Picture this: a UAE hospitality giant skyrocketed their Return on Ad Spend (ROAS) by 30 per cent on Meta, simply by flaunting their brand elements in the first 3 seconds of their ads.
Meanwhile, a major fast-food chain slashed their costs per lead by 34 per cent on the same platform, just by ramping up the food products in their display banners. These success stories aren’t about dictating what to do with creatives but rather guiding on how to optimise assets, pushing creative quality scores from a decent 60 per cent to a stellar 90 per cent.
AI in creative analytics isn’t just about making assets; it’s about making assets that drive the bottom line with game-changing tweaks.
The creative quality of an ad, its actual impact and therefore value, has long been a challenge for advertising professionals and their clients. Despite being a critical factor in the performance and success of a campaign, measuring and improving creative quality has traditionally relied on subjective judgments and gut feeling rather than hard and objective metrics.
So how is it that some advertisers increase the CTR of their ads by up to 40 per cent by adding a specific word in their copy?
Well, they had a little help not from a shrewd copywriter but a trained AI. In recent years, AI has emerged as a powerful tool for measuring and improving creative quality.
AI-powered creative quality measurement tools can help brands to understand their target audience and their preferences, identify the creative elements that are most likely to resonate with this audience and test different designs to optimise campaigns before serving them.
Isolating the elements of ad creatives and their impact on conversion, AI reveals and highlights ways to improve the performance of an advertisement. For a telco, the magic word was speed, for a hospitality group, it was staycation, but AI discerns all manners of improvements, like including or removing individuals, changing colours, or the position of an element.
These systems work by analysing large amounts of historical campaign data from digital platforms, which is then used to train AI models so that they understand the relationship between creative elements and performance. These AI models can be used to score creative assets and predict their performance.
This information can then be used by brands to select the most promising combination of elements within creative assets for their campaigns, like word themes and specific human profiles, and optimise their creative strategies.
By quantifying the impact of creative elements on the engagement of our audience and identifying the main drivers of performance, AI increases efficiency and reduces costs.
To be really helpful, a system also needs to be localised to the region and process Arabic language and Arabic ad copies contextually with an accuracy that exceeds 90 per cent. We helped a large Arabic telecommunication brand decrease its cost per impression by 80 per cent by using the word speed in Arabic in ad copies for a stronger product offering on 5G.
Far from being a threat to creative folks, AI actually facilitates experimentation and testing, supercharging innovation in creative design. AI has been used to empower an offline A/B tester that predicts the performance of ads before they are launched on platforms.
This enables advertisers to fine-tune their content and explore unconventional approaches, fostering creativity and differentiation in a crowded advertising landscape.
Another exciting dimension is creative quality measurement. As AI tools become more sophisticated and more data becomes available, brands and media companies will be able to use AI to even greater effect to improve their creative quality and campaign performance.
For example, AI will be used to develop more personalised creative measurement solutions that take into account individual users’ preferences and behaviours. Building on past users’ behaviour, AI will predict the optimal messaging to deliver to each audience segment in each market. The technology will be used to integrate creative development and measurement more closely, enabling brands to create and test creative assets in real time.
Eventually, AI will measure the effectiveness of creative assets across different channels, such as TV, digital and social media. We will finally have a real chance to get it right the first time, optimising the three seconds that brands have to capture consumers’ attention.
By Dr Hoda Daou, general manager of Annalect MENA
