Analytics is better than A.I.
Imagine this: you’re in a new town, and you’re looking for a place to eat — not a franchise, but something local.
It just so happens that a bunch of people you know live, or have lived in this town — people who understand what you like and don’t like (what’s important to you, and your general financial and dietary situation).
In looking for a place to eat, which of these would sway your opinion more:
- Opinions from random people in town
- Top results in a Google search
- Recommendations from your friends
For most people, it’s going to be recommendations from their friends, right? For all the reasons I listed above.
Strangers don’t know you, so they’ll basically just tell you their favorite restaurant.
Google and other search engines don’t know you personally. They rely on averages and a different set of signals.
I’m speaking at Podfest later this week and I noticed something interesting about the schedule: there are at least a dozen talks on how A.I. can help you with your podcast.
But as far as I can tell, there’s only one talk on using analytics to measure the success of your podcast.
Analytics Gives Your Positive Signals from Your Audience
Over the last few months, I’ve been thinking a lot about analytics. I revamped my CALM framework for easier content creation, and the M stands for, “Measure.”
You should measure the success of your content so that you understand what resonates with your audience. This is a far better signal than A.I. guessing what you should write about based on averages.
Now, don’t hear what I’m not saying. I use A.I. all the time for ideation and research. But A.I. can’t tell me what my audience wants more from me — not alone, anyway.
In Episode 454 of Streamlined Solopreneur, I talk about how you can leverage analytics, while still experimenting, to come up with a content schedule that your audience actually wants from you:
Combine Analytics with A.I. for the Right Mix
There’s a reason YouTubers and Short Form video creators tend to produce the same type of content — it follows a very familiar script with what’s basically a different punchline.
It’s the same reason Sitcoms, procedural cop shows like Law and Order, and House MD are all formulaic but still popular: people like familiarity.
They like knowing what to expect, with the occasional surprise.
You can use A.I. to figure out the occasional surprise, while looking to your analytics for the familiar.
Relying solely on A.I. is like asking 100 people on the street where you should go to dinner. Or Googling, “What should I write about?”
Analytics is like getting recommendations from friends — they know what you’re about, and they’re ready to recommend something similar.
Just A.I. isn’t going to cut it. I suggest you capture good ideas as they come to you (the C in CALM). But when you’re stuck, using analytics in conjunction with A.I. is better than just asking ChatGPT what you should write about:
I’ve been writing about TOPIC for two years, and my best performing content are these 10 titles: (LIST TITLES). Can you please help me come up with 20 ideas similar to these?
A.I. Can’t Replace Real People
I almost called this piece “A.I. won’t make you better at your craft.” I know it’s a hot topic these days, but we need to remember that YOU are the content creator — whether it’s thought leadership for your business, or a new episode of your podcast.
And if you’re not starting completely from scratch, your audience is there, ready and willing to learn from you — and recommend you to the right people.
Why would you outsource such a crucial part of the content creation process to A.I. when you have a goldmine of data at your fingertips?