What marketing decisions are most critical for your organization?  Are these priorities predictable or do they fluctuate with day-to-day changes in your market?  Do you have the right data, tools and support systems at hand to make the best decisions on an ongoing basis?  Is your organization wired to optimally deploy these tools to enable different organizational silos with access to  a common view of your demand environment?

Unique challenges require customized solutions

Increasingly, sales & marketing decision makers find themselves in need of highly customized solutions to the problems that are specific to their organizations and their place in the value chain.  Manufacturers of food and beverage products may primarily be concerned with making more productive trade spend decisions, while CPG producers might focus on shoring up brand equity for premium products facing competition from substitute offerings.  Wholesalers may prioritize increasing the value of each transaction basket by inducing customers to purchase products from them that they currently purchase from competitors.  And retailers might want to better understand how their promotional campaigns resonate with target customers and what they can do to earn a better return for each campaign dollar.

These challenges are distinct, yet they share two common attributes.  First of all, whatever problem you are solving depends in turn on a multitude of factors including those you can control – setting prices or allocating advertising dollars for example – and those which you cannot control like competitor actions or seasonal influences.  You need analytical tools that can help give you a holistic understanding of these factors rather than merely isolated fragments.  A holistic understanding comes from the ability to disentangle the many different factors at play and focus on the ones that have the most impact. For example, how do pricing decisions interrelate and influence product campaigns, and how can these decisions be calibrated to optimal effect?

Second, the challenge you face for one set of customers and products is not the same as the challenges you face for other customer-product combinations.  You cannot solve nuanced micromarket problems with solutions that play to aggregate market averages.  Every sales & marketing decision you make should proceed from the individual behavior of every customer in relation to every product.  Perhaps you sell something as thoroughly commoditized as salt or margarine.  There still are likely to be substantial areas of variance between the needs and preferences of any two of your customers in relation to every purchase they make of these or other products.  You will be well-served by analytical tools that help you see and quantify those differences on the problem at hand.

Piecing together the micromarket story

Ten years ago opportunities hidden in the fabric of your micromarkets would most likely have remained undiscovered.  Today we have robust computational technology to help unearth them.  We really can view the most basic units of activity in our markets.  But that by itself is not enough for us to turn insights into actionable guidance for decisions.  The challenge is this: from the information we now have at micromarket level, how do we piece together an understanding of what really drives demand?  How do we distinguish between the truly relevant factors that affect sales outcomes and those that are in effect little more than noise?

Every record in our purchase history has a story to tell, but the story is often incomplete.  Some product-customer combinations have many data points resulting from frequency of activity over close time intervals.  Others – the so-called “long tail” – offer up very little to help us put the story together.  Incidence of customer-product activity in the long tail can be so infrequent that making any truly insightful analysis would seem to present an impossible challenge.  How can we accurately quantify the effects of a price reduction, sales promotion, seasonality trend or something else when there are only two or three data points over the course of an entire year?

This is where the “science” of scientific marketing earns its name.  Problems like data sparsity can be solved with the application of advanced methods such as Hierarchical Bayesian modeling, which “borrows” information from similar transactions to develop demand models with associated probabilities for specific outcomes.  Think of these methods as storytelling aids, helping us to fill out and enrich those incomplete snippets of information found in the transaction records.

Connecting decision points in the organization

Robust computational technology can help us peer into the micromarket and see activity  at the most granular level of every customer and every product.  Advanced science can help overcome problems like data sparsity and identify what is driving customer behavior in the unique circumstances of your own market.  But there is one final challenge to address: how to connect all the different places in the organization where marketing decisions are made and align them for optimal responses to the opportunities which present themselves every day.

Marketing decisions typically are not made in one place, but rather in many places throughout the organization.  Pricing analysts as a rule don’t work within the same formal departmental structures as salespeople, advertising managers don’t share common information channels with trade spend executives, and so forth.  The advantage of understanding the factors that influence customer behavior can only be converted into predictive actionable recommendations if all the decision-making agents relevant to these areas are able to look at the same data and align their decisions for optimal effect.

Now, no piece of decision support technology can make people in different silos talk to each other.  To a large extent the capability to overcome the silo mentality lies within the organization’s own communications processes and related activities.  Most likely, organizations that are serious about this will be investing in some form of enterprise-wide technology like Enterprise Resource Planning (ERP) systems to ensure that the right data are available to anyone who needs it anywhere in the organization.  To provide the most precise guidance for customized marketing decisions, the analytical tools and support systems you use should anticipate the existence of an enterprise-wide platform acting as the system of record, and be able to seamlessly link into it.

Customizing micromarket-level decisions is not easy.  You need a combination of robust computational capabilities to crunch billions of terabytes of data, creative scientific approaches to solving daunting practical challenges like long-tail data sparsity, and organizational alignment to facilitate optimal decision-making from integrated decision support systems.  As tough as the challenge is, the effort has the potential to return a substantial multiple of your investment by directly addressing and providing guidance on those problems that are your own top priorities, in the context of your own unique marketplace.