Tomorrow’s businesses won’t thrive without AI.
In one form or another, competitive organizations are already using artificial intelligence to automate tasks, improve efficiencies, or drive more revenue. For sales and marketing teams, AI and its subfields have finally fused science into the art of selling.
Natural language processing (NLP) for example, can help determine customer intent. Machine learning (ML) can recommend the best response for sellers, and A/B testing can identify the most effective playbook for a scenario or customer demographic.
Indeed, the time for pitching AI’s business case in sales has long since passed. Instead, the moment calls for businesses to make the next move towards full transformation. The problem is: how exactly should you do it?
Below, I’ll discuss a three-step process to get it done: 1) Standardization, 2) Automation, and 3) Optimization.
Standardization is the foundation for unleashing the power of AI. If every sales rep operates on their own, AI can still add value by helping the individual, but that value is limited. The full impact of AI is realized by having it analyze the whole sales process at scale, understand the strengths and the weaknesses, pick out the techniques of the best performing reps and share them with everyone via automation and recommendations. This requires that all reps follow a standardized sales process.
One way to achieve such standardization is to embed the sales process in technology tools such as sales engagement platforms. When such tools are used on a regular basis, fact-based best practices can be determined and shared across the organization.
Some caution, however, is required when standardizing the sales process. It is important to strike the right balance between the amount of structure and the amount of freedom the reps
have. Implement standardization in a way that enhances human creativity, performance, and wellness. Otherwise, misuse of AI can stifle human potential and impede business growth.
Sellers spend much of their time doing manual, repetitive, and tedious administrative tasks such as updating CRM data and verifying customer info. These soul-sapping activities not only breed discontent, they also prevent sales professionals from doing what they do best: sell.
Artificial intelligence changes the calculus. By automating repetitive tasks, AI frees sellers to perform activities that generate higher value — such as meeting prospects or preparing unique proposals for customers who exhibit high levels of purchase readiness. The ultimate goal of automation is to completely spare sellers from repetitive tasks so they can allocate 100% of their time building relationships and having meaningful conversations with customers.
For example, extracting phone numbers from email contacts takes an awful lot of time but can augment lead generation efforts by thousands of new numbers per year. The trouble is, delegating this task to already burdened sales reps erodes their productivity. As Amplify shows, the obvious solution is to automate the process using ML to extract contact info from email signatures and alert sales reps that a new number has been discovered.
Automation improves the efficiency of human sellers by accelerating the process and reducing the effort required to complete a task. But fundamentally it does not change the sales process nor directly improves the success rates of sales activities. Doing that is the goal of the optimization stage.
Optimization has several building blocks: value metrics, reporting, A/B testing, and recommendation.
A. Value metrics. Using AI to optimize the right metric helps deliver positive business outcomes (e.g., increased revenue). However, the wrong metric — one that isn’t predictive of long-term success such as increase in sales and revenue — can undermine the process and lead to a drop in performance.
For example, “reply rate” might indicate email effectiveness but it does not accurately paint the whole picture. A high reply rate can — and often does — include many
negative comments and unsubscribe requests. “Positive reply rate” is a far better metric to track and optimize.
B. Reporting. Accurate reporting on value metrics enables sellers to determine the effectiveness of content, approaches, and sequences. This not only makes optimization possible but also provides a platform for demonstrating value to buyers. By simply counting the number of steps for example, FitBit helps people stay healthy. Similarly, tracking an actionable metric can drive dramatic improvements in the sales process.
C. A/B Testing. A framework for testing new ideas is an invaluable business tool and serves as a catalyst for innovation. A/B testing is a scientific way of validating what works and what doesn’t in different sales contexts. For example, we ran and experiment to determine whether an aggressive selling approach really trumps a socially polite one. Using an intent classification engine, we found that it really pays to be polite on and off the sales floor. The naive “reply rate” metric, however, pointed in the opposite direction, underscoring the importance of having the right value metrics.
D. Recommendation. Having full visibility into what really works and what doesn’t ultimately enables AI to give the right recommendations. Advanced solutions can recommend the best email template, sequence, or channel to use given a certain selling scenario. For example, we use NLP to make recommendations on how to handle out of office replies: auto-pause the sequence and advise reps to engage referrals found in the email.
The era of AI in sales has arrived. There is no question about it. Companies that see the potential are investing in new tools and processes. Those who have taken the three steps outlined above already see orders of magnitude in the productivity improvements of their sales teams.
Sooner or later, every company will embrace AI. The only question is — will your company be at the forefront of AI adoption, or will you wait and allow others to pass you by?
Pavel Dmitriev is a Vice President of Data Science at Outreach, where he works on enabling data-driven decision making in sales through machine learning and experimentation. He was previously a Principal Data Scientist with Microsoft’s Analysis and Experimentation team, where he worked on scaling experimentation in Bing, Skype, and Windows OS. Pavel received a Ph.D. degree in Computer Science from Cornell University in 2008, and a B.S. degree in Applied Mathematics from Moscow State University in 2002.
About the Author
Pavel Dmitriev is Vice President, Data Science at Outreach