How ‘AI as a service’ helps Social Soup scale

Cloud-based image analysis helps deliver insights for influencer marketing agency

Australian influencer marketing agency Social Soup has leveraged cloud-based AI services to help it analyse thousands of images to determine what elements are driving the greatest engagement for brands.

Social Soup has been in the market for over a decade and during that time has evolved significantly as social platforms and technology have changed, says founder and CEO Sharyn Smith.

When the agency launched it was very much about people catching up and sharing in the “real world,” Smith says.

“Now it’s much more around social connections and how people connect on different platforms and how they share and what inspires them and who they follow.”

The agency has worked with a number of well-known brands across a variety of industries. Clients have included LG, Nutella, Stan, L’Oréal, Bonds and Dyson.

“Ultimately at the end of the day, for us it’s finding who are the people that are influencing decision making in terms of someone buying or trying, or changing behaviour about a particular product, and partnering with them to try the products, create content and share that with their followers,” Smith says.

The agency works with influencers ranging from high-reach “Insta-famous” celebrities through to everyday people — “a mum that might be sharing it within her mum’s group”.

Since Social Soup’s founding, the scale of its business has grown significantly — with cloud services helping to underpin that growth.

“People in the past worked with a handful of influencers,” Smith says. “Now it’s very much part of their marketing and business strategy that drives their content, that feeds into their social and digital ecosystem and they’re working with hundreds, if not thousands, of influencers at a time.”

Social Soup uses a bespoke, in-house-developed system to “tap into conversations that are happening out there,” Smith says, and find the right influencers for a brand to work with. The company is able to “analyse hundreds of thousands of social media profiles very quickly for particular keywords and hashtags”.

“We wouldn’t want to have that server capacity all the time, so certainly from that side of things — being able to go up and down in terms of capacity — working on the [Amazon Web Services] platform is incredibly important for us,” Smith says.

The other key advantage from being a cloud-based business is that Social Soup can quickly integrate new services and tools and they become available, the CEO adds.

One example is Rekognition. Amazon Web Services in March announced the availability of its Rekognition service in its Sydney Region — making it the first of the company’s AI services to be available from its Australian data centres.

Amazon Rekognition provides image and video analysis. AWS says the service can be used to identify objects, people, text, scenes, and activities depicted in a still image or video as well as conduct facial analysis and facial recognition.

Smith says that Social Soup was already making extensive use of the service before its local launch — switching it on around 24 hours after AWS first announced it in late 2016.

Smith says the company was “pretty much the first in Australia” to start using the service, thanks in part to the agency’s close relationship with AWS.

“Straight away we were able to analyse vast amounts of data and look at images and analyse what’s in them, then look at what images and what influencers are driving the greatest engagement – and report back to our clients on the insights that sit behind that,” Smith says.

A campaign for an instant coffee brand’s organic range found that the best performing images featured food alongside the coffee, for example.

Social Soup’s cloud-based system is not only able to achieve scale that wouldn’t be possible with a manual approach but can also find patterns across images that might escape a human, Smith says.

“At the moment Rekognition can look for up to 2000 attributes,” the CEO says. It was used to establish, for example, that the best performing images for a major soft drink brand were those that had “high proportion of natural environment and greenery.”

“It was able to pick up on plants, and trees and that sort of side of the images that were coming through for that particular campaign,” Smith says.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

Tags cloud computingAmazon Web Services (AWS)artificial intelligence (AI)Social Soup

More about AmazonAmazon Web ServicesAustraliaAWSDysonLG

Show Comments