Tried automating a business's whole workflow. Failed miserably.
We found a company using 8 employees to do manual work (mainly outreach focused). We said "This whole thing can be automated". They were shocked. The idea of not paying for and managing 8 employees, not paying for electricity and rent. It felt like a burden was removed from their shoulders. Had a meeting, got an idea of the workflow, and got to researching. We found out that yes EVERYTHING could be automated, but we kept researching for 2 days straight. Because there was just this one HUGE issue: it was too costly and slow. Much slower. That beats the whole point if the new "solution" ends up being slower and making them less revenue. The bottleneck was today's browser agents being slow, and the initial part of their workflow required interaction with sites. The entire workflow became not possible because of this. And the cost kept skyrocketing as AI is expensive. We took a step back, and realized that no, we don't have to automate everything, like the gurus always say, that a 50 node n8n workflow won't matter if you don't save them time and money. They always advise to keep it simple, and that's what we needed to do. Hence, we proposed 3 solutions that all worked independently and could work in kahoots too. What outcome would they give? They would increase the cost from $3k to $5k, but why did we propose something that would increase costs? Becayse the revenue went from $6k to $13k. That is a 1.6x increase in cost for more than double the revenue. This made us learn. Don't try to automate everything. If it can be automated doesn't mean it should be automated. Find gaps, find ways to make their workflow smoother, easier. Allow them to get double the work done with half the effort. That is what Automation truly portrays.