How Visual Mentions Complement Social Media Monitoring?

How Visual Mentions Complement Social Media Monitoring?

Many companies are already monitoring online content. It turns out, however, that tools typically used to track mentions tend to underperform on platforms that were built to share photos or videos, like Instagram, Pinterest or YouTube. Fortunately, visual content monitoring can help where standard analytics are losing ground. DeepBrand allows you to track visual mentions of your brand and find specific objects in online content. Read on to see how visual content analytics fills in the gap in your online content tracking.

Is something wrong with mentions?

It’s nothing wrong with mentions! Monitoring keywords is a great way to discover conversations people have online, especially those regarding brands and products. Also, there is plenty of keyword monitoring tools that do a great job and are used by social media managers or PR experts on a daily basis.

Mentions give me a lot of data already, why would I need software to read images?

One obvious limitation to keyword tracking is that it’s text-only. Today, visual content makes for a major chunk of online conversations. For websites like Flickr or Instagram, monitoring keywords is simply not enough to capture the true meaning of posts. To get it, you need to be able to see what the pictures represent.

Of course, text-searching algorithms can provide some insight into what’s in a photograph as long as it’s properly tagged. But people don’t caption their content for algorithms — they do it usually for friends and followers.

The essence of a picture is to be seen, not read.

Using hashtags can be culture-dependent and does not always accurately describe the contents of an image. The most used hashtag in 2019 is #love, but such a keyword can be used to describe virtually anything, from a happy couple to one’s favourite meal, to a puppy, to a car one enjoys driving. On the other hand, some of you might probably never use any hashtag at all.

If text-based tools track mentions, then image-based tools, like DeepBrand, track visual mentions.

What exactly are visual mentions?

If you are a marketer, you probably know that if the recipients don’t have a chance to actually see your brand, they are lost contacts. Let’s suppose a friend publicly endorses the new L’Oréal mascara and you can actually see her showing the product. The impact of a visual mention on brand or product awareness will be much greater compared to when she would just mention the product in the comment. A proper presentation of the brand’s logo or product is therefore crucial for creating the right associations with your brand and, basically, making it more memorable.

A visual content analytics tool should be able to track visual mentions and provide you with additional data on brand placement. Such information may contain the size of the brand logo or a list of other objects detected in the picture.

What about finding objects other than brands?

As a marketer, you obviously want to reach out to your audience with your offer, in hopes of converting them into paying customers. So, if you run marketing activities for a bicycle brand, you will probably want to find people who are already active bikers. The easiest way to do it would be to type in keywords like bike or cycling in your social listening tool and run a search query. This will give you access to heaps of posts, articles and user comments.

The hard part is to determine where the highest chance of finding active bikers is. In our case, a social listening tool would normally show you mentions and let you reach out to authors via DM. But not everyone discussing bikes is your future customer. This means you will need some additional desk research and makes the customer finding job more time-consuming.

A solution to this challenge is to use image search. DeepBrand uses image search which is directly run on objects detected in photos from social media. So, if you type the phrase like bike, you will get a listed view of all the photos with bikes. You can also see the results on a profile level, to get the most popular profiles of people who posted photos with bikes. This opens an opportunity to better understand your target audience and start conversations with natural opinion leaders.


We are living in a visual world where a major part of online communication happens through images and videos. While keyword monitoring is a great way to discover brand or product-related discussions, it has several shortcomings. Those limit marketers’ ability to gather meaningful data from image content. This is where visual content analytics may come in very handy. Monitoring brand or product mentions can successfully complement your marketing research and strategy planning activities.

Interested in taking DeepBrand for a test run? Sign up here.

Leave a Reply

Your email address will not be published. Required fields are marked *