AI makes your content work smarter, not harder
Generating quality content, ensuring it reaches the right audience and stays top of mind is a constant struggle for marketers and ecom managers alike. How can we make sure that the best, most captivating visuals are above the fold and helping to boost sales?
Enter the machines. Artificial intelligence and machine learning have been trending since the early 2000s. And as technology has evolved, its potential has branched out from the realm of sci-fi to more practical real-world applications. Now is the time to ask ‘what can AI do for me?’.
At Flowbox, we think the answer is simple. Help pinpoint the best content, and keep it in front of your customers.
Content that converts
Let’s break this down. First of all, what do we mean by ‘best content’? You can define best in many different ways.
As a visual marketing platform with collection, curation and analytics tools built-in, we can directly measure performance and conversion of individual photos and videos onsite. Meaning that for us, ‘best’, means content that converts, adds value and boosts sales. But while we can measure a post’s performance after the fact, how can we select the best content from the get-go?
Finding quality content
Another way to determine ‘best’ is to look at the quality of the content. Which, of course, can be very subjective. However, there are some universal truths that can help, especially when collecting user generated content for use in marketing. You don’t want images that are too dark, where you can’t make out a person or a product, or images with a lot of blank space. And sorting through content to pick out gems can be time-consuming and take attention away from other tasks. This is where machine learning can be your friend.
Ahead of launching this feature, Flowbox fed the AI algorithm hundreds of ‘good’ vs ‘bad’ images and video thumbnails to learn from. These good/bad examples were based on content typically approved or rejected by real Flowbox customers from around the world. Using this tactic means that the AI has been exposed to and learned to recognize the type of content that end users deem high quality, which performs well and engages website visitors.
The AI has learned to recognize patterns, such as rejecting images that are too dark or contain little to no information. Each piece of content is now designated a ‘Quality Rate’ – a number between 0-100, meaning that customers can quickly filter out poor quality content and focus their attention on the higher scoring images. This is a great example of AI and machine learning reducing workload and making a process more efficient.
Relevance is key
Another way to define ‘best’ is ‘most relevant’ – after all a customer looking for a professional handbag is unlikely to be swayed by even the most attractive suitcase and traveling content. It simply isn’t relevant. This is where categorization and personalization come into play. And in the backend or digital asset management system, content needs to be organized to ensure our handbag shopper is inspired by the latest fashion trends, and not content from hiking trails.
AI-powered smart tagging at Flowbox helps identify and display relevant content. Each image is assessed by our AI tool, which identifies concepts and objects in the image and automatically assigns a set of smart tags (from a list of over 40,000 different keywords). This means that you can easily search and find relevant content and distribute it to your site/email marketing channels. With a visual marketing platform like Flowbox, you can easily embed dynamic visual galleries filled with authentic content onsite that match one or many keywords with each category page.
We really like how premium bicycle brand SQ Labs is displaying relevant, inspirational content from their community on their site. And if you compare their mountain biking category page versus the road bike category page, you can see how different the content is, ensuring it’s relevant. After all, a customer looking for a mountain bike has very different needs and desires than a road racer!
An omnichannel experience
Finally, we’ve broken down what ‘best’ content is and how AI and machine learning can help with the sorting and scoring of images in the backend to optimize customer experience. How about on the frontend: how can AI help ensure that the best content is present in your omnichannel marketing?
At Flowbox, we like to let machine learning do the heavy lifting. Our Flowscore algorithm uses statistical analysis based on a post’s conversion and engagement and sorts content displayed onsite based on performance. This means that rather than just displaying visuals in chronological order, the content with the highest clicks and conversions moves to the top. This sorting algorithm can also be applied for content exported to email marketing, social advertising, in-store displays and more. Because Flowscore analyses the content in real-time, it’s possible to export and distribute the highest scoring content directly into other feeds and channels. Your website audience then becomes your original testers and then, as a result of their interactions with the Flows, the most attractive content gets redistributed elsewhere.
Letting AI do the heavy lifting
Machine learning and AI are great tools in today’s digital-first marketing and customer experience. By analyzing images, sorting and scoring content and continually learning and developing, AI cuts manual work and streamlines your marketing efforts. By measuring data in real-time, AI solutions allow you to react quickly and ensure that the best performing, most relevant and highest-scoring content is top of page and top of mind.
Soon the question will be: ‘What can’t AI do for me?’.
Interested in getting started with Flowbox? Click here to book a demo.
More information:
Cecilia Rehn, Digital Marketing Manager, Flowbox
cecilia.rehn@getflowbox.com
