Expert Perspectives: Applying Data Science & Analytics to Your Sales Strategy
Learn from an industry expert about why you should apply data science and analytics to your sales strategy and why this shift will help you meet your goals.
On the surface, sales and data science might seem like oil and water, with sales focused on softer social skills and data science on harder quantitative skills. However, they can make powerful companions.
Specifically, data science can augment the existing sales processes and experience of the sales force to help make strategic decisions. Products and services are often highly complex, with customer needs similarly difficult to disentangle and understand.
In this fragmented environment, data science techniques can provide a sales force with the information to make better decisions about which customers to focus their efforts on, the ways in which to engage them, and how to best differentiate product/service value from competitors. Below, I discuss three key ways of applying data science and analytics to help your sales strategy achieve its goals.
1. Track propensity to buy.
Keep your finger on the pulse of the propensity for market segments. Take for example, Cisco; their sales and marketing force uses a series of propensity to buy (P2B) models to decide where they should focus time and resources. These models forecast demand for their products, everything from routers to servers and beyond. Cisco’s data science team builds these models by training them over their existing customer data to recognize potential customer characteristics and identify emerging trends.
By keeping tabs on the propensity for different market segments to buy their products, Cisco’s sales team can then focus on selling products to those with the greatest disposition to buy them, thereby producing maximum profitability. Furthermore, Cisco retrains these models every quarter, providing the sales team with the latest and most accurate information on their potential customer base.
2. Optimize sales channels.
Sales teams often employ many different channels to sell their products: retail, online, resellers, different geographic locations and so on. Each of these has the potential to reach different market segments and help scale a business, but how do you choose which ones to invest your time and resources into?
Data science and analytics techniques can be especially useful in optimizing various sales channels for your business. For instance, predictive models can be trained to identify possible channel partners based on their access to customers, finding those with the likelihood of buying your products. Once you have existing channel partners (for instance, individual retailers), you can use anomaly detection algorithms to compare their performance against regional and industry benchmarks, terminating the relationships that do not perform up to standard.
Identifying channels that diverge from the industry/regional standard may help identify under- and over-performance for a given Sales Channel. The graph illustrates how data from one channel could be compared with that of the industry standard via their probability distributions.
Additionally, within a given sales channel, optimization algorithms have been shown to help figure out the ideal number or subset of products to focus on selling (i.e., that maximize profit for given logistical constraints like production costs, storage costs, price, and inventory). By combining these approaches, a sales team can strategically identify partners and channels that perform well and maximize their profit within those channels.
3. Quantitatively identify unmet needs.
In order to close a sale and differentiate your value, you need the ability to address unmet customer demand in a way your competitors do not. The challenge becomes identifying these needs across so many fragmented sales channels (local, international, online, and more).
One way to better understand unmet customer needs in a cost-effective way is to use data science methods to analyze public opinion from social media services like Twitter. Natural language processing techniques make it possible to quantify the number of times words and ideas appear on the service in real-time. For example, if you were interested in public opinion of the Pandora jewelry brand for a given day on Twitter, you might find the following (where larger words are used more frequently than smaller ones):
By training models to learn to identify words and sentences related to relevant product-specific topics, your sales team can greatly enhance their ability to sell to customers, emphasizing how your product meets their unmet needs (and those of the Twitter population more generally).
As I have briefly illustrated in this piece, sales strategy can benefit enormously from data science and analytics by helping sales forces make informed, data-driven decisions that improve overall sales. There’s no need to see data science and sales as opposed, non-mixing forces; they can form a happy and profitable union.