Sales teams are vital to most enterprise businesses, but expensive. Many companies spend tens of millions of dollars, or up to half of their budget, on sales. Salesforce, to cite one example, poured more than half of its revenue — nearly $4 billion — into its sales and marketing efforts last year.
But they’re even more expensive if they aren’t performing. Typically, only 20 percent of sales reps make 80 percent of a firm’s revenue. That means billions of dollars and untold resources are being spent in an area that’s not performing at its highest level of productivity.
The biggest problem that slows productivity: outdated tools and technology. Sales reps spend nearly 20 percent of their time inputting data into their customer relationship management (CRM) system, according to InsideSales Labs. Sales managers need these central repositories for all customer relationship and interaction data, but most salespeople hate data entry and other administrative busywork.
San Francisco startup and Inception program member SalesHero has developed a business AI assistant that uses GPU-powered AI to automate repetitive tasks that often detract from selling. In the process, it resolves the conflict with managers over the lack of data collection and CRM updates.
Meet Robin, the Business AI Assistant
CRM systems can hold mountains of useful information about customer and sales rep activity. IDC predicts that by 2021 AI-powered CRM activities will boost business revenue by $1.1 trillion. However, if that data isn’t getting collected properly, or at all, the system and the data lose value.
Robin improves efficiency and productivity by sending a daily email that includes a list of suggested tasks that it can do for the rep. Examples include adding a contact, updating account information or logging emails and calendar events. The rep simply needs to review and approve the tasks it would like Robin to perform.
To get started, users only need to connect their company-hosted emails and Salesforce accounts. There’s no IT involvement or any software to learn. Robin easily integrates into any sales process since it can work in harmony with any other tool.
Robin processes big data from users’ CRM, customer interactions and SalesHero’s proprietary database of 45 million global records to:
- improve CRM data by collecting important details previously missing or outdated,
- boost prospecting by offering up new and pertinent accounts and leads, and
- accelerate customer engagement with automated scheduling.
One SalesHero customer, a global financial institution, reports they’re rapidly automating repetitive daily tasks for their 2,500 sales reps — and seeing productivity gains of up to 20 percent. By saving sales reps hours of time each week, the company estimates Robin has boosted efficiency by $150 million.
Deep Learning in Action
Behind Robin is an event processing system that can handle billions of events and records each day. Structured and unstructured data such as CRM records or emails are continuously cleaned, analyzed and run through a series of machine learning models to automate business processes. Complex deep learning models personalize each user’s interaction and are constantly updated by a grid of powerful NVIDIA GPUs.
SalesHero’s system processes 10 million AI-powered predictions per user each month. These predictions include classifying emails, recommending new accounts, drafting email responses and much more. Robin delivers dozens of to-do suggestions to a user each day. Approved tasks are automatically performed, and the system incorporates user feedback to continuously learn and improve the machine learning models in the backend.
“We believe that in less than 10 years all information workers will have a personal AI business assistant,” said Stefan Groschupf, founder and CEO at SalesHero. “AI will continue to seep its way into every industry, boosting productivity and increasing job satisfaction.”