Using machine learning to increase sales velocity

Case: Transfluent

 

View PDF

 

About the company

Transfluent is a Finnish technology company that offers an almost real-time service platform specializing in language translation. The company was founded in 2012, and in addition to its head office in Finland, has a subsidiary in California. See more. 

 

Challenge

Transfluent's sales team manages over a thousand sales opportunities annually, and the sales team is expected to sign a certain number of new customers each month. As a major part of the sales is new customer acquisition, making the understanding of which cases to focus on crucial for their success. Therefore, data-driven support for focusing on the right sales cases has high value for the company.

The project aimed to bring clarity on which sales cases to focus on.

 

Our solution

Transfluent's marketing and sales activities generate lots of customer acquisition data which can be used to improve the sales team’s performance.

Since the company was using HubSpot Marketing Hub there was good traceability through the pipeline, all the way from first clicks on the website and content to sales qualified lead and won customers. This customer footprint and other data points, such as company information and sales engagement, were used to calculate deal scores. A deal score represents each sales case's likelihood to buy and can be used for prioritizing sales opportunities in the pipeline.

Scoring sales cases requires an organization-specific Machine Learning model. The algorithms in the model were trained with existing sales and marketing data.

  • The first step was to collect the data and prepare it for training the scoring model. Based on this, the model selects the most significant variables for scoring.
  • After the deal scoring model was validated, existing sales and marketing tech solutions were integrated for automating the deal scoring.

In production use, required data points – variables – are pulled from the source systems on daily basis and the model is run. The Deal Scores are then returned to CRM and used for sales case prioritization and nurturing.

Outcome

The predictive deal Scoring solution helped Transfluent's sales team to prioritize their activities and to focus on the most valuable deals in the pipeline. 

Prediction accuracy has varied between 75 % - 95% which means that on average more than four sales cases out of five have been scored correctly by our automated system. 

By using a highly accurate ranking system, Transfluent's sales team has been able to focus on the winning cases and has greatly improved sales velocity.

InlineMarket's Predictive Scoring

Predictive Scoring is a shared sales and marketing methodology for ranking and prioritizing leads and sales cases in the pipeline on a day-to-day basis.

Predictive scoring is based on the interest your leads show in your company, the current stage of the deal at your pipeline as well as several company attributes, such as the company size and industry. See more. 

 

Firmographics

 

Interested to hear more about this project and predictive scoring? Schedule a meeting with Ari. 

 

About InlineMarket

 

Our mission is to increase our customers’ data-driven businesses with a strong focus on technology. We help sales and marketing professionals to be more successful by using Data & AI. Our platform benefits machine learning for automation and higher business performance. Read more about us. 

 

microsoft partner hubspot partner