Brands Rankings: pollution of hashtags and products missed without visual recognition on Instagram

Since the raise of social media, it has become crucial for brands to manage their presence on visual platforms such as Instagram. Traditional listening solutions follow hashtags to monitor brands’ popularity.

However, this methodology is fraught with problems.

First of all, Pollution of hashtags is viral on Instagram, i.e. a large majority of posts with a brand hashtag have in fact no relationship with that brand.
Second of all, a large majority of Instagram users do not include hashtags on their posts, i.e. brands miss products.

As a consequence, unless brands follow all of their clients’ accounts, it will be impossible to find their products.

As an example, let’s take the #Balenciaga on Instagram.

Here we just retrieved the most recent results for #Balenciaga on Instagram. The first 12 results were:

x3 Balenciaga

x1 Chanel

x1 Dior

x1 Hermes

x1 Gucci

x2 Fendi

x2 Nike

x1 other

Only 3 posts out of 12 were relevant for Balenciaga. But there is more: the majority of results did match Balenciaga’s competitors.

Why is there so much hashtag pollution? Have we been just unlucky with this specific search?

In fact, the root of the problem is caused by the way hashtags work. Ranking of posts is done according to popularity of hashtags. Because many users and vendors want to be noticed, they constantly “hack” trendy brands’ popularity to come up in searches.

That these hashtags are irrelevant to the content of the posts is secondary.

Methodology of Heuritech’s study

Heuritech’s visual recognition technology, based on the latest deep learning research, is trained to recognize hundred of branded products — such as the Nike Air Jordan — in images.

For this study, we analyzed 1.5 million of Instagram posts from April to July 2018 thanks to our image recognition algorithms.

We then analyzed the relationship between the products recognized and the hashtags mentioned.

We called it Content Missed each time a product from a brand was found by our visual recognition technology without a brand’s hashtag.
We labelled posts as Spam when our algorithms identified a competitor’s product instead.

Visual recognition for brands and products: Valentino’s case study

Let’s take Valentino as an example: how much are specific and general hashtags relevant to find products from this brand?

Specific hashtags being for example #valentinokittenheels while general hashtags are for instance #fashion or #instagood.

Here, searching for #valentino allows to retrieve only 10.99% of the Valentino products posted. Meanwhile, there is 90,07% pollution posts in the feed, caused by products from other brands.
General hashtags tend to be heavily polluted by spam, while specific keywords are too narrow to retrieve a significant volume of Valentino products.

The brand’s own hashtag is in between these two extremes, yet fails to provide satisfactory metrics.

Brands’ ranking

This phenomenon is in fact not specific to Balenciaga or Valentino. Hashtags for all fashion brands show high pollution while missing a lot of products at the same time. For these rankings, we analyzed 30 fashion brands.

On average, brands’ hashtags contain 77,5% of pollution on Instagram. Without visual recognition technology, brands miss on average 78,4% of their products.

Do you want to lift the veil on your brand’s Instagram presence, without spam and without missing any products?

Heuritech is here to help. Day and night, we collect and monitor all the content associated with your brand on Instagram. Our platform allows you to visualize moodboards of your products worn in real life to understand your customers better. You can also benchmark your products with competitors over time and measure your products’ adoption among influencers and consumers.

Drop us a line at to know more, or subscribe to our newsletter below!