Learn more about the RichRelevance DataMesh Cloud Platform Watch the RichRelevance Product Suite Overview Watch the RichRelevance Story Avail is now part of the RichRelevance family Learn more about DataLove Learn more about the RichRelevance DataMesh Cloud Platform Watch the RichRelevance Product Suite Overview Watch the RichRelevance Story Avail is now part of the RichRelevance family


We exist to help shoppers find what they need—quickly and easily—as they shop online or in-store.

As a B2B services company, all of our personalization technology helps our customers—leading retailers such as Target, Walmart and Marks & Spencer, as well as brands such as Patagonia, Disney and Williams-Sonoma—improve shopping for their customers.

Our team's retail DNA means we thoroughly understand the complexities and challenges of omni-channel. Our core team was responsible for key technology developments at Amazon, Overstock, PayPal, eBay and Yahoo!. We built our infrastructure directly leveraging the experience and knowledge from these retail powerhouses. RichRelevance is recognized for our expertise across the industry. Recent accolades include: Intel Premier IT Knowledge Award for Innovation; #1 in Personalization by Internet Retailer.

Check out our Engineering Blog for our latest thoughts on technology.


InfrastructureUnder the HoodAlgorithms Away!Omni-ChannelBlocking & Tackling


RichRelevance Infrastructure

Our Global Footprint & Network Architecture

Underlying Technologies

{rr} Infrastructure

Google and Amazon have measured that every one-tenth of a second (100 ms) delay in latency equates to 1% decline in sales.

RichRelevance consistently delivers personalization actions in 65 ms or less. This blazing-fast service is delivered via a proven infrastructure of 10 globally-distributed data centers to ensure 100% up-time.


Global Footprint

The IT team at an $800M online merchant reported that page load times for a top‐three competitor (built on a single data center) were on average 450% slower (400 to 500 ms) than RichRelevance’s average of 40 to 80 ms—one of the key reasons this merchant switched to {rr}.

InfrastructureUnder the HoodAlgorithms Away!Omni-ChannelBlocking & Tackling


What's Under the Hood

Hadoop and the Open-Source Ecosystem

Using Hadoop, Kafka, Hive, Pig, Hbase, and R as our underlying open-source technologies—for reliable, scalable, distributed computing—means we’ve chosen non-proprietary, less expensive, and more accessible ways to develop our software. While our customers’ needs always come first, we actively contribute our knowledge and expertise to improving these collaborative tools so the larger open-source community, which includes academics, researchers, and non-profits, can innovate as well.

To give you a further idea of scale, we use the same distributed computing infrastructure as Google and Facebook, and we rank #7 in use of Apache Hadoop nodes (following Yahoo, LinkedIn, Facebook, and eBay).


How We Got Started

Our collective involvement with the open-source community began in 2008. In our CEO’s words:

“We’d just gotten Sears and Walmart live in 2008. Scott Carey (Principle Software Architect and Apache Avro Project Management Committee Chair) and Tyler Kohn (VP of Engineering) came to me and said, 'We’re going to start migrating to this thing called Hadoop.' I pointed out that we needed to align our technology investments with our revenue, but they pushed back, and I finally conceded. And Hadoop has been the key for us in a way I did not anticipate. Now when we go the Strata Conference, we’re the pretty girl at the dance. We’re one of the largest implementations of any NoSQL technology anywhere.”

– David Selinger, CEO

Our Open-Source Leadership

We continue to offer expertise, experience, and leadership to various parts of the open-source community/ecosystem, from national conferences to local Meetup groups:

  • Darren Vengroff presented at the Big Data World Conference in London.
  • Tyler Kohn recently spoke at a Hive Big Data Think Tank Meetup.
  • We recently addressed a Bay Area Scala Meetup on what SCALA will and should look like five years from now.
  • Scott Carey is the current Avro PMC Committee Chair.
  • Murtaza Doctor (see video above) recently presented to the LinkedIn Kafka Meetup group. (We adopted Kafka even before LinkedIn officially released it.) He and Senior Software Engineer, Giang Nguyen, demonstrated at the Hadoop Summit how the company’s unique HBase architecture and design enable RichRelevance to capture and store nearly 1 billion daily events through HBase in real-time, including data ingestion, schema design, and access patterns, as well as how to trouble-shoot major problems like sharing and hot spotting.

“ Big Data technologies like HBase and Hadoop are making a dramatic impact on the broader technology community and the way we think about how to design and architect systems. As a transformative technology, we have invested time and energy in developing and nurturing the HBase and open source community. We look forward to sharing our knowledge and best practices with others to continue to help innovate the HBase, Hadoop and broader Big Data ecosystem.”

– Murtaza Doctor, Principal Architect

InfrastructureUnder the HoodAlgorithms Away!Omni-ChannelBlocking & Tackling


Algorithms Away!

