How We Differ from “Real-Time Experience” Engines (And Why You Should Care)

happy customers

Solid Customer Recommendations Are the Key

There are dozens of “real-time experience”(RTE) engines for e-commerce on the market right now, but they are all very different than Loyalty Builders. We solve different problems and are usually complementary to RTEs. But it’s a busy marketplace out there and it can be confusing. Here is a look at the RTE process – advantages and limitations – and how we differ, particularly on customer recommendations that reliably work.

The RTE Process

RTEs generally customize the e-commerce shopping experience by analyzing website activity. They are platforms that include mechanisms and templates for getting product recommendations in front of site visitors as they browse/shop. This includes emails triggered on cart or browse abandonment. The products that are recommended to a visitor are based on the most recent products viewed or placed in the cart by that customer, or on top selling or top trending products viewed or purchase by other visitors recently, or by specific “rules” defined by your merchandisers. It’s a reasonable way to improve the shopping experience.

Key Differences

Loyalty Builders (LB) also provides product recommendations, but that’s where the similarities end. Here are some differences:

  • LB recommends products based on actual purchasing behavior of individuals and compares that to all customers over several years. This is a deeper analysis of what actually results in purchases. RTE recommendations are based on most recent website activity. But unless they purchase, this is a step removed from real purchase intent. The value for predicting purchases is suspect and temporary at best.
  • LB can score all customers immediately, based on purchase history. They are always available for every customer. Most RTE systems require website activity to accumulate over time before it can start to infer what customers who behave in certain ways seem to want. An RTE system would have nothing to say about a customer who exhibits no recent website activity.
  • LB calculates every single customers’ probability to buy every product you sell and ranks those products as recommendations for each customer by probability to purchase in a given (configurable) time period. RTEs can look only to recent browse behavior, and would not know to make recommendations for products the customer has never browsed. One exception is rules your merchandisers maintain based on segments they define. But this is not automated and relies on merchandiser skill, not data-driven opportunities.
  • LB calculates loyalty metrics, such as potential lifetime value, risk of defecting, likelihood to buy in several different time windows, amount expected to be spend on next purchase for every customer. This is important for customizing messages and incentives by lifecycle stage and value. RTEs do not factor expected loyalty, value, or risk.
  • LB models and makes individual predictions on every single customer. RTEs tries to group customers into segments based on similar behaviors, and treats each of those segments the same.
  • LB recommendations are based on transaction records, which are always accurate, accessible, and complete for every customer. No personal customer data is required, and no additional data sources to integrate, manage, govern, and maintain..
  • Because LB looks at customer health and lifecycle trends over time in terms of purchasing, regardless of web browsing activity, it’s easy to address customers according to increasing or decreasing loyalty.

 Some questions to illustrate the differences:

  • Could an RTE be used to personalize email campaigns, i.e., send an email blast to all customers at once with the top 6 product recommendations that are unique to each customer, regardless of whether they have visited the website or not?
  • What would an RTE recommend to a customer with a purchase history, but no recent browse activity?
  • Could an RTE target and message customers differently based on lifecycle stage, expected value, and risk?
  • Could an RTE’s analytics be used to identify best customers for a product promotion (any product within minutes, any time)?
  • Could an RTE identify appropriate marketing budget for specific customers based on expected spending level, such as whether they are worth a catalog or direct mailing?
  • Could an RTE prioritize opportunities for account reps?
  • Does the RTE rely on cookies to track personal customer information?
  • How does the RTE factor in purchases that occur outside the e-commerce site?
  • What is the cost to maintain this system and feed it the data that it needs?
  • Is the revenue uplift (ROI) easily measurable and obvious?

Some RTEs talk about using social media data, weather data, and other data sources to make recommendations, but I’m willing to bet these features are not used in practice. You can check with your webmaster. If they were, there would be a lot more data integration, system configuration, and maintenance required than it’s typically worth.

In general, the problem with these “real-time experience” systems is that they must infer “preferences and intent” of visitors from their click/browse behavior and other time-sensitive inputs. Assuming someone wants to buy golf clubs because they visited a page displaying golf clubs is a leap of faith and fleeting at best.

Most of these vendors admit that only the most recent browse behavior — within a few days — is potentially useful. But if it’s good technology, it can increase conversions. The question is, how much is it costing in data sourcing and integration, system maintenance, etc., for the results you are getting, assuming you can conclusively measure and attribute the lift properly? Several of our clients are questioning the ROI from the systems they have tried.

Loyalty Builders is simply an information service designed to integrate with your existing marketing systems. On your website, our recommendations could be integrated into an RTE’s mechanisms for rendering recommendations and generating emails such that you get the best of both worlds — some products recommended because of recent browsing and others recommended because of our deeper analysis of actual buying. And as mentioned, our analytics can also drive email campaigns (we also integrate with ESPs), mobile, CRM, catalog lists, direct mail, etc.

For a small monthly fee, Loyalty Builders provides invaluable, multi-purpose customer intelligence to supplement and improve the results of whatever marketing to customers that you already do.


Tags :