A picture is worth not just a thousand words, but a thousand tweets, or texts, or anything not visual. Life is fast-paced and social images are part of – and a result of – that pace. So, are you analyzing images? You should be. Otherwise, who knows how much of the conversation you’re missing?
It was predicted that 1.4 trillion photos would be taken in 2020. And 90% of that number were taken on mobile phones. How many of those do you think made it onto social media? Whatever the number, it will undoubtedly rise in 2021, every photo not included in your social media analytics is an opportunity missed.
Our “What Is?” series continues with the importance of image analytics, and options for brands to bring this technology into their data operations.
Why Image Analytics, and Why Now?
We have stepped into an increasingly visual era with everyone having a cellphone in hand and a social platform such as Instagram and Facebook to share them on. And social platforms are growing daily. In 2020 there were over 3.6 billion people using social worldwide, a number projected to increase to almost 4.41 billion in 2025. So, if you aren’t including visual data into your social media monitoring, everything you think you know about your audience is off.
Still not convinced? Over 50% of shared content on Facebook is visual, with 15.09% being video and 34.73% being photos. And on Instagram, 59% of posts are pictures. Now imagine the number of those visual images that include brand products and logos – without any mention of them in the text. In many of these images, a popular brand may be represented, even if it isn’t the point of the post. That information is valuable – especially for understanding the lifestyles and passions of your target audience.
Ignoring image analytics doesn’t just affect your audience analytics. It affects competitive intelligence, product development, influencer strategy, and more. And these images help organic reach or locating potential influential partnerships you can use to extend that reach even further.
Looking at the hashtags, this micro influencer Eliza Melendez, is obsessed with fitness, nature, and photography. And her close to 5K followers stay tuned in to her posts. So, you know what she’s about, but none of the information mentions Nike – even though the image shows her relaxing in Nike pants.
This could be a potential micro influencer for Nike to engage another demographic of fans. They may want to know that Nike is her brand of choice. And this reinforces that you must include images in your social analytics. And with NetBase, you can.
The Foundation of Image Analysis
Image analysis is used every day by most people without them even realizing it. Whatever you use to backup photos on your phone – Google Photos, for example – is already categorizing your images, whether you asked it to or not.
Smartphones are constantly updating and getting better at reading information, it can differentiate, as an example, which pictures are of the zoo you visited with your nephew versus your boss’ retirement party.
And image analytics takes that technology up a notch, categorizing images from social media, and sorting them according to everything applied to text: gender, age, facial expressions, objects, actions, scenes, topics, sentiment and brand logos.
In just a few short years social media analytics has progressed in leaps and bounds and we’re just at the beginning when you imagine where image analytics will be as technologies like AI, AR, VR, etc. become the standard.
NetBase already uses machine learning and artificial intelligence (AI) technology to power our image analytics – and our commitment to grow, innovating to include anything new on the horizon, is unwavering. That’s the least you can ask in a social media monitoring platform – that it commits to grow as technology does, versus relying on third-party solutions to keep up.
So, Do We Still Need to Analyze Text?
Of course! Images are posted without accompanying text all the time, and there are still instances where text is the main focus. Consider blogs, news articles, and reviews – text is still the bulk of the equation. But when you combine text and image analysis, your data becomes that much more powerful.
Think of image analysis as another layer to your social analytics tool, just as ability to decode emojis and natural languages does for text. And sentiment isn’t just for adding context to text, it can also be applied to images. And this is a huge advantage in several ways. For instance:
- Images transcend language barriers. You may not speak Mandarin, but images are universal. That’s a boon for global brands, or those looking to expand geographically.
- Images offer quicker accuracy. There’s a lot less to decipher or question with images. It’s all right there. Where social text is often limited character-wise, images are limitless.
- Images are personal! And social data and sentiment is all about getting personal with consumers. Nothing brings you closer to your audience than the glimpse into their lives provided by the images they share.
Don’t Brands Have Enough on their Plates?
Every day there seems to be some new technology brands are told to urgently embrace. While some brands are barely sold on the notion of social media analytics – savvy companies understand the importance of social but many still aren’t using it in all the ways they could. Or they have trouble convincing stakeholders of its value.
With social data, the proof is in the pudding. And you’ve got to have a handle on all the ways image analytics further your brand if you’re going to convince anyone to invest in a suite of tools. They need to know what they’re getting for the money and with that in mind, here are some points you can share:
1. Image Analysis Offers Greater Accuracy
We already know accuracy is paramount to the usefulness of social data. Without it, there’s no point in bothering. Social media analytics is more than number of brand mentions, it’s also about quality and context of mentions, including visual mentions.
Image recognition allows image analysis tools to identify your logo in social posts, letting you know when your brand is “mentioned” visually. But it’s not just about your logo. Visual image “mentions” analyze objects, scenes, locations, etc., meaning you can look at images related to your category. For example, any image that features a cup of coffee – or the specific logo of your brand or a competitor.
And with the many ways you can view these, you’re sure to find details to inform you on your brand health. Some of the views NetBase offers via our social media monitoring are:
- The Objects view which displays objects identified in post images, such as “person” or “sandwich/burger.”
- The Scenes view identifies scenes in post images, such as “football field” or “arena.”
- The Emotions view pinpoints emotions identified in post images, such as happiness, sadness, and neutral.
- The Logos view displays secondary logos identified in post images. For example, a logo-based topic for Starbucks might contain the Starbucks logo along with an Oreo Logo.
As an example of our objects view, below we’ve analyzed Starbucks. As we can see it pulls out some great visuals in connection to person and cup. And these searches can be filtered and adjusted to high, medium, or low to get more precise findings.