The Science Behind {rr}’s Unique Ability to Create 1:1 Predictive Personalization

  • Ensemble Learning

    The enRICH Personalization Engine is the first commercial application of ensemble learning. The approach is the only one that facilitates competition among 100+ independent algorithms (recommendation types), wherein each makes use of different kinds of user behavior and catalog data to select the highest-performing strategy for each unique placement. (The FAA uses ensemble learning for airspace management when multiple requests for airspace use need to be automatically reconciled and managed.) In contrast, most recommendation systems leverage one highly complex algorithm for use across the entire customer base. The enRICH Personalization Engine performs thousands of multivariate experiments to decide, in real time, which algorithm best matches a particular customer’s needs at a specific place and time. On many pages, several recommendation placements enable us to display a combination of high-performing messages, based on the best-performing algorithms.

  • Personalization and the wisdom of the crowds

    The enRICH Engine doesn’t just depend on the wisdom of the crowds to make recommendations relevant to each customer. We also consider current and recent site activity in our recommendations, so we serve as a real-time aid in the customer’s discovery process.


    Even our largest retailer integrations are completed and deployed within 4 to 6 weeks (or even as fast as 2 to 3 weeks, depending on individual variables). No other enterprise e‐commerce provider—let alone personalized recommendations provider—can make the claim.


    Our response to real-time intent and customer micro-trends is unparalleled in the industry. Our enRICH Personalization Engine rebuilds product recommendations up to 12 times a day, based on complex mathematical models—adjusting for subtle changes in shopping behavior, inventory, pricing and more. The result is sustained relevance without retailers needing to manually manipulate content.


    As shoppers interact with recommended products, RichRelevance’s built-in feedback loops inform the system about how products and recommendation types perform.

  • Data Integration

    To deliver more relevant recommendation types and content to shoppers, the enRICH Engine lets retailers integrate data from offline transactions with existing data on individual and group behavior. The technology can also be integrated with marketing tools such as email, site navigation, ratings, and reviews.


    We are the only vendor to provide third-party developers with access to the data and technology necessary to create new applications that leverage consumer behavior. This introduces limitless possibilities for retailers and brands to access data inputs such as mobile and social that require user-based filtering, and the display of product catalog or ad inventories. Learn more about the RichRelevance DataMesh Cloud Platform.

InfrastructureUnder the HoodAlgorithms Away!Omni-ChannelBlocking & Tackling


Omni-Channel (Including Mobile)

Consumers Demand Seamless Connections Across Devices …
And They Expect the Same From You

The volume of data consumers are creating is growing exponentially. Now, information comes from website info; PCs. tablets, and mobile devices; and e-mail. With new smart phone technologies integrated into cars, with TVs now as powerful as computers, and with the advent of wearable devices, data volumes will continue to expand for the foreseeable future.


Retailers and brands need solutions that can address this challenge and operationalize responses to it. Companies that want to personalize everything—across all channels—for their customers will need partners flexible enough to scale and grow with their businesses. It's paramount that these partners see each new channel as an opportunity.

InfrastructureUnder the HoodAlgorithms Away!Omni-ChannelBlocking & Tackling


Blocking & Tackling

Reporting, Analytics, and Merchandiser Dashboards


    The enRICH Engine’s unique architecture lets us provide merchants with immediate, tangible feedback. Our site analytics tools provide precise insights into website traffic and the effectiveness of enRICH solutions on a retailer’s site. Modeled after industry leading analytics interfaces, our dashboard integrates easily with any web analytics platform and offers filtering options, graphs to visually identify trends, and easy export of reports in Excel format.


    While the enRICH Personalization Engine does all the hard work—automatically personalizing each customer’s shopping experience with relevant recommendations—online retailers can control how and when items are featured by any of the enRICH solutions. The {rr} Dashboard allows retail merchandisers to “tune” the enRICH Engine to achieve specific revenue, profit, and conversion objectives for individual products, brands, categories, and page types. For example, retailers can easily adjust personalization strategies to account for high-margin products, excess inventory, black-outs; and product pairings. The Dashboard provides transparency into the performance of the enRICH Personalization Engine while enabling the controls essential for online retailers to achieve their business objectives.


    Our Shopping Media solution provides retailers and brands new ways to maximize profitability and optimize KPIs by leveraging real estate on and off retailers’ sites to attract both customers and brand/manufacturer dollars.

InfrastructureUnder the HoodAlgorithms Away!Omni-ChannelBlocking & Tackling

RichRelevance allows us to showcase products that meet our customers’ needs by providing insight and offerings that are relevant to them on an individual basis…RichRelevance impressed us with their technology, results, and pedigree in retail, as well as their commitment to us as a partner. We had finished a testing solution and were planning to continue testing in a variety of solutions from different vendors. But after working with RichRelevance, we knew we had found what we were looking for. RichRelevance was able to deliver the right solution for both our internal business needs and customer expectations.

— Greg Anderson

Head of Ecommerce, Motorcycle Superstore

RichRelevance has customization abilities that are 4 to 5 times that of competitors, and has targeting capabilities built into the solution—with no separate vendor needed—and that was very appealing. The system had the equivalent of a lot of hard coding we were already doing. The fact that copy could be swapped out by our merchandising team and there was no heavy JavaScript required was ideal.

— Brandon Proctor

VP of Marketing, Build.com

Contact Us