SessionM launched in 2012 as a platform that increased user engagement by adding gamification and loyalty rewards to mobile apps. The system has since expanded to support more channels and message types. This puts it in competition with dozens of other customer engagement and personalization systems. Compared with these vendors, SessionM’s loyalty features are probably its most unusual feature. But it would be misleading to pigeonhole SessionM as a system for loyalty marketers. Instead, consider it a personalized messaging* product that offers loyalty as a bonus option for marketers who need it.
In that spirit, let’s break down SessionM’s capabilities by the usual categories of data, message selection, and delivery.
Data: SessionM can gather customer behaviors on Web and mobile apps from its own tags or using feeds from standard Web analytics tools. It can also ingest data from other sources such as a Customer Data Platform or CRM system. Customer data is organized into profiles and events, which lets the system store nearly any type of information without a complex data model. SessionM can also accommodate non-customer data such as lists of products and retail stores. It can apply multiple keys to link data related to the same customer, but requires exact matches. This works well when dealing with known customers, who usually identify themselves when they start using a sytem. Finding connections among records belonging to anonymous visitors would require additional types of matching.
Message Selection: SessionM is organized around campaigns. Each campaign has a target audience, goal (defined by a query), outcome (such as adding points to an account or tagging a customer profile), message, and “execution” (the channel-specific experience that includes the message). SessionM describes the outcome as primary and the message as following it: think of notification after you’ve earned an award. Non-loyalty marketers might think of the message as coming first with the outcome as secondary. In practice, the order doesn’t matter.
What does matter is that campaigns can include multiple messages, each having its own selection rules. Message delivery can be scheduled or triggered by variables such as time, frequency, and customer behaviors. This means a SessionM campaign could deliver a sequence of messages over time, even though the system doesn’t have a multi-step campaign builder. Rules can draw on machine learning models that predict content affinity, churn, lifetime value, near-time purchase, and engagement. Clients can use the standard models or tweak them to fit special needs. Automated product recommendations are due later this year. Messages are built from templates that can include dynamic elements selected by rules or models.
Delivery: Campaign messages are delivered through widgets installed in a Web page or mobile app, through lists sent to email providers or advertising Data Management Platforms (DMPs), or through API calls from other systems such as chatbots. Multiple campaigns can connect through the same widget, which raises the possibility of conflicts. At present, users have to control this manually through campaign and message rules. SessionM is working on a governance module to manage campaign precedence and limit the total number of messages.
The system can generate presentation-ready messages or send data elements for the delivery system to transform into the published format. It supports real time response by loading customer profiles into memory, limiting itself to information required by active campaigns. External systems can access the customer profiles directly through JSON API calls or file extracts, but not through SQL queries.
About that loyalty system: it’s sold as a separate module, so only people who need it have to pay for it. It includes the features you’d expect: points, promotions, awards, status tiers, reward redemption, and so on. SessionM added the ability to deliver and redeem personalized coupons through retail Point of Sale systems when it bought LoyaltyTree inDecember 2016,
SessionM has about 70 clients. The company originally sold to large enterprises, which are still about half its customer base. It is now pursuing mid-market clients more actively. The company has raised $73.5 million in funding.
_______________________________________________________________________
* You’ll note that I’m using “customer engagement”, “personalization”, “messaging”, and other system categories interchangeably. It’s probably possible to distinguish among them, but, in practice, all assemble a customer profile, use rules to select messages for individuals, and deliver those messages through execution systems such as Web sites. Most marketers will want to pick just one system to do this sort of thing, so they’ll evaluate vendors from all those classes against each other. This makes distinguishing between them largely an academic exercise.