Sales remains the last business area essentially unchanged by technology. Most companies have viewed investments in CRM as their answer, but CRM has done little to make salespeople more effective. CRM can even impede effectiveness by taking time to update applications that often hold incomplete and inaccurate data.
Marketing has changed fundamentally with the availability of rich consumer data, algorithms to segment and target consumers, and digital campaigns that intelligently interact with consumers. At Experian, we referred to the shift from “Mad Men to math men”. The phrase may not reflect gender parity (the team’s leadership was mostly women), but accurately captures how marketing has shifted from creative content to applying quantitative techniques to optimize digital marketing investments. No similar shift has occurred yet with sales.
My experience with sales analytics began in the Global Equities business at Citi. We analyzed “low touch” clients who sent order flow electronically while rarely interacting with sales traders or research analysts. We found that a small subset of these clients, with specific characteristics and transaction patterns, would increase their flow significantly with “high touch” service.
Subsequently, I led sales technology for Barclays Capital, where the merger with Lehman Brothers gave an opportunity to re-engineer sales workflows. We created a trusted and comprehensive data lake covering all transactions and exposures across all asset classes, then built a rich business intelligence platform that explained what had happened. Using this base, we applied somewhat basic classification and prediction models to optimize what could happen. We embedded these analytics transparently into how salespeople worked. For example, a key element of the business case for the merger was achieving cross sell. When sales traders call clients now, the VOIP system determines who they are calling and a window pops up on their desktop proposing the top cross sell opportunities.
Later, I led product and technology for Experian Marketing Services, providing data, predictive analytics, and cross channel campaign execution to sophisticated airlines, banks, ecommerce businesses, political parties, and retailers. This experience showed the most effective analytics to generate or retain concrete sales included:
- Lead identification and scoring
- Cross sell and up sell
- Churn or defection
Sales leaders have different needs. The most practical and highest ROI uses include:
- Smart pipeline prediction
- Win/loss drivers, often through visualization
- Sales person behaviors analysis
Analytics are most effective when embedded seamlessly within how people work. Salespeople can be slower relative to other business functions to adopt a data-driven culture, so analytics often work best within an “invisible UX”. For example, leads can go to salespeople via a text. When an app on a smartphone senses that the salesperson is calling on a client, the CRM system can be automatically updated. After a client meeting, a salesperson can get a reminder text asking them questions about the call. Each step is simple and proactively involves the salesperson.
Analytics can trigger a series of simple transactions that cumulatively form complex workflows. For example, I’ve advised an office furniture manufacturer that uses a sophisticated sales performance model emphasizing building trusted advisory relationships with facilities managers for large companies. However, the sales performance model is not always immediately clear to junior salespeople at dealers, who often sell to smaller clients. Using “invisible UX” allows applying a relevant subset of the performance model to dealer salespeople, e.g. the system can text about salespeople about expiring office leases, prompt them to look into the depreciation status of existing furniture, and then follow up asking whether they’ve secured support from both solution (facilities) and financial influencers.
Looking forward, we can expect artificial intelligence to complement and progressively replace many sales functions. Currently, chatbots are beginning to perform customer service functions. The trend will extend to replace much inside sales with synthetic intelligence. Next, salespeople will have AI avatars that complement client interaction. As deep learning and reinforcement learning evolve and combine with natural language generation, we can expect limited forms of general intelligence that enable wholly autonomous sales via artificial intelligence.