It's not enough to know how your brand is selling. Increasingly, managers need to know what the marketplace's perception of the brand is. This is critical for everything from fixing flaws to revamping advertising. After all, you can't manage what you can't measure. The real question is: How do you go measure brand perception?
Until the advent of the Internet, perception measurement was anecdotal. A pharmaceuticals brand manager might go to a physicians' conference and talk with as many doctors as possible. A consumer brand manager might poll field managers on what they heard on the "street" or subscribe to a clipping service to see product mentions in magazines or newspapers.
Previously, some data related to perception arrived through quantitative means. For example, pharmaceuticals managers studied physicians' prescribing patterns using data purchased from third parties. Marketing firms generated coupon offers based on shopper profiles. Special purpose quantitative studies, some based on focus groups, could build a data profile on various brands.
Over time, brand use data accumulated, but slicing and dicing this wealth of statistical information doesn't really tell managers why people buy a product, or why doctors prescribe a drug. It certainly can't predict future patterns-positive or negative. Medical journal articles praising a competitor's product could change prescribing patterns within weeks ... or even days. A cascade of negative blogs could impact consumers who increasingly Google product names before purchase.
So how do companies successfully track perception? They do it by combining natural language searching capabilities with statistical analysis and then formatting it a way that any brand manager can easily understand and share with peers and executives. This is a critical research frontier for companies. The Aberdeen Group estimates that leading companies gain significant advantage over trailing companies in the capture and measurement of unstructured text and are actually four times more productive at this kind of measurement activity.
The Internet: a free and highly searchable database
For a pharma brand manager, the simplest way to track prescribing patterns has been to buy third-party data on physician prescribing practices. This is helpful, but not comprehensive. Some doctors opt out of reporting and some countries don't allow data to be shared. More importantly, data doesn't tell why a doctor prescribed a certain drug.
Pharma salespeople can influence prescribing patterns, but articles in leading journals pack a more potent punch. With the right solutions, companies can analyze journals to find out what physicians say about their drugs. With graphic presentations, brand managers can discover which physician thought leaders write about their drug, how they write about it, and who they typically co-author studies with. This gives brand managers the knowledge to best promote their products.
Mining call centers and consumer generated media
Consumer product manufacturers have different issues: To save money and shore up perception, they need to quickly find and fix product flaws. They also need advance knowledge of customer behavior and perceptions. Companies in this sector-particularly those building large products under warranty-look closely at warranty data for early warning of issues that need to be fixed. The most successful of these solutions use text analytic methods to overcome complaint-coding errors and quickly unearth problems.
While this analysis is critical, it can't warn companies as early as a solution that searches the Internet for complaints and criticisms. Sometimes consumer opinions vented on the Internet aren't related to a warranty issue. Maybe refrigerator produce bins stick, or the leather in a car ages poorly. Maybe it is difficult to get computer support staff on the phone, or running shoes fall apart faster than comparable brands. Many of these conversations are captured in consumer blogs or other Internet sources.
By applying a text analytics solution to vast quantities of Internet information, brand managers can assess how they compare to competitors' brands and what issues are potentially driving future customers. This can also be applied to customer interactions with your company-via Web or call center.
Surveys can mislead
If you work in brand management, you might think, "we've got surveys to cover that." But survey participation is typically 10 percent or below. Surveys might also capture people who are either really unhappy or very happy-missing those who fall somewhere in between.
The potential for analyzing non-quantitative text is limitless, but only if companies choose solutions carefully. The best solution is one that not only works with non-quantitative text, but can draw in quantitative results to produce graphics that demonstrate both perception and reality.
About the Author
Barry deVille (firstname.lastname@example.org) is a field technical consultant for SAS. He is an expert in pairing quantitative and qualitative data, and visual presentation of such data. For more information, go to www.sas.com
Please note that the Viewpoints listed in CRM magazine and appearing on destinationCRM.com represent the perspective of the authors, and not necessarily those of the magazine or its editors. If you would like to submit a Viewpoint for consideration on a topic related to customer relationship management, please email viewpoints@destinationCRM.com.