Choosing What to Automate First
Picking the Best First Project, Not Just the Biggest One
If you've run a few processes through "Is Your Business Ready for Automation?" and come out with two or three reasonable candidates, this is the next decision: which one do you actually start with?
It's tempting to pick whichever one looks like the biggest win on paper. That's often the wrong instinct for a first project specifically. Your first automation project has a job beyond solving the process itself: it needs to teach your team how this works, build confidence, and create a visible result that makes the second project easier to get buy-in for. The highest-ROI process isn't always the best process to learn on.
The Four First-Project Criteria
List your candidate processes (the ones that scored "likely ready" or higher), and score each one low / medium / high on these four questions:
1. Automatability — How cleanly does this match the readiness signals (repetition, rules-based logic, volume, data availability)? A process with cleaner logic and fewer exceptions will move faster and hit fewer surprises — valuable when your team is still learning the process. 2. Reversibility — If this doesn't work the way you expect in the first few weeks, how easily can you roll it back to the manual process without real damage? A first project with a low "blast radius" lets you learn without betting something critical on getting it right immediately. 3. Visibility of the win — Will the people affected — your team, your customers, or both — actually notice and feel the improvement? An invisible back-office fix might be real progress, but it won't build the internal momentum a first project needs to earn support for the next one. 4. Learning value — Will doing this one teach you something useful for the next automation project — about the tool you're using, about gaps in how your processes are documented, about where your data actually lives? Your second project should be easier because of your first one, not a repeat of the same discoveries.Putting It Together
A candidate that scores medium-to-high across all four — even if it isn't your single highest-ROI process — is usually the better place to start than a high-stakes, low-reversibility process that happens to promise the biggest number.
Save your highest-ROI-but-higher-risk candidates for your second or third project, once your team has one successful, visible, well-understood automation behind them. That sequencing — start smaller and build credibility, then tackle the bigger one — is also exactly what "How to Pilot an Automation Project" in this Resource Centre walks through in practice.
A Simple Worked Example
(Illustrative only — not a real client's data.)| Candidate process | Automatability | Reversibility | Visibility | Learning value | Good first project? |
|---|---|---|---|---|---|
| Weekly invoice data entry | High | High | Medium | High | Yes — strong first choice |
| Customer onboarding emails | Medium | High | High | Medium | Good second choice |
| Core pricing-approval workflow | High | Low | Low | Medium | Save for later — too much is riding on getting it right immediately |
Notice the pricing-approval workflow might genuinely have the biggest dollar impact — but low reversibility and low visibility make it a poor place to start. That's the whole point of this framework: ROI matters, but it isn't the only thing that matters when the goal is a strong first project.
---
Want help applying this to your own candidate list? We can walk through the scoring with you in one short conversation. Book a free consultation →