Maximizing Insights from Customer Satisfaction Data

Market Metrix is a leader in feedback solutions for the hospitality industry, with more than 100 enterprise clients, in 70 countries, from lodging, car rental, airlines, and allied fields.

Situation

Market Metrix specialize in “high touch” scenarios, where a superior level of personalization in customer service is required. Their internal teams have earned a reputation for innovative research to enhance customer and employee satisfaction. In 2008, Market Metrix began a successful collaboration with EC Wise and its China partner EC Wise Sichuan to match sophisticated human analysis capabilities with the right computerized tools.

Challenge

Market Metrix’s central service to clients is a highly customizable, web-based survey mechanism, which results in a wealth of data for analysis. They wanted to leverage the data in a manner as precisely targeted to its clients’ needs as are the surveys. They needed accessible reporting tools, both for their internal expert users and client end-users with less insight and ability. Additionally, Market Metrix envisioned a number of new products, including sale of research results from multiple sources and in new formats.

An important source of data for Market Metrix’ analysis is a web-based survey administered for Market Metrix by Harris International. This anonymous “benchmarking” survey makes Harris’ interactive polling engine and pool of participants available for comparison with Market Metrix customer survey data. Harris’ delivery of survey results was in an idiosyncratic format, which changed quarter to quarter. As with any anonymous survey, extremely poor data quality represented a constant challenge.

Market Metrix engaged the EC Wise team to help modernize their processes and pave the way for integrating their various offerings. Their primary goal was to standardize the Harris survey intake procedures and provide analysis models for the benchmarking data.Market Metrix also wanted to “enrich” the data from third-party sources, to be incorporated ad-hoc. As an important feature of the changes, the new analysis capabilities for the benchmarking data would drill-down to individual venues; previous benchmarking analyses were at brand or corporate level. For an optimal resale strategy, the analysis would offer further drill-down into individual responses.

Approach

The EC Wise team first analyzed the way Market Metrix produced surveys with Harris, and then designed a suitable data warehouse structure for the results. In conjunction with the data warehouse implementation, we created an ETL process that would adapt as the survey structures changed.

The team provided initial load processing for 5 years’ worth of previous survey results. Loading the historic data required arduous data cleansing of many types. The overall result was greater accuracy and a higher degree of confidence in any analysis results from this wide and deep data resource.

EC Wise examined the brand-to-property relationships in existing data, and compared them to an industry-standard third-party data source for hospitality brand affiliations and property-level data. EC Wise created over 20 algorithms to match source property designations in the external data against Market Metrix’s own view of properties and brands. Users prioritized the algorithms based on perception of their reliability, received match statistics and results, and refined their matching order to request new trials.

Market Metrix was able to use the result to normalize their brand and affiliation designations. The combined and branded data resource offers not only a new, up-to-date set of survey selectors for Harris surveys but also an authoritative source for internal lookups of brand and property-level data for additional analysis.

Then EC Wise team members analyzed the survey metadata and designed a set of XML schemas for it. The team envisioned a phased introduction of the metadata. They provided a simple-to-implement change to the business process to start, but included a comprehensive plan for evolving Harris survey design to merge with Market Metrix’s standard administrative process for customer survey management.

With metadata and base methodology for data intake in hand, we provided a light-weight initial way to accessing the new repository. Using tutorials based on Market Metrix data and analysts’ wish-lists, Market Metrix users learned to use Excel to retrieve both the data warehouse relational layer and the cube layer. The approach allows Market Metrix staff to experiment and provide feedback, while the team continues to refine their design. It also allows Market Metrix to prototype more formal reports for later implementation.

Resolution

Market Metrix is succeeding with the new benchmarking repository. It loads Harris data each quarter at a low cost, in a consistent process, with increased operational efficiencies, and to normalized storage structures. Market Metrix derives new, publishable insights, and actionable recommendations for its client base. They have achieved a different form of partnership with Harris, and a better model for expectations in dealing with other data vendors in the future.

With the new repository, Market Metrix has added the desired property-level analysis capability to their previous overview of brands and affiliation. They use the new analysis cubes at this level to capitalize on its market-leading sophistication in research areas such as customer loyalty. With finer control of the benchmarking data, they find new productizing opportunities for the Harris data.

EC Wise’s business user-centered design process enables Market Metrix to drive requirements at the pace they need. EC-Wise’s tight coordination and unique blend of US and off-shore talent make the process cost-effective and high-quality.

The benchmark resource provides a cleansed, authoritative and accessible basis for hospitality data for Market Metrix customers and in the general industry pool. Its organization facilitates the resalable retrieval of both raw data and informative metadata. The cubes allow Sales staff to gauge the availability of different slices of data and for pricing and packaging. After the sale, Market Metrix reaches back into the data warehouse for the purchased details.

Market Metrix’s internal database design staff will be counting on this accessibility, as they embark on their new version of their customer-facing product lines. Afforded organized feedback from their own customers, able to compare themselves against the correct competitive market segments, and empowered with well-presented analysis, predictions, and suggested remediation, Market Metrix’s clients will have a complete business intelligence view of their customers’ experience and needs.