Now the final Part of our Audience Fundamentals series!

Today, we're putting on our data scientist hats. Well, the AI will be wearing the hat - don't worry, I won't make you do any complex calculations. It’s not like I could…

Let’s get started:

Base level strategy

Here's the basic idea: Do more of what works and less of what doesn't.

OK we’re done here right? Probably not…

Sounds simple, right? Yet you'd be surprised how many people don't follow this golden rule. They either:

1. Have a success and leave it at that (one-hit wonders)

2. Have flops and just... keep going! (Definition of insanity)

Remember when we talked about the power of volume? This is where it really pays off. All that content you've been pumping out?

It's not just content - it's data - and used properly it’ll help us avoid these mistakes.

Dealing with Successes: Rinse and Repeat

First up, your winners.

When you hit a winner, here's what I want you to do: Do it again.

No, I don't mean posting the identical post or video. That’s duplicate content and will get you dinged. But rewrite or rerecord it. And post it again.

"But Kyle," I hear you cry, "won't people get bored seeing the same content?"

We don’t care. Because there are three situations here. None of them matter:

1. If they've seen it before, they'll move on. No harm done - it's one of 1000 posts they'll see today. They’ll live.

2. If they've seen it before and are happy for a repeat - great! They get to review, refresh, or see it with new eyes.

3. or most likely scenario: It's a new person. Horray.

Let me show you what I mean. Check out this screenshot from my TikTok:

See these two top videos? Literally the same concept. 5M+ views.

That reminds me, I need to do the same video again…

Most of the time, you're reaching new people.

Those 5M views? 99% were new viewers of my content. They hadn't seen the first video. They don't know or care that the second is similar.

So, repeat your winners. Add graphics, change the lighting, wear a silly hat. Doesn't matter - add a little variation and do it again. And keep doing it (and similar content) as long as it works.

Dealing with Flops: Pivot and Learn

Now, for the not-so-fun part: failing posts. If something isn't working, you need to make a change. I know right…ugh.

The question is: When do you know if something is working or not? This is why I personally think TikTok is great - each post has its own chance to shine. It'll be shown to 200 people, and if it's engaging, it'll be shown to more. If not, it'll stall in "200 view jail". Very clear cut! You know what’s crap and what isn’t!

On other platforms, you have to judge win/loss compared to your other content. Does one spring ahead of the average? That's a winner. Does one underperform? It's a loser.

Assess this weekly to work out what’s not helping you…and then: do less of the losers. It's that simple. Really.

The Power of This Approach

If you follow this basic plan, your content will get better - it can't not. You're using feedback to follow what people want. It's like evolution, but for your content strategy.

Once you have sufficient data, we can add another layer of feedback. We'll use a prompt to review your social media account performance and give you insights.

But before we can do that, we need the data. Where you get the data depends on the platform - Google how to download your data for the respective platform. Or ChatGPT the info!

Generally the data comes in a CSV file. Get as long a period as possible.

Once you have your data, feed it into this prompt:

You are an expert social media analyst with a deep understanding of content performance metrics across various platforms. Your task is to analyze the provided social media data and offer insights to improve the user's content strategy.

I have a CSV file containing my social media performance data. Please analyze this data and provide insights to help me improve my content strategy. Follow these steps:

1. First, tell me you're ready for me to upload or paste the CSV data.

2. Once I've provided the data, perform the following analyses:

Quantitative Analysis:
a) Identify the top 10% performing posts based on engagement rate (likes, comments, shares, etc.).
b) Calculate the average performance metrics for all posts.
c) Determine the best and worst times for posting based on engagement.
d) If hashtags are used, identify the top 5 most effective hashtags.
e) Analyze the correlation between post length and engagement.

Qualitative Analysis:
a) Identify common themes or topics in the top-performing posts.
b) Analyze the tone and style of the most engaging content.
c) Determine what type of content (e.g., questions, how-tos, personal stories) performs best.
d) Identify any patterns in the use of media (images, videos, text-only) in successful posts.

3. Based on your analysis, provide the following insights:

a) Top 3 content topics or themes that resonate most with the audience.
b) 3 specific characteristics of high-performing posts.
c) 3 common elements in underperforming posts to avoid.
d) Recommendations for optimal posting times and frequency.
e) Suggestions for hashtag usage (if applicable).
f) Advice on content length and media use.

4. Offer 5 actionable recommendations to improve overall content strategy, based on what's working well.

5. Suggest 3 new content ideas that align with the successful patterns you've identified.

6. Provide a brief summary (3-4 sentences) of the overall content performance and the most critical area for improvement.

Please present your analysis in a clear, organized manner, using bullet points where appropriate. Avoid technical jargon and focus on providing actionable insights.

This prompt (plus your data) will come out with a whole load of info for you. For example a qualitative analysis:

It also identified my best hashtags:

And ends with general recommendations:

The Last Word

You're not just creating content anymore - you're running experiments.

Think of each post as a hypothesis about what your audience wants. Some will fail, some will succeed, but each one teaches you something valuable.

That means that your “failed” posts are actually really useful. They give you more intel.

Hopefully that helps release you from the curse of making only “perfect” content. Because that’s impossible - and only leads to paralysis.