Ipsos Encyclopedia - Social Intelligence and AI Enabled Consumer Intelligence
AI enabled Consumer Intelligence (AICI) is a new category defined by Forrester as a way to craft insights from various data to optimise the experience of current customers and discover emerging trends, outliers, and unexpected shifts or changes in consumer behaviours.
AI enabled Consumer Intelligence (AICI) is a new category defined by Forrester as a way to craft insights from various data from outside their firm (e.g. social conversations, web, and/or consumer data) combined to their own data (e.g. CRM, website) to optimise the experience of current customers and discover emerging trends, outliers, and unexpected shifts or changes in consumer behaviours.
AI-analytics can help to spot unknown phenomena – answers to questions you didn’t yet know to ask. Or the so-called “unknown unknowns.”
AI-CI has broad use cases and can support the business from front-end innovation, product development, brand building, creative development and customer experience providing both strategic and tactical Insights value for Market and Category Understanding, Trends and Cultural Insights, bringing unique understanding on how people express themselves and discuss.
Social Intelligence players are leading the AI-CI new market-place with platforms agencies like Talkwalker, Quid Netbase, Blackswan, Brandwatch. Social Intelligence has been a rising research practice that leverages conversational data from Social conversations and Search, Natural Language Processing amongst others analytics boosters, along to qualitative frameworks to answer specific business questions, solve problems, spot new business opportunities.
Social Intelligence differs from “Social Listening” or “Social Media Monitoring” which is the process of tracking mentions of a brand, product or competitor online, so very much KPIs oriented monitoring, which typically sits as a function within Marketing, Digital often as an extension of social media management.
Whilst Social Listening (1st wave) is typically the automation of the collection of large social media datasets, Social intelligence (2nd wave) is positioned higher on the insight value chain as it provides actionable insights for specific business cases empowered by the combination of machine and human intelligence. AI-enabled consumer intelligence marks the next stage (3rd wave) with the blend of diverse data sets, AI models at scale and predictive analytics.
Ipsos Point of View
Due to its organic and unprobed nature, social and search data can bring value to market research by answering to our clients' business questions via a consumer-centric point of view. The Social Intelligence insights quality comes from:
- the best blend of artificial intelligence, machine and human analysts,
- the right data-set, that very often requires to go beyond social media and includes other profiles of conversational data (like blogs and forums, search).
Social Intelligence can be particularly useful in:
- Hearing people in their own words: Discover how people talk and express themselves (words, definitions, jargon, expressions, organic hashtags) in relation to clients' products, category, issue or policy.
- Catching and tracking social trends: spot early stage signals and trends as they crystallize online
- Seeing what matters, what is engaging: Social data is polarizing, and that is a good thing in some contexts - people talk about what matters most to them.
- Setting the stage and Enriching context: People are posting about the occasions, issue, brand or category when interacting in real life
- Understanding needs and segments: catch the attributes, benefits, barriers that matter.
As with any market research approach, Social Intelligence has some disadvantages and limitations, mainly:
- Generalization to the broader population is constrained, as:
- Social is often anonymous, so we can't control for background variables
- Demographic data is not available or is not reliable in social media
- Not everyone is online, nor on social media
- Not everyone on social media is actively posting
- Not all topics are born equal when it comes to being discussed or searched online. Some issues, brands or categories are just not mentioned enough to make social intelligence possible
With these in mind, executing a good quality Social Intelligence work requires a set of operating procedures and best practices aimed to use Social Intelligence at its best.
Ipsos Views - AI meets Consumer Insights
