Artefact’s data scientists, data engineers and consultants worked with Reckitt’s transformation team to implement a powerful and scalable solution that will enable Reckitt to make intelligent use of data to drive sales effectiveness from its digital marketing activities. Called the Audience Engine, the platform is now used every day by Reckitt’s teams around the world.
CHALLENGE
Transform into a digital-first business and use data effectively to increase sales
Reckitt is a multinational consumer packaged goods (CPG) company, with offices all over the world. Its purpose is to create a cleaner and healthier world and its products include household favourites such as Dettol, Durex, Harpic, Lysol, Nurofen and Vanish.
Like many businesses today, Reckitt recognised that it relied on traditional methods to understand its customers – consumer panels that delivered audience insight based on surveys, brand knowledge, demographics, consumption and market statistics; shoppers were predominantly reached through TV advertising.
Reckitt’s mission was to transform itself into a data-first business, a move that would enable it to exploit the value of its own data to strengthen its understanding of its consumers, and from there make its digital advertising more effective in order to increase sales.
SOLUTION
The Artefact Audience Engine: machine learning and AI models turn data into actionable insight
The Artefact Audience Engine takes a scalable and artificial intelligence (AI)-driven approach to first-party data.
Looking at the user journey (on its websites) provides Reckitt with insight about the user’s intentions, and their propensity to purchase. Artefact’s custom built machine learning models use this data to create hyper-targeted audiences for digital marketing campaigns. These audiences are built and sent directly to a demand side platform (DSP) of choice through API connections and are used as seed segments for DSP lookalike models to find scale. The Audience Engine allows the brands to answer their specific business questions, better understand their customers, and target them according to their needs.
One of the very few Google partners to be certified on both the Google Cloud Platform (GCP) and Google Marketing Platform (GMP), Artefact used both platforms to build the Reckitt Audience Engine.
The Audience Engine helps Reckitt’s brand marketers deliver campaigns that are more effective at reaching core consumers. But the benefits also extend across the business: the media team is able to create more efficient media campaigns; communication strategists can better understand audience performance and use this to prioritise future business objectives and activities; while the CRM team is able to build stronger programmes and content.
DEPLOYMENT
Multi-disciplinary teams roll out and scale the Audience Engine across Reckitt’s global business
Artefact worked with Reckitt’s transformation office to deploy the Audience Engine at a global level. Data scientists developed the product in terms of its capabilities and algorithms; consultants rolled it out to all the markets and brands; and data engineers worked to industrialise the tool, building a platform so that anyone from Reckitt can run the Audience Engine for any brands automatically. This took place in four streams, with a robust project management methodology implemented by Artefact to deliver a programme of this scale:
The rollout team worked directly with the martech leads, the media managers and the brands to deploy the tool.
The product innovation team, made up of data scientists and consultants, listened to the needs of the business and developed new capabilities for the Audience Engine that enabled these to be resolved.
The product industrialization team worked to scale the product, developing a tool so that the Audience Engine runs automatically.
The adoption team demonstrated the value of the Audience Engine to the broader business, consolidating results and talking to different stakeholders about how it could help them in their own day-to-day tasks.
RESULTS
Test and learn campaigns show an inverage of 21% increase in media spend ROI
Reckitt has big ambitions for its Audience Engine, wanting to ensure it will be used across all key brands and markets. To achieve this, it adopted a methodical approach, first creating digital centres of excellence in specific regions to cover key markets, and then using agile teams to drive development and innovation.
In Mexico, sexual health brand Sico believed its core consumers to be young males however the Audience Engine debunk this long standing idea and showed that young females were equally as important and that the message to consumers shouldn’t only align to male interests. The Audience Engine also unearthed an even more exciting piece of new insight – it found that those who concentrated their time on articles related to safe sex and STI prevention on the website, the core audience was dominantly young females. This led to the realisation that in Mexico, there was an appetite for young women to understand how Sico would help them with safe sex practices and there needed to be a purposeful effort to reach young women differently.
The second example is in the US, where media planning and buying is based on the Oracle Data Platform’s pre-determined purchase data segments, which look at the specific products or categories of products that people buy. But trials showed that modelling raw data using the Audience Engine machine learning algorithms could increase ROI on media spend by 30%. By taking control of what model to use on a given dataset to drive a specific objective, the Audience Engine has demonstrated to be significant more effectiveness in impacting ROI while also driving greater media buying efficiency. By lowering the cost of media by 20% in this test & learn, we were able to reach more relevant people that were more likely to impact our sales – the Audience Engine makes media budget work smarter and harder.
The Audience Engine’s purpose is to enable Reckitt to use data in a highly sophisticated manner to be more effective than conventional digital targeting. The test and learn campaigns set out to prove the value of Audience Engine has solidified the product’s credibility and inspires confidence to drive adoption to Reckitt’s benefit. Test and learn campaigns such as these inspired confidence by showing the Audience Engine to be highly credible; having quantified the benefits, Reckitt could then start to drive adoption so that it is used in as many campaigns as possible.
FUTURE PLANS
Using the Audience Engine to continue to evolve as a digital and data led company
Reckitt’s objective is that the Audience Engine continues to deliver incremental value, making its business more effective and efficient; this will see it support both large and small brands across all its markets by enabling them to gain an in-depth understanding of who their customers are. As technology develops, it will look for new ways to leverage data, adapting the Audience Engine to fit as the company becomes more digital and data led.
