Online retailers like Netflix and Rhapsody generate nearly as much revenue from unconventional titles as from mega-hits. Having escaped the shelf-space constraints of bricks-and-mortar stores, retailers are finding unforeseen moneymakers among their vast inventories.
Chris Anderson, editor of Wired Magazine, coined the term “the Long Tail,” referring to the myriad niche products whose collective market share rivals traditional blockbusters in the online market. Retailers like Amazon.com have found indie films and obscure music to be great moneymakers because the cost of inventory and shelf space approaches zero.
However, applying the Long Tail to most online or traditional markets faces one major obstacle: By selling online, the distribution processes have become more efficient, but marketing has not. Businesses seeking to tap specialty markets face the same conundrum: How do I cost-effectively reach the right prospects? Today, niche markets are built mainly through word of mouth and online recommendations, not the ideal approach for maximizing sales.
Businesses first need to understand who is the ideal customer for their products. Advanced predictive modeling holds the key. Specialty marketers traditionally have mailed to targeted lists based mainly on a prospect’s affluence profile. For example, a housecleaning company sent mailers to every homeowner earning more than $100,000 yearly. This can be costly and inefficient.
Using predictive analytics, the company learned that homeowners had a much higher tendency to use cleaning services if they had a child. If the child was younger than 6, the likelihood increased. If the child was a boy, response rates improved further. It was even better if the family had a dog, and still better if the homeowner were a single parent. With this knowledge, the company weeded out large categories of people from its mailing list who were unlikely to respond. It created a precise model of its most likely, and most lucrative, customers and avoided mailing to households unlikely to respond.
It doesn’t require a plethora of customer information to create such models. Demographic data are widely available for most households, and new technologies such as genetic algorithms automate the predictive modeling process. By coupling existing customer lists with this rich demographic data, businesses can glean a wealth of knowledge about those most likely to respond to their campaigns.
Instead of spending money marketing to the masses, many marketers are cultivating these lucrative “micromarkets.” By uncovering those most likely to have an affinity for their products, businesses can realize the promise of the Long Tail.
