How a tech major removed cognitive dissonance in buying journeys and created $1B in incremental CLV
The Big challenge
What is Cognitive Dissonance?
An ambiguous but very real set of reasons prevents customers from furthering or completing their purchase journey. Typically – confusion, lack of information, too many or insufficient choices, cumbersome navigation, and poor artifacts. 70% of buying journeys start digitally; however, each session results in only a 1% purchase rate. The reasons are difficult to understand, are highly dispersed, and hidden deep behind behavioral patterns left behind by the customer. This is especially true with complex purchases – financial products, insurance, telecom, technology, healthcare, etc.
Fractal developed and deployed an innovative patent-pending technology called AIDE to uncover the reasons for dissonance and enabled operationalization that resulted in significant turn-around results of the digital-first buying journey.
A Fortune 100 B2C Tech company that sells hundreds of products, subscription plans, and accessories, with 1.5M average daily digital visits. Conversion rates hovered between 0.8-1.2%, and the omnichannel purchase was low. Therefore, there was genuine leakage of sales opportunity. Digital teams used 100s of reports detailing journeys, heat maps, exit rates – however were in the dark on the root causes and experiential improvements to provide. What works in the digital-first experience, what doesn’t, and why? What improvements to prioritize?
AIDE: Automated Insights for Digital Evolution
AIDE is patent pending AI technology that mines through millions of digital touchpoints, along with external and omni-channel data. Unique pattern recognition AI algorithms are deployed to uncover several micro-scopic factors that cause or prevent a sale. Such as:
- Anomalic deviations in hundreds of behavioral KPIs,
- Stage of journey that had the highest impact. Further, isolating the specific touchpoint from among thousands of possibilities,
- Root cause of the issue (as best as possible, we aren’t mindreaders!),
- Size of the impact on lost revenue, or increased costs.
AIDE assembles five AI modules, that are fully automated, and fully scaled up for millions of unstructured and structured data points created daily –
- Sensorize: An automated AI/ML pipeline to derive meaningful business indicators using the click activity across customer journeys.
- Detect: Deviations from expected behavior across digital journeys get captured by applying pattern recognition algorithms to the key digital indicators.
- Locate: A suite of supervised machine learning algorithms to identify drivers of key customer journey outcomes (drop-off, clear cart, etc.) and measure relative impact at a page & click level on a customer’s experience.
- Reveal: NLP module to perform sentiment analysis and entity extraction on the voice of customer data such as chat; feedback etc. to identify the root cause of the friction and generate actionable insights.
- Prioritize: Quantify the insights with respect to loss in revenue or incremental overhead costs to prioritize hypotheses for improving website design
AIDE’s uniquely visual; highly actionable granular insights added a critical layer of understanding of visitor’s digital behavior. AIDE also created an automated pipeline to measure performance of pages and content elements daily, which enabled the team to routinely generate new UX ideas instead of merely fixing bugs.
An Agile POC to Scale approach leveraged, results being iterated with business partners to create a highly intuitive consumption layer. A war room setting further helped absorb insights quickly, resulting in reducting in action from 2 weeks to 1 hour (in some cases).
Client’s in-house technology stack was used to build and deploy AIDE.