post
April 2, 2019
Read time: 6min
Artificial and human intelligence for insights management: Ipsos and Market Logic
The world of insights is changing, fast
In the past, the role of an insights manager was to find precision in methodologies. Now, the task is to find expert ways of synthesizing and applying the vast ocean of data at their fingertips.
Insights managers have to be willing to take risks if they want to be ahead of the curve, says Radhecka Roy, Curation Expert at Ipsos. “Infobesity” – the inundation of data that digitally empowered consumers create – means that insights managers have to be able to hone in on what matters and turn that data into action. Read on to find out how Market Logic and Ipsos have been working together to harness the power of Artificial Intelligence to tackle infobesity.
The power of AI
AI can help take large volumes of unstructured and structured data and bring it all together with a purpose-built machine for market researchers and insights generation.
Take Coca-Cola, for example. At IIEX EU in February 2019, they presented a snapshot of how their Market Logic platform helps them make sense of the mountain of consumer data they have. On their platform, they have over 95,000 research documents accumulated over the past few years.
In December, they integrated social listening data into the platform – at a whopping 945,000 post count. Needless to say, human brains are no longer sufficient to process that amount of data. “AI is essential to help us generate insights from huge volumes of consumer data,” stated Begoña Fafian, K&I Director, WEBU. But how can insight managers be sure that AI is really working for them?
Trust, but verify
AI has to be at least as likely to get the right answer as a member of the insights team. In the case of uploading, auto-tagging and summarizing documents, AI is so advanced that its accuracy is on par or better than a human’s. Until the day when users trust AI completely, they will still have full control over the data and can access it at any time to verify the conclusions or draw their own.
At Coca-Cola, insights professionals needed to know that the insights generated with AI would be at least as reliable as the insights a human could generate. To test the quality of the insights, they wanted the machine to sort them into three buckets:
- Insights that confirm what we already know
- Insights that surprise us, but make sense
- Insights that surprise us, but make us feel uncomfortable
Coca-Cola’s insights team checked the results to make sure they made sense, and more importantly, to evaluate if they were useful. The results for each bucket exceeded their expectations.
Bucket one results found, for example, that “millennials like living near coffee shops.” Bucket two contained results like, “coconut flavors are often associated with breakfast,” and bucket three provided insights like “stevia is associated with an unpleasant taste.”
Where human intelligence come in
AI capabilities are impressive, but machines can’t solve complex business challenges alone. Human Intelligence (HI) is required to bring the curiosity, creativity, culture and wisdom to insights management. It’s what Jack Ma calls LQ – the “Love Quotient.” Working hand in hand, Human and Artificial Intelligence can produce much richer output than working alone.
A June 2018 McKinsey study identified three types of organizations: the idlers, isolators, and integrators. Idlers are enterprises that are still struggling to cope with huge volumes of data. Isolators are companies that have strong data capabilities but low levels of creativity.
Integrators are the champions of marrying creativity and data and, according to McKinsey, data-driven creativity is as significant as doubling a company’s growth rate. And this is the space that Human Intelligence and Curation plays in.
From infobesity to impact
This combination of AI power enhanced by Human Curation was piloted through a study that Intel and Ipsos partnered. Intel, the multinational technology corporation, realized that their customer base was fragmented, so they wanted to capture real-time micro-journeys.
They worked with Ipsos to develop a platform that would help keep their finger on the pulse of these journeys as they happened, provide them insights in a language their business would understand, and present a menu of opportunities to drive impact and growth. This successful case was presented by Intel and Ipsos at ESOMAR Fusion in November 2018.
First, Ipsos started with social listening and AI, which took two hours to come up with over 75,000 responses. Then they used the machine to clean up the responses and develop them into 30,000 logical thematic clusters, which took two days.
Then human curation came into play: within two weeks, Ipsos Curators built eight archetypal journeys and brought them alive as compelling stories with emotions, motivations and latent needs, as well as opportunities for marketers to act as a starting point for activation.
Self-service insights were presented to marketers on the Ipsos Insight Cloud, run by Market Logic. The Insight Cloud is seamless and easy to navigate – it is a one-stop-shop to action insights and multimedia assets, and links to larger databases.
AI liberates the curator
At Intel, insights managers have saved time while producing better, more inspiring outputs that are ready for marketers to apply. Currently, Intel is in the process of integrating the purchase journeys with other fundamental frameworks used – with a view to eventually link it to CRM.
Some of the applications underway for activation include sharper media engagement, selection of influencers for influencer marketing, content co-creation with users/influencers, initiatives for key touchpoints and a review of product range description on marketer websites. As Sunita Venkataraman, Director of Global Insights and Analytics at Intel states, “Data is not magic. You need to know how to connect and bring it to life … we need an agile system that helps us make real-time decisions. And AI with Human Curation makes a great combination to deliver that.”
The agile edge
The best companies know that agility does not only refer to speed. Agility means progressively building upon hypotheses to find what works. It means intelligently calling upon what you already know to make more informed decisions, and getting them to market quickly.
Agile companies search before they research and invest in expert Curators to build meaning into their analysis, and it shows – their post-curation insights are ready to activate at a moment’s notice.