Contact us

A world-first artificial intelligence experiment

How AI can be integrated into a social campaign

news and insights
#Artificial intelligence

A world-first artificial intelligence experiment by Artefact UK, proves how AI can be integrated into a social campaign to improve speed to insight and allow marketers to create real-time personalised content.

At Artefact UK, we are data driven, creative led, and are always keen to come up with innovative ways to improve our services. At the start of this year, we set an experiment to show how Artificial Intelligence can be integrated into a Social Campaign – to collect data that inspires innovation and creates innovative campaigns.

In recent years, the ever-increasing pool of consumer data has created a massive opportunity to better understand customers and deliver highly contextual messages or offers. But the sheer volume of data often erodes the customer experience; data is only valuable and actionable if it’s matched back to a customer with a high degree of accuracy and analysed across all channels.

Brands therefore find themselves unable to harness the insights they need for effective engagement.


At Artefact UK, we know that when raw data is quickly converted into useful insights, businesses see the most dramatic results. We wanted to show, therefore, how Artificial Intelligence can be integrated into a Social Campaign to allow insights to be reached more quickly and inspire innovation.


To prove our concept, employees were given the chance to be guinea pigs in a world-first Artificial Intelligence experiment.  AI hasn’t been used on social media before, so we created a platform that could analyse posts on Instagram.

To get involved, you simply posted a picture on Instagram of your breakfast with the hashtag #4Psbreakfast. (Please note that this experiment took place before the transition to Artefact UK and so the 4Ps hashtag was used) The AI software could then recognise not only what was in the picture posted but also further information such as demographic information of people posting, location, time of post, etc.


The AI produced some very accurate results and our experiment clearly illustrated how AI can be integrated into a Social Campaign, but what else could we do with this concept?

Often marketers create buyer personas based on who they think their audience should be and what that audience are looking for, but they have no data to back it up. The experiment shows how marketers can turn data into real insight. By using data to create personas that match the brands existing customers or ‘lookalike’ audiences, marketers can help identify trends and behaviours of current and prospective customers.

In our experiment we found that other datasets could be integrated to collect different results. For example, we could look into whether differing weather conditions affects what people eat for breakfast – do posts of porridge increase when it’s cold and fresh fruit when it’s hot? Do people have hot drinks with their breakfast whatever the weather? Is toast eaten all year round?

We could also look at when posts are made to see if the time of the day affects what people eat. The AI often picked up superfoods; we could see to which demographic they appear, as well as seeing whether they are consumed more on a particular day of the week (do people try to start the week healthily, for example).

For a food and drink brand, you can see how valuable collecting data in this way can be. It allows marketers to personalise their marketing efforts at a large scale and in real-time. For example, if in England we had a particularly cold blast and it was snowing up and down the country, we may notice people posting photos of hot breakfasts like porridge. Marketers can then use this information to create targeted ad campaigns.

In conclusion, AI can be used to look at what people are posting on social media, collect data and analyse trends to target consumers and create real-time personalised content. The insights drawn from an AI Social Campaign will allow marketers to design tailor-made content for each group, on every channel; real-time advertising means you can dynamically engage your customer when they need the product you are selling.


To discuss how machine learning technologies are evolving alongside new developments in user interaction and cross-platform behaviours, give us a call on +44 (0)207 607 5650 for a no-obligation coffee and chat about marketing possibilities across all inbound channels. How could our unique brand of data insights thinking benefit your business?