Home/Solutions/Digital Experience Index (DXI)

What is DXI (Digital Experience Index)?

DXI is the single, explainable score at the center of Quantellix.AI's platform. This page defines exactly what it is, how it's calculated, and how ecommerce, travel and payments teams use it day to day.

Short answer: DXI is a 0-100 explainable AI score, generated by Quantellix.AI's LNM-AI model and SXI++ engine, that predicts how likely a specific website visitor is to buy — broken into four sub-scores (Visual, Behavior, Transactional, KPI) so it's never a black box.
Quantellix.AI·Reviewed by the Quantellix.AI product team·Updated September 2026

DXI in one sentence

DXI (Digital Experience Index) is a 0-100 score that Quantellix.AI assigns to every visitor on your website, predicting how close that specific person is to buying — and explaining exactly why, instead of hiding the reasoning inside a black-box model.

It is the core output of LNM-AI (Quantellix.AI's Large Numerical Model), which reads every click, scroll, search and session on your site, and SXI++, the engine that compresses those signals into the single explainable number your team sees.

0-100
DXI score range
4
explainable sub-scores
20-80%
typical conversion lift
25+
live deployments

The four sub-scores that make up DXI

  • Visual DXI — how visitors perceive your site's design and clarity.
  • Behavior DXI — how visitors move, scroll and search across your pages.
  • Transactional DXI — how visitors actually behave once they reach checkout.
  • KPI DXI — how that engagement ladders up into revenue, AOV and repeat purchase.

Each sub-score is reported separately, which is what makes DXI explainable: your team can see which layer is losing value, not just that a visitor scored low overall.

How DXI is calculated

DXI is generated by a three-agent pipeline. A data ingestion agent pulls in multi-modal signals — clicks, scroll depth, search terms, images, even A/V where relevant — into one clean dataset. An AI scientist agent applies SXI++, a blend of proprietary machine-learning and statistical methods (refined with reinforcement learning), to turn that dataset into the four sub-scores and the overall DXI number. A guided-execution agent then turns the score into a recommended next action, with a human always in the loop.

"DXI is the credit score for digital engagement — for any platform, any industry." — how the Quantellix.AI product team describes it internally.

What a DXI score actually means

A DXI of 78, for example, means that visitor sits in roughly the 78th percentile for propensity to buy, based on how their Visual, Behavior, Transactional and KPI sub-scores compare to your historical buyer population. Scores like this let a team retarget the highest-intent 20% of visitors, rather than treating every visitor identically.

DXI vs traditional web analytics

  • Traditional analytics report totals and averages after the fact; DXI scores individual visitors in real time.
  • Traditional dashboards need a human to spot a pattern; DXI already ranks visitors by propensity to buy.
  • General-purpose LLMs and AutoML tools are typically less explainable; DXI is explicitly built to show its reasoning, sub-score by sub-score.

Frequently asked questions

What is DXI (Digital Experience Index)?
DXI is Quantellix.AI's explainable, 0-100 AI score that predicts how likely a specific website visitor is to buy. It is built from four sub-scores — Visual, Behavior, Transactional and KPI — so a single number never behaves like a black box.
How is DXI different from Google Analytics or a CDP?
Analytics platforms and CDPs report what happened in aggregate. DXI is a per-visitor, forward-looking propensity-to-buy score, generated in real time by Quantellix.AI's LNM-AI model, so a team can act on an individual session while it is still happening.
How accurate is DXI?
Across Quantellix.AI's benchmark set of ecommerce case studies, DXI-based scoring reached 98-99.8% accuracy on visitor conversion and pricing-efficiency metrics, outperforming general-purpose LLM and Auto-ML baselines in the same tests.