I think you need to focus on testing what works and what doesn’t. Don’t define your strategy but test possible strategies, see what works and what doesn’t, and iteratively move towards what works. The good news is that you have a lot of followers to work with – that means data! Look to what you already have.
What I would do is create a sheet in Google Sheets. Create columns, column 1 is post id, column 2 is a summary, column 3 is your estimated quality, column 4 views, column 5 likes, column 6 comments, column 7 is the actual quality, column 8 is the date of the post.
Let me define those terms so you know what I’m talking about:
1. post id – A link to the post so you can easily review your spreadsheet and click on the link.
2. summary – describe the post in one or two words. You want broad categories that effectively say what the post it. Think: “Mom picture” or “Beauty product”.
3. estimated quality – Create a scale 1-5 and just give it a number based on how “good” you think the post is. Don’t overthink it, but it will be good for you to get a handle on what constitutes a “good” post.
4. views – Use Instagram’s analytics to find out how many views each post got.
5. likes – How many likes did the post get
6. comments – How many comments did the post get
7. Actual quality – This can get more complex, but you need to define a method for grading the posts. Use columns 4, 5, and 6 and define a formula. Something like X * (views / avg views) + Y * (likes / avg likes) + Z * (likes / avg likes) – where X, Y, and Z are coefficients that you pick to weight the relative importance of each value to you. Then, normalize the scores and go back to the 1-5 scale so you can compare it to your estimated quality in column 3.
Fill in the spreadsheet with all of the previous posts on the Instagram account. Then, start making graphs. You want a bar graph showing the actual quality of topics. You want a graph showing posts with the highest likes/views ratio. You will also want to graph followers over time and actual quality over time. Pick the metrics that concern you and track them.
With the graphs you’ll know what “summaries” perform best for you. See if you can start to dial that in. Create more of the content that did well. Track your quality over time. If it is increasing, that suggests your current choices are working and you should continue. If it is stagnant or decreasing that suggests you need to change.
With each post read the comments and try to figure out why the response to the post was what it was. If this was a post that performed well, try to figure out why. What are people saying about it? How did you rate the quality subjectively? How did it rate objectively? Based on your intuition, talking to the influencer, and reading the comments, come up with a theory that explains what happened with the post. This post did poorly because we posted it at the wrong time of day, because the shot had low production value, because it looks too manufactured, because whatever. Then test that theory going forward – if you think the shot was too “produced” try something more candid next time, etc. Keep track of your theories and the results over time.
In this way you can try different things over time, isolate what is working and do more of it, and discard what doesn’t work. You need to be flexible and adaptive, but also track key figures over time so that you aren’t just groping about in the dark but intelligently going somewhere.