How Images Are the Next Big Data Source for Analytics and Business Insights

Images are the new currency. Don’t believe me? Take a look at these statistics

• 95 million images are uploaded on Instagram daily – Source
• There are over 330 million active Twitter users and tweets with images receive 150% more retweets – Source
• 60 million emojis are used on Facebook daily – Source

The proverb ‘a picture paints a thousand words’ has become even more relevant in today’s sharing age. But why? We could say that it is easier to express a feeling with an image than text. Or that the human brain processes visual content faster.

We are now living in the world of hashtags, emojis, limited text characters. Images complete the stories these are trying to tell. Images are not only enhancing text but are often standing in place of it entirely.

And wouldn’t it be a shame if organizations today did not leverage the information from this huge ocean of visual data?

It is estimated that by 2021, the global image recognition market is expected to touch $38.92 Billion. The video analytics market is expected to touch USD 8.55 Billion by 2023. So where does image analytics feature here?

Much like how sentiment analysis raised the bar for social monitoring, image analytics is raising the bar for social listening. There is a goldmine of data stored in these images as these help organizations understand visual sentiment especially when text is absent.

Take this tweet as an example.

The entire sentiment of the tweet is summed up in the hashtag #PerfectDay. Nowhere is the airline brand mentioned but only the word ‘plane’.

Now imagine the kind of opportunity Emirates could have created by leveraging logo recognition resulting from image analysis! A tweet like “Thank you for the click! You’re such an amazing mom. We hope you and your son had an amazing time plane spotting”.  It is this human-to-human kind of conversations that customers are looking for.

So, what exactly is image analytics?

Image analytics isn’t some futuristic technology. You are probably using it without even knowing it. Your smartphone is already categorizing your photos. Your iPhone will tell you easily which photos are from your office party or from your latest adventure trip.

What image analytics essentially does is categorizes images from different sources and sorts them according to contexts such as facial expression, age, action, topics, sentiment, gender, and brand logo.

How does it do so? Quite simply by leveraging automatic algorithmic extraction and consequent logical analysis of information found in the image employing digital image processing techniques.

And why should you care about image analytics?

Fritz Venter and Andrew Stein say that the objective of image analytics is to “bring an unstructured rendition of reality in the form of images and videos into a machine analyzable representation of a set of variables.”

Here are a few reasons why image analytics is something to look out for

Source authentic data

There is data everywhere. But how authentic is that data? Organizations across the US spend almost $10 billion each year on third-party authentic data. This sourced data has its accuracy limitations and yet forms the basis of many personalized marketing campaigns. The result? Limited accuracy.

With image analytics, the data organizations source will not be mere numbers from a survey, but actual customer data derived from first-hand sources, think facial expressions. The data is also captured real-time when the customer is experiences something, often before she makes a purchase. Such data becomes more relevant when personalizing offers as the insights derived are deeper, more accurate and also real-time.

Improve customer journeys

The age of digital transformation puts the customer in the center of all focus. And for that, it is imperative to improve customer journeys and customer experience. Leveraging facial recognition businesses can create a positive impact on the same in several different ways.

• The airline industry, for example, can leverage image analytics to replace passport checks or deliver travelers from annoyance caused by straggling passengers. Changi airport, for example, is putting facial recognition technology to work to find lost passengers, detect and find people who are on a particular flight or leverage camera-based scanners to automate passport gates. The airport is also using this technology to offer self-service at check-in, immigration, and boarding.
• Insurance companies can use image analytics and facial recognition to improve the insurance claims process. An insurance provider in the US is doing the same by allowing customers to upload pictures of their damaged vehicles. The company analyzes these pictures and processes the claims.
• Insurance companies can also leverage this technology to reduce liabilities from workmen injuries from dangerous risk assessment areas such as rooftops and use drones instead. They can also determine the extent of damage and claims estimation reporting and enable faster assessment of claims especially in the wake of natural disasters.
• The retail sector can unlock the power of image analytics to validate customer identity at stores using cards such as Mastercard. This can help prevent cart abandonment by eliminating challenges such as OTP’s sent via text messages.
• Facial recognition and image analytics can also be leveraged to gauge customer dissatisfaction by analyzing the customers’ facial expressions and movement. It can also be used to activate customer loyalty programs and preferences and improve the customer experience and customer journey.

Optimize retail initiatives

Image analytics has a huge potential in the retail industry. The Consumer-Packaged Goods industry, for example, could leverage some technology advantage to optimize shelf monitoring, store checks and audits. With this technology, retail outlets can

• Effectively track and monitor in-store operations using shelf images.
• Get real-time insights into key performance indicators such as stock outs, on-shelf availability, compliance metrics or pricing changes.
• Improve store coverage for field sales representatives by replacing manual checks with image recognition.

The analytics derived from accurate in-store insights help to optimize store coverage, identifying performance issues and thereby improving retail execution and recovering lost sales.

Improved sentiment analysis

Sentiment analysis is gradually becoming an essential contributor to improving customer journeys. Using image analytics, businesses can gain a context that goes beyond words.

Image analytics makes sentiment analysis more complete by giving you the capability to understand what a picture is ‘saying’. So, while any text or hashtag adds sentiment context, image analytics can come into play when there are no textual clues. What happens only when the image does the talking? You need to be able to hear the sentiment, right? And that is what image analytics helps you achieve.

Image analytics also complements social analytics. Social analytics is a complex puzzle where every piece is important (including video thumbnails). By adding image analytics to your arsenal, you can understand the visual side of the social analytics coin and stay ahead of the curve leveraging the insights.

Image analytics and healthcare

This technology is rapidly becoming a rapidly growing in the health IT sphere. According to reports, the global medical image analytics market is poised to grow to $4.26 billion by 2025 as the industry shift towards value-based care increases.

Accurately classifying complicated texts, images, and other clinical data to improved disease identification capabilities (such as detecting small tumors in gigapixel pathology slides, etc.), to improving patient journeys in the hospital environment, assist clinical decision-making using actionable insights generated from images, image analytics has huge potential in this sector.

As we go deeper into an economy where competitiveness is fueled by technology, image analytics definitely helps businesses unlock insights that would otherwise go unnoticed. Would it then be too audacious to say that while image analytics still is relatively new, it makes sense to add it to your analytics arsenal – because at the end of the day, seeing is believing?