence increases. You've proven that automation works in your environment, with your systems, with your people.
- The pipeline grows. People across the organisation start bringing automation ideas to you, because they've seen what's possible.
- The next project is easier. You've built integration patterns, established governance processes, and developed organisational knowledge that accelerates everything that follows.
The first project isn't just about the value it delivers. It's about the capability it builds. Choose wisely.