← Back to all posts

The Power of Positive Sentiment: Tatcha's Social Media Strategy for Successful Product Launches

Does the sentiment toward a new product launch (which could be observed through Instagram likes) foreshadow the future sales of that product? If we find evidence that the positive sentiment toward a new product's content (from its launch until the present) correlates with the product's current total sales quantity, it suggests that the amount of positive sentiment toward a product over time can drive sales. Or, if we find evidence that the first few weeks of positive sentiment toward a new product's content correlates with the product's current total sales quantity, it suggests that social media can predict whether a new product will be successful.

The Power of Positive Sentiment: Tatcha's Social Media Strategy for Successful Product Launches

Purpose of Case Study

One of my favorite Instagram pages of all time has to be Tatcha’s — I’m a huge sucker for a gorgeous aesthetic and color scheme. And as I was scrolling through their feed, I noticed that every time they create a new product, they typically make a series of posts to advertise the launch.

__wf_reserved_inherit

So it had me wondering — does the sentiment toward a new product launch (which could be observed through Instagram likes) foreshadow the future sales of that product?

If we find evidence that the positive sentiment toward a new product's content (from its launch until the present) correlates with the product's current total sales quantity, it suggests that the amount of positive sentiment toward a product over time can drive sales. This implies that ongoing positive sentiment can influence and potentially increase a product's sales over time.

Or, if we find evidence that the first few weeks of positive sentiment toward a new product's content correlates with the product's current total sales quantity, it suggests that social media can predict whether a new product will be successful. This implies that initial positive sentiment can serve as an early indicator of a product's future sales performance.

Both of which are interesting hypotheses.

Analysis 1

To test our first hypothesis, we looked at Tatcha’s entire feed, starting in 2020, and identified the first 7 product launches that the brand made content for. We then added together the likes of every promotional post for each product starting from its launch date and ending in 2024.

We scanned Sephora’s Website and Tatcha’s Website (their two biggest sales channels) for product reviews of each product and made a general assumption (so take these results with a grain of salt) that the total number of reviews per product represented approximately 10% of total sales.

Then, we graphed each product’s total Instagram likes (representing positive sentiment) against the estimated number of sales for that product. As a result, we ended up with this graph.

__wf_reserved_inherit

What we’re seeing here is that the patterns between both metrics actually do mirror one another pretty well. The last two products, Indigo Midnight Repair and The Texture Tonic, seem to deviate slightly from the pattern, but it may be because these two products are relatively newer than the others.

When we include every product in our calculation, it seems that on average, the sales for a product lies at around 31% of that product’s total positive sentiment measured as likes on Instagram. If we exclude Indigo Midnight Repair and The Texture Tonic, that number rises slightly to 34%.

Analysis 2

We then graphed estimated sales against only the total likes for the first 3 Instagram posts marketing each new product. Our goal here was to see if initial positive sentiment could predict the eventual success of a product. Our graph ended up like this.

__wf_reserved_inherit

As you can see, the results are much more iffy here. The patterns don’t match one another at all, sometimes going in the opposite direction. This implies that our second hypothesis might not be true after all.

Meaningfulness of Results

Looking back on a results, we can make two inferences from our data.

1) Yes, ongoing social media content that drives positive sentiment may increase sales for a new product over time.

But 2) No, loads of initial positive sentiment toward a new product won’t necessarily mean it’ll lead to tons of sales in the future — and it unfortunately won’t be able to predict anything right away.

Our results might even suggest that continuing to post more content (quantity-wise) about a specific product could boost its sales, seeing as positive engagement typically increases with the quantity of content made. This makes sense considering customers may need to see a product a couple of times before locking in a purchase.

__wf_reserved_inherit

And so although social media engagement might not be able to predict how successful a new product will be, consistently making great content that helps people form positive associations toward that product will, over time, drive more and more sales and help bring that brand-new item to the top.