Do you know what drives your customers' behavior? More specifically, do you understand the impact each marketing investment you make has on your customers' incremental buying behavior? Intuitively, you know this is a complex equation that depends of both the type of marketing treatment that reached the customer and how much time elapsed between the marketing treatment and buying event.
Unfortunately, until recently, you have been handicapped by a lack of tools of analyzing customer behavior. Hence, as marketers, we typically rely on simplistic methods for aligning marketing activities with sales results. Techniques that we intuitively know are wrong: crediting "last click" with most or all of the order value, evenly splitting credit amongst all marketing treatments within a timeframe, arbitrary rules (often based on the opinion of the most powerful person in the room) for allocating order credit to various marketing activities.
Enter Big Data
With the advent of big data, it is now possible to analyze all your marketing activity, every email, every web impression, every click and more, to build a holistic picture of how marketing impacts customer buying decisions. Analyzing this data will quite naturally lead you in the direction of advanced statistical techniques, perhaps exposing you to terms like "survivial model" but fear not, with current technologies, this brave new world is now accessible to you without needing a PhD in statistical methods!
Tip #1: Embrace complexity... your customer isn't a simple animal!
As leading marketers learn to "let the data do the talking", they are often surprised by the results. Intuitively, we all know that much of a customer's decision to purchase a particular product or service is beyond our control (for example, I'm going to buy flowers on Valentine's Day regardless of who does or does not market flowers to me). This realization that marketing isn't 100% responsible for a customer's buying decision (or even 150% in those organizations that double-count sales as an attribution technique) can be a big scare for some marketing organizations. Naturally, a model is only as accurate as the data you give it so be as comprehensive as possible when thinking about what data to look at. This will lead to a more accurate model and a better understanding of your specific customers. Be prepared for some surprises - not everything you have been doing will look positive under this new light, but savvy marketers understand that improvement can only come if they are willing to change. Avoid defending the status quo and be prepared to look at your marketing investments in a new light.
Tip #2: Trust your data and your model
Armed with a new understanding of your customer, it's time to look for results. Remember, information can lead to insight but only action can lead to results. As marketers, we all know that results are what matters. Models that explain the past can give us the information critical to gleaning new insight about our customers. However, models that help us predict future behavior can drive breakthrough results. Leading marketers understand that trusting your data doesn't mean betting the farm on some new fangled modeling technique. They combine the recommendation from new predictive models with classical techniques and A/B testing to accelerate the adoption of advanced attribution modeling inside their organizations.
Tip #3: Don't settle for a rear view mirror approach to modeling customer behavior; insist on actionable information that you can put to work to drive results for your organization.
Most marketers still use outdated attribution methods to divvy up orders, but don't know how to apply their learnings going forward. The bottom line is that big data and advanced statistical techniques now offer all marketers a level of sophistication that was previously available to only the largest of organizations. With these new tools, marketers can identify the right customers to contact in the channel that each customer is most likely to respond to at the right time for the customer.
Accurately measuring marketing effectiveness has important implications for a company. DataSong's patented approach is actionable and accurate by modeling behavior at the individual consumer level. http://www.DataSong.com
Article Source: http://EzineArticles.com/?expert=Traci_Chu