For those of a technical nature, SOMA implements complex probability and predictive analyses, such as Bayes' Theorem to improve the accuracy of results.
SOMA does not rely on any single data point to create organisational successes. Instead, SOMA measures the outcomes for every possible permutated variate step combination across all of your automated processes. It self-learns these networks to generate the best personalisations and branches for each User Cohort (auto-assigned too!) inside every one of your automated journeys, 24x7x365.
For instance, when SOMA was still machine-learning, for a subset of Cohort A, users were presented with a sequence X of personalised touch points over a set of self-determined time periods within an automated journey and SOMA measured improved outcomes over sequence Y for another similar subset of Cohort A. Once the optimised paths have been determined (primary learning stage completed) for each Cohort (based on predictive analysis), SOMA self-selects sequence X (representing specific personalisations, emails, SMSs and CTA variates) for all members of Cohort A in the future. However, it still randomly trials alternative sequence variants for a small percentage of users of each Cohort to ensure that previous AI-based determinations remain optimal over time.
If this fully automated processing increased your conversion rate and reduced your OPEX by just 5-10%, what would that actually represent in real terms for your organisation?
Chris Rodbourne, CEO S-Digital