Plug MARS into your existing marketing stack

Integrate your existing tools and send out the right discounts to your customers within your budget

INTEGRATE NOW

What's needed for MARS to work?

Data source

MARS uses several types of data to learn from and recommend discounts to maximise conversions with minimum spend

  • Transactional Data

  • Transactional data includes app events that are directly associated with revenue. These revolve around billing and monetary transactions.

  • Examples:

  • Order value

    Checkout started

  • In-App Behaviour

  • MARS collects in-app customer behaviour data in the form of specific app events.

  • Examples:

  • Products added

    Products searched

  • Realtime Campaign Data

  • MARS tracks realtime campaign data and changes in customer’s app behaviour during the course of the campaign

  • Examples:

  • No. of users converted

    Spend per customer

  • Discount Usage

  • MARS uses coupon and discount data to get a deeper insight into customers’ spending behaviour and coupon preferences.

  • Examples:

  • Discount value

    Minimum cart value

  • External Data

    BY MARS
  • All the learning from customer data is supplemented with external data provided by MARS itself to identify patterns associated with environmental changes like weather, traffic, world events etc

  • Examples:

  • Traffic

    Weather

  • API

  • The MARS API can be used to craft your own customer experiences eg. Personalised social media retargeting ads, physical discount coupons etc

  • Marketing Channels

  • Use Push, SMS, Email to send out MARS recommended discounts to your customers

  • In-App Discounts

  • Create personalised discounts for every user directly reflecting in your app UI

Output Channels

MARS recommendations need an output channel to send out offers to the end customer.

  • Individual Level Privacy Guarantee

  • Differential privacy ensures each individual gets roughly the same privacy that would result from having their data removed.

  • Highest Privacy model allowing best results

  • The statistical functions used by MARS in making discount recommendations do not overly depend on any single individual’s data. i.e. MARS’ output is not affected by removal of a user from the data set

  • No PII Information

  • MARS doesn't use any PII information in it's reinforcement learning models

Why trust MARS with MY data?

MARS uses a differential privacy approach to ensure an output of the highest quality with minimum privacy loss

List of Integrations

Don't see your integration here? Request one now!
REQUEST INTEGRATION

“In just one month we boosted user engagement and saw significant reductions in churn rate”

Sushant, Product Manager

Add MARS to your marketing stack today

INTEGRATE NOW