Setting a new standard for applying Social data
Every day millions of us share our needs, wants, frustrations and desires on social media networks. This rich and robust data source has the power to transform how we surface and deploy consumer insight. So why are brands still struggling to unlock its full potential?
Transforming unstructured data into scientific predictions
Our always-on data platform ingests millions of real-time, publicly available conversations from a variety of Social data sources.
Irrelevant content and noise are removed, whilst clustering analysis and Natural Language Processing tools structure the data into a consumer-defined category taxonomy.
The data is scanned for emerging and new topics. Each isolated topic is analysed by a variety of factors, including how volume, growth rate, sentiment and associations are trending over time.
Our Trend Prediction Value (TPV) algorithm analyses each trend’s maturity and calculates its future growth potential based on seven years of historic category data.
All category trends are ranked against each other using their TPV score, giving you one global metric by which to prioritise new and emerging trends.
Users can access this data and analysis to track and monitor trends, using our dynamic, always-on, trend prediction software: Trendscope
Trend Prediction Value (TPV) is a ranking metric to help brands identify which trends to prioritise for their innovation and marketing strategy. TPV is driven by data science, assigning a value to each of the thousands of trends within an industry, enabling trends to be ranked objectively based on real consumer conversation patterns and machine learning.
The algorithm includes the size and growth of conversation over a two-year window, and a projected future growth forecast. TPV utilises our behavioural conversation datasets which are cleaned to ensure only relevant conversations about your industry are included, minimising the noise often present in social conversations.
Find out more about our Social Prediction tool, Trendscope.Learn more
Accurate trend prediction in a pandemic
TPV accurately ranks and prioritises individual trend manifestations, such as ingredients, benefits, and product types as well as trend clusters and macro category drivers – the big ideas that will shape the category for the next three to five years.
COVID-19, though, represented a new challenge. Consumer conversations soared on previously low-volume topics like banana bread, canned goods, and homemade beauty products. While it was clear some trends would stay and have a lasting impact on the category, the noise made it impossible to know which was which.
We created a COVID Classifier to enhance TPV and filter out short-term noise from COVID-19 conversations. It means only the trends that are most likely to sustain growth are prioritised, giving marketers and innovators the insight and confidence to know which trends to plan and activate against.Find out more
What are the benefits of this approach?
Ethnography at scale
No more vanilla or biased outputs by asking leading questions. Our insights are derived from organic consumer conversations using machines to find underlying patterns and connections that human analysts can’t see.
High-quality data means high-quality insight
The accuracy of our data is the foundation for our 90%+ accurate predictions. We achieve this using our proprietary technology which enables us to create the most accurate Social datasets in the market.
Never miss a trend again
Our AI approach means you don’t need to be aware of a topic to analyse it. Instead, every category trend which consumers are talking about is identified and prioritised in one simple metric. This is how you can identify ‘unknown unknowns’.
Scientific predictions not prophecies
Our TPV model accurately predicts the trends with the highest ranking future trajectory. This gives your brand a competitive advantage through faster, fact-based decision making on which trends to back.
- Pre-emptive questions validate what you already know
- Prompted answers = ‘claimed’ behaviour
- Human analysts interpret and make sense of the data
- Offers breadth or depth – but not both
- Narrow view of a moment in time
- Retrospective results, gives context to the now
- Minimises risk
- Ethnography at scale surfaces new and unknown insight
- Unprompted observations = ‘actual’ behaviour
- Machines scientifically analyse underlying relationships in the data
- Both robust and granular
- Looks forwards and back for a wider perspective
- Scientific prediction, lets you action the future
- Maximises opportunities