Gomez Scorecard

Despite some conflict-of-interest grumbling in the brokerage community (top-ranked broker, E*Trade chalked-up nearly a half-million dollars in consulting fees with Gomez in 1999), Gomez Advisors  www.gomezadvisors.com  has achieved near-adjective status in the financial services community, e.g.,“We are working on increasing our Gomez.” The company’s quarterly rankings of Internet stockbrokers, banks, mortgage lenders, and credit card issuers, are widely sited in the press and touted on the home pages of the winners. Even a high score in a single category can be leveraged into PR. USABancshares, despite finishing 20th overall, and 3rd in cost, rotates this message on its home page, “Twice Ranked #1 for overall cost by Gomez.com.”
 

News flash on the home page of USABancshares.

To be included in the Gomez Scorecard, banks must provide Web-based account access and bill payment. Gomez calculates an overall score based on ratings of five attributes: ease of use, customer confidence, on-site resources, relationship services, and overall cost. The company assigns a score through direct observation of the bank’s Web site, results from a questionnaire sent to the bank, and interviews with bank customer service reps. Table 18 on the next page shows the rankings for fourth quarter 1999 (Note: the Spring 2000 Scorecard is now available at gomez.com). Table 20 summarizes bank scores by Web platform vendor.

Table 1

The Internet Bank Ratings

overall score on the Gomez Advisors scorecard

Sources: Score -- Gomez Advisors, 1/00, <gomez.com>;
Web platform vendors – Online Banking Report, 4/00
¹Overall score from Gomez; see its Web for scores across all five attributes as well as scorecards designed for specific customer segments, Contact: John Robb, jrobb@gomez.com  (978) 287-0095;

²Web platform = vendor for the transaction area of the Web; many of the “in houses” are in-house systems

³Security First Network Bank, <sfnb.com>, is an Atlanta-based subsidiary of Royal Bank of Canada

Cross-referencing Gomez ratings with our database of Net-banking platform vendors, we generated Table 19 showing the average Gomez score for each vendor. However, the sample size is so small for most vendors, it’s difficult to draw many conclusions. It is evident that banks using S1 or S1-owned Edify score consistently higher than average. But we don’t really know the cause and effect, e.g., does the S1 platform cause Gomez scores to rise, or do banks that would normally score high on Gomez choose S1 for other price/performance reasons.

Perhaps a better indicator of platform performance is the ratings assembled by the consulting firm Celent (Table 4). In its interviews with banking execs, Celent found a strong preference for Corillian among large and medium banks and Digital Insight. among community banks. But there are so many variables in vendor selection that you must take these popularity contests with a grain of salt.

Table 2

Largest Public Net Banking Vendors*

dollars in millions

*The majority of their revenues derived from Net banking software

Table 3

Internet Bank Service Providers

overall score on the Gomez Advisors scorecard

Source: Gomez Advisors, 1/00; OBR, 1/00 (see table 11 footnotes)

¹Provides the platform for the top site; ²Merged with S1; ³Category includes banks that use 100% proprietary systems, those that have a blend of vendor/custom software, and those that have not publicly identified the software used; ⁴Merged with Digital Insight

Table 4

Platform Rankings According to Celent

overall score on the Celent scorecard

Source: Celent Communications, 2/00, <celent.com>

Note on Web Traffic Methodology

Web traffic data is less precise than it may appear. PC Data derives Web traffic numbers from a sample of 100,000 primarily home-based users whose actual surfing habits are tracked by PC Data software. A Web site must have more than 100 visitors across the panel before being ranked. Sites below the threshold show as blank in our charts, even though actual traffic could be as much as 100,000 users or more. Another difficulty interpreting the trends is that PC Data has altered its methodology and panel make-up over time, so results in the current month aren’t directly com-parable to previous month’s data. Readers must be careful in reading too much into a given data point. For the complete methodology, see  www.pcdataonline.com/methodology