Unlimited tests in any channel, Bayesian statistics, dynamic traffic allocation (MAB) and predictive targeting by segment. Client-side and server-side in a single platform.
The probability that a variant is the best one, and the expected loss if the call is wrong. Decisions come earlier and with more confidence than with classic p-value thresholds.
Multi-Armed and Contextual Bandit algorithms shift traffic to the winning variant in real time and account for user context. A/B testing and personalization at the same time.
A hypothesis that wins in one segment can lose in another. Read results by audience and serve each segment the variant that works for it.
Machine learning predicts which hypothesis will succeed in each segment and forecasts the potential uplift — testing becomes manageable and less expensive.
Gravity Field supports an unlimited number of tests across every format of online communication: content blocks, product recommendations, banners, pop-ups, notifications and many others.
Test any UX element in any channel: on your site, in your mobile app and even in physical stores — on self-service kiosks, on sales assistant devices on the floor and at the checkout.
Dynamic allocation is an engine built on machine learning algorithms — Multi-Armed Bandit and Contextual Bandit — that automatically serves the most relevant offer to each user at each moment.
The engine picks the optimal variant itself: it takes personal interests and targeting rules into account and automatically favors the variants that optimize your metric in real time — conversion, revenue, CTR and so on. That combines A/B testing with personalization in a single run.
Gravity Field analytics lets you review test results across different audiences. Segments often react differently to the same campaign, and a hypothesis that wins for one segment can lose for another.
Once you have statistically significant results for your key segments, you can serve each segment its winning variation and move on to testing new hypotheses.
Predictive targeting is the logical next step after A/B testing by audience. A machine learning engine predicts which hypothesis will perform best for each segment and even forecasts the potential uplift on your key metric.
That makes the testing process more manageable and helps optimize the cost of validating hypotheses.
With Gravity Field you serve relevant offers not only within a single page, but throughout the session and across following sessions too.
The platform lets you build an end-to-end personalized journey for individual segments across your entire site or app, and measure the effect of every interaction in the funnel.
Gravity Field can run client-side campaigns (in the browser) and server-side campaigns via API at the same time. Client-side gives you fast development and launch — ideal for quickly testing targeted interface changes.
Server-side campaigns deliver higher site performance and are used for substantial structural changes to pages.
We will show you how to set up an experiment, switch on MAB autopilot and read results by segment.