NetBase has the highest precision image analytics logo recognition that can distinguish partial, hidden, rotated, and even modified logos for greater brand awareness. On top of this it also
offers the largest logo library, which includes 2000+ logos of global brands across CPG, Luxury, Retail, Food and Beverage, Travel and Hospitality, Media and Entertainment, Sports, Financial Services and Technology.
2. Makes Sentiment Analysis More Complete
Text posts rely on NLP (Natural Language Processing) to convey context beyond the words. Now you can do that for images as well.
With posts identifying your brand visually, text – even a hashtag – adds sentiment context. And so does the expression on a consumer’s face, for example.
That detailed analysis includes posts that might tag your brand without providing any other text-based clues. If the image is doing the talking, you need the right social analytics tools to “hear” what it says.
If you were a gamer, wouldn’t you be excited about a Ewin Racing gaming chair? But the brand name is never mentioned. And this is a great example of UGC – User Generated content – one of the biggest indicators of brand passion out there. And social media analytics readily spots and analyzes it.
3. Helps Identify Better Sponsorship Partners
Brand sponsorship opportunities are a great way to increase awareness and earn a little consumer love by association – but only if you partner with the right organizations and events.
How do you know whether it’s worth putting your ad dollars on the line? This is another way logo recognition comes into play. NetBase allows you to view engagement surrounding brand sponsorships. For example, viewing a competing brand’s logo for a particular event helps you to know if it’s worth a similar partnership for your own brand.
For great results, decide on a timeframe and then make a list of #hashtags, content/messaging, influencers, and even emojis to search. Here we’ve chosen Paris Fashion Week and their partnership with L’Oreal as our topic to see how this beauty line is being represented.
And our social analytics tool reveals posts such as this one, which helps L’Oreal measure real-time campaigns to see how much engagement and exposure their brand is getting, even while not the main focus of the event.
Loreal has to be happy about the mention in this post, with their brand splashed across this Tweet. The engagement could be higher, but keep in mind this even isn’t until Fall.
Tracking success of upcoming events is critical as it allows you to adjust as needed to ensure a great outcome.
But what about tracking the success of a campaign that’s just passed? 2020 was a year of tumultuous navigating for brands of all types. And having social media monitoring made the going much easier for brands to find their way.
Take Black Friday for example, while many brands were slicing prices to win favor, Cards Against Humanity didn’t skip a beat when given the chance to shine. They chose instead donate their entire Black Friday budget to nonprofits instead of sales. And the engagement on this is high- something we’re sure they are glad to see, and other brands can take a page from. In giving, they won consumers’ favor – and a good bit of loyalty.
How’d they know what to do? That’s something that NetBase combined with Quid products can help us with. Maybe Cards Against Humanity dug deeper into their demographics – to see what other interests unified their audience. Using Quid Social, we’re given a timeline of conversation that illuminates consumer discussion around card games, such as Cards Against Humanity. This helps a brand understand which areas are being talked about the most and how they’re being talked about.
Highlighted in our timeline above are individual posts by consumers, each with valuable information categorized by color and segment. It’s a great way to boost your social analytics game – having both market research and social media analytics provides you the whole picture. And perhaps it’s within these colored bars that Cards Against humanity correctly spotted the trend of giving back in 2020.
4. Assesses the Value of Potential Influencers
Influencers are a ticket for brands to reach more consumers and image analysis can be a part of your influencer marketing strategy. Both celebrities and non-celeb influencers sharing images of their lives reveals opportunities beyond the obvious.
Consider this Starbucks post we located using media analysis:
The intent of the post is to share how thoughtful this social media user found Starbucks baristas to be. This may be a way for the brand to gain some more exposure through his devoted followers. With the number of retweets and likes this post experienced, it’s reach goes beyond just this user’s follower count.
5. Let Brands Understand How, Where and When
Consumers post about everything from a product purchase to finding a great price, or the result of a customer service situation. However, they may not take the time to write about the way they use products. Instead, they show, not tell, through lifestyle posts on social, as many of the posts above illustrate.
Not only do these pictures – hopefully – indicate brand love just to be visually mentioned, but you can understand how consumers use your products, and why. And that is a key and critical detail which tells you a lot more than a simple brand mention, especially if the visual mention is negative.
Though this post has negative sentiment, and it’s low on the engagement meter, it is still worth thinking about. The negative sentiment doesn’t seem to be about KFC as a brand but that this fan can’t afford the KFC gaming console. These posts are still worth investigating as they speak to what brand consumers’ think. Do other KFC or gaming enthusiasts feel the same? It’s worth finding out and potentially creating a contest to gift a console to a deserving fan!
Social media analytics is a complex puzzle, and every piece matters. Today, images are a massive part of the consumers world and their impact must be counted as a daily part of your social media monitoring.
Image analytics was relatively new a few years ago, now it’s becoming just as valued as sentiment analysis, so brands will want to add it to their arsenal. You need every edge you can get to keep ahead of competitors and understanding the visual side of the coin might be just the ticket to stardom.
This has been part 6 of our Social Media Analytics Guide for 2021, designed to keep you in-the-know on the tools, metrics, and skills necessary to compete in an increasingly global arena.
Read the other parts of this comprehensive series by clicking below:
- What is Social Media Analytics and Why Is it Important?
- What Is Social Media Monitoring?
- What Is Social Campaign Analysis?
- What Is Social Sentiment Analysis?
- What Is Social Listening & Why Is It Important?
- What Is Image Analytics?
- What Is Audience Analysis?
Ready to see how our image analytics work as part of our comprehensive social analytics platform? Get in touch!