AUTHORS

Brandon_Mills-10924_headshot_small
Brandon Mills
Manager,
ZS Associates
Jason_Bell_11099_headshot.jpg
Jason Bell
Associate Principal,
ZS Associates
John_DeSarbo_thumbnail
John DeSarbo
Principal,
ZS Associates
Kyle_Heller_thumbnail-1
Kyle Heller
Associate Principal,
ZS Associates

Latest Posts

How Data Science and Machine Learning Can Improve B-to-B Sales Organizations’ Success

Posted by Yogesh Sharma on Tue, Jul 25, 2017

shutterstock_513585007.jpg

As tech companies face increasing competition, more complex selling processes, fast-evolving technology and an abundance of data, increasing sales force effectiveness is a constant need. Unfortunately, current B-to-B selling is still stuck in the ’90s, when the salesperson was the first point of contact with customers. Today, by the time first contact is made, the prospect is already two-thirds of the way through his buying journey. However, recent advances in data science, machine learning and increased data proliferation are making it possible to improve sales gains.

For instance, ZS worked with a global telecom company to develop product recommendations for the sales managers to target their global accounts. The company had more than 40 high-tech products in unified communication, mobility, IoT, cloud infrastructure and data security, which were being sold to more than 1,400 accounts globally. The sales force was challenged with targeting the right products to the right accounts, and the efficiency of the sales process was largely driven by gut-based product targeting.

ZS developed a recommendation engine that prioritized products that sales managers targeted to respective accounts. And these recommendations often varied for the same account across different countries. We used data on products sold to these accounts in the past two years, account firmographics, industry vertical and employee strength. Using collaborative filtering, we determined the nearest neighbor for each account to identify products with high purchase probability. The revenue potential of each product was added to create a composite priority of products for each account-country combination. We created some improvements in the recommendations by using an ensemble of support vector machines, Bayesian networks and artificial neural networks.

When the telecom company applied the algorithm to actual sales, it determined that of the new sales opportunities identified, 63% came from the new algorithm. This represented 56% new revenue. 

For another high-tech hardware client, ZS improved the product recommendations that it used to direct products to its channel partners, thus optimizing the product portfolio in its indirect selling strategy. 

With increasing buyer sophistication, increasingly complex buying processes, high-tech products, and an abundance of big and real-time data, the commercial sales process has tremendous opportunities to introduce data-driven selling actions to increase the selling ROI. B-to-B companies now have the chance to replicate the success of B-to-C companies in using such sophisticated data science techniques.


RELATED CONTENT 

CASE STUDY: Shifting Sales Efforts From Art to Science

BLOG POST: Stem the Tide: How Financial Services Firms Can Prevent and Predict Churn


 

Topics: data science, B-to-B, machine learning, sales organizations, sales force effectiveness, High Tech, recommendation engine, cross-sell, collaborative filtering

Click here to subscribe to Tech Bytes & Insights
AUTHORS
Brandon_Mills-10924_headshot_small
Brandon Mills
Manager,
ZS Associates
Jason_Bell_11099_headshot.jpg
Jason Bell
Associate Principal,
ZS Associates
John_DeSarbo_thumbnail
John DeSarbo
Principal,
ZS Associates
Kyle_Heller_thumbnail-1
Kyle Heller
Associate Principal,
ZS Associates
SUBSCRIBE

Get 'TECH BYTES & INSIGHTS' Updates

Subscribe to receive email notifications whenever new blog posts are published.

×

Subscribe by Email

Search by Topic

see all