I timed it. Manually scheduling one Facebook post in Meta Business Suite, with everything already open and the post already written, took just under two minutes. A script that does the same thing took six seconds.
That is the kind of gap I look for in a business. This time, I found it in my own.
The Bottleneck
I post to two Facebook pages for my marketing consulting business in Bradenton. My target is roughly six posts a week between them. At two minutes per post under ideal conditions, that is twelve minutes a week just on scheduling. Nothing complicated. Just friction. The small kind that compounds.
Multiply two minutes by 24 posts a month and you get close to fifty minutes. Do that for a year and it is hours. Not a crisis, but a real number. And real enough to fix.
What I Had Built
Once I identified the bottleneck, I knew what kind of solution to look for. Facebook has a direct API. A script that connects to it can schedule a post with one command. No login screen, no scheduling interface, no third-party platform in the middle.
I used Claude Code to write the script. I described what I needed: post to two pages, accept a message and a scheduled time, confirm it worked. Claude Code built it. I tested it. It ran in six seconds.
That distinction matters. I did not write the code. My job is not automation building. My job is knowing what to look for and figuring out what kind of fix fits the situation. The build is secondary to the diagnosis.
Why Not Zapier or Make.com
Most automation tools for social media run on top of platforms like Zapier or Make.com. They work. I am not saying they do not.
But there is a dependency layer that does not always get discussed. If you build automation on a third-party platform, your workflow runs on someone else’s infrastructure. If that platform changes pricing, breaks a connector, or discontinues a feature, your automation breaks with it. You are not building something you own. You are renting a process.
A direct API call does not have that problem. The script talks to Facebook. That is the whole chain. Nothing else sits in the middle.
The Math on Time Saved
At six seconds per post, 24 posts a month takes about two and a half minutes total. Compared to the manual approach, that is roughly 45 minutes a month returned to the business. Over a year, the difference is around eight hours on the low end. Closer to twenty if you factor in realistic starting conditions: finding the right tab, navigating to the planner, beginning from cold.
That is a meaningful return on a few hours of setup work. And it is the kind of math that is worth running before you decide a task is too small to automate.
What This Has to Do With Consulting
This is a small example of what I do with clients. Not building scripts. That is not what I offer. But finding where the time is going and figuring out what a realistic fix looks like. That is the work.
A lot of business owners know they have inefficiencies. They just have not stopped long enough to look at where they actually are or what the options look like. This Facebook scheduling example is simple. The principle behind it is not: stop the process, measure it, and figure out if the cost of the fix is worth it.
In this case it was. Six seconds versus two minutes, multiplied across a year, is a meaningful number. And it came from identifying the problem first, not from buying a tool and hoping it helped.
I have written before about the difference between AI consulting and AI automation and why mixing those two things up tends to lead to the wrong investment. This is a live example of that line in practice.
Questions to Ask Before You Pay Someone
A lot of the AI automation being sold to local business owners right now is built on Zapier or Make.com. That is not automatically a problem. But if someone is charging you to build automation for your business, you should know what it runs on.
Ask what platform the automation actually lives on. Ask what happens if that platform changes its pricing or discontinues a feature. Ask who owns the integration when the project is done. Ask what you would need in order to run it yourself if you had to.
Those questions are not about being difficult. The answers tell you exactly what you are buying. If the answers are vague, that is useful information too. I have written about what small businesses actually need from automation and where the line is between a tool that helps and one that adds complexity without a clear payoff.
Finding Your Own Bottlenecks
Most of the time, the opportunity is not obvious until someone looks at it. That is what getting a business ready for this kind of work actually involves: slowing down, mapping where the time goes, and figuring out which problems are worth solving and which ones are not.
I use these tools in my own business. I am not just advising clients on things I have not tested myself. If you want to think through where your own business has gaps like this, I offer a free clarity call. No pitch. Just a real look at how your business runs and where the clearest opportunities are.