Every single week. I make 25% of my videos something new, because I never know what's going to work. That's how I found carousels: I tried them for one week, they outperformed everything, and now I post them every day.
How much of your content should be an experiment
Someone at Friday Strategy Sips asked me how I noticed my carousels were outperforming my talking-head videos, and what the data was telling me.
It started by accident. I was testing out different things.
Every single week, I do 25% new videos and try something new. I never know. Carousels were what I tried that week, and they just outperformed. Now I do those every day.
Why you have to test content on your own account
You have to test it out on your account. When my carousels went out, the saves were extremely high, and I noticed something. When my saves are high, I make money from it. Either people show up here, they purchase something, or they DM me.
I could tell you what's worked for me, but until you actually test it and get 30 days of data, there's no way to know. Many times, the blanket statements haven't been true for me in my business.
I manage around 5 social media clients personally, and not one of them has been successful on the same platform. They hate my videos on LinkedIn, but on TikTok they love them.
How to test new content without confusing your audience
Change one thing at a time. That's how you can really see what's working.
You can A/B test your content. Do 50% to one audience and 50% to the other. The way the algorithm works now, you can test without confusing your followers. I do the same: I'm either talking to someone just starting, or to people about client retention and boundaries.
Test it for a full 30 days, and go all in on whatever bites.
Conclusion
Save a piece of every week for something new, give it 30 days of data, and let your own numbers tell you what to do more of.
Have a question you need answered? Every Friday at 6:30 PM Eastern, I do free live business consulting at Friday Strategy Sips. Just show up and ask.