What #2015BestNine can tell you about your brand
I just finished reading a great book called Listening Brands: How Data is Rewriting the Rules of Branding and the recent hashtag going around #2015BestNine got me thinking about what it means for brands.
The basic premise of the book is that the old inituition, big idea, "talk at the customer and make them like you" point of view in advertising is old fashioned. Instead, you should listen to your customers and engage in a conversation with them proving that you provide value for them and deserve to be a part of their lives. One way to do that is with data.
So what does this have to do with #2015BestNine? #2015BestNine gives you a mini-dataset that you can base some limited conclusions on. For example, do certain filtered photos perform better than others? Is there a certain type of photo that performs better than another?
I will caveat that clarify of the data goes down as the volume of photos go up. A high school friend of mine is a model and instragrams alot, taking many shirtless photos in his underwear. His best nine was filled with these and while it gives a strong signal that his followers love to see him with his shirt off, the sample size is too small to give any other insights to his other photos.
Below is my 2015 Best Nine. I only started sharing on instagram about 6 months ago with an average posting schedule of once every 3-4 days(~57 posts) with a small amount of followers(~ 90). So 9 posts will be statistically significant as it is roughly a 15% sample.
My Best Nine tells me that people see me as athlethic. Seven out of nine pictures relate to fitness, running, or obstacle course runs. If I were trying to maximize likes with this group, I would endeavour to post more pictures from runs and maybe think about what peripheral stuff I could post (going to the gym, etc). One of the best nine is a darkly lit video of myself doing burpees in front of a Christmas tree from four days ago. I think it was either the timing of the post or the novelty of the Christmas tree that made it perform well.
The only data that is too small is no data al all.