Automation is where AI stops being “interesting” and starts being “this saves us real money.” But automate the wrong thing, or automate a broken thing, and in our experience it costs you more than doing nothing at all.
The owners who win aren’t the ones who automate the most. They’re the ones who are deliberate about what they automate first, and in what order.
The golden rule: never automate a broken process.
If a process is a mess, automating it just makes the mess happen faster and cost more. Automation amplifies whatever it touches. The same is true of the data underneath it: Gartner points to poor data quality as a top reason generative AI projects stall and get abandoned after proof of concept (Gartner). So fix or simplify the process first, then automate it. Amplifying garbage just gives you more garbage, sooner.
Is this a good candidate? Run it through four questions.
A genuinely good automation candidate is usually all four of these at once. Run any task through the checks below. If it stalls on the broken-process question, stop and fix that first. If it needs real human judgment, keep a person in charge. What survives the gauntlet is what you automate first.
flowchart TD
A(["A task you might automate"]) --> B{"Is the process broken?"}
B -->|Yes| F(["Fix it first, then come back"])
B -->|No| C{"Repetitive and rule-based?"}
C -->|No| G(["Needs judgment, keep a human on it"])
C -->|Yes| D{"Eating real hours, and checkable?"}
D -->|No| H(["Low payoff, leave it for now"])
D -->|Yes| E(["Automate this first"])
So the four things you’re looking for:
- Repetitive. It happens often, the same way each time.
- Time-consuming. Your people spend real hours on it.
- Rule-based or pattern-based. There’s a logic a machine can learn.
- Tolerant of occasional checking. A human can review the output where it matters.
Where owners get fast, visible wins.
- Reading and sorting documents. Invoices, forms, applications, receipts. Pull the information out, put it where it belongs.
- Answering routine questions. For customers and for staff, instantly, around the clock, in Arabic and English.
- Moving data between systems. The dull, error-prone work that quietly eats hours.
- First drafts. Quotes, replies, summaries, reports that a person then finalizes.
What to be careful with.
- Anything where a mistake is expensive and you can’t undo it. Keep a human in the loop.
- Work that needs real judgment, empathy, or relationships. AI assists here. It shouldn’t run the show alone.
- Anything touching sensitive personal data without proper care for PDPL and security.
Sequence it by value and ease.
Start with something that’s both high-value and straightforward. An easy, visible win builds confidence and frees up time and budget for the harder, bigger automations next. Don’t open with your most complex, mission-critical process. Earn your way there. It’s also the surest way to avoid joining the projects Gartner expects to be abandoned after proof of concept, at least 30% of them by the end of 2025 (Gartner).
Want to know what to automate first in your business? SDCG helps owners find the automations that pay off fastest, and warns you off the ones that’ll cost more than they save. Independent advice, grounded in what actually works. Book a free 30-minute review.
Sources