Back in May of this year, Google’s AlphaGo defeated the worlds best Go player in a 3 game match-up. It wasn’t so long ago that this game was considered far too complex for a computer to compete at the professional level in real time. The MiniMax strategies that work for games like Tic-Tac-Toe or Chess (with the addition of alpha-beta pruning) didn’t apply here due to the number of potential moves.
Similar to the complexities of Go, you’ll often hear Sales being excluded from the list of jobs soon to be lost to automation. As a Regional Director, I can’t say that I agree.
The claim is that teaching a computer story-telling, empathy and emotional intelligence — key skills for making a sale, are beyond the abilities of near-term AI. This assertion feels as familiar and as fallacious as the claim about Go being strictly in the human domain, before it was roundly trounced by AlphaGo.
Deep Learning, Big CRM Data & Rebuttal Lists
Around the same time AlphaGo was mopping the floor with the world’s greatest Go players, Salesforce launched a product called Einstein.
Einstein is a CRM AI that attempts to apply deep-learning, predictive analytics, natural language processing, and image processing to assist sales reps in closing deals and increasing sales revenue. By processing billions of data points, repetitions and images, this tool can help with key selling steps such as identifying new opportunities, scoring leads, and even which email marketing materials are best at lead conversion.
And this is just the beginning. Once Einstein and similar tools land in the hands of sales reps, they will be able to ingest mountains of data on how sales are conducted within an organization — reams of information overlooked by human eyes but discernible by an AI’s cold logic. Perhaps it can learn the common concerns of prospective clients and the most effective rebuttals of sales reps. Or using the historical data of similarly sized companies or sectors, it may be able to build a heuristic model and select tailored products or services that best fit the prospect. Combined with natural language and image processing, it can even create smarter pitches and learn from the client response. How long before the machine goes from assisting a sales manager, to holding their hand, and eventually supplanting them altogether?
In fact, Salesforce is banking so heavily in the future of AI that over the past three years the company has spent over $4 billion acquiring a string of machine learning startups and e-commerce developers in its quest to perfect Einstein and gain a head start in business AI. And they’re not alone. In 2012, capital raised by AI startups totaled around $589 million. Last year, it soared to $5 billion, with commerce, sales and CRM applications alone accounting for 10% of the investment.
Canary’s in the Coal Mine
Despite AI’s burgeoning potential, you might think that in today’s present reality, all of these sounds a little far-fetched, and probably a long ways off. It may well be, but even as we speak, simpler solutions are already starting to replace some segments of the sales industry.
Telemarketers are actively being replaced with robo-calling AI’s. Retail is quickly moving to smart self-service, and companies like IPsoft are taking over the customer buying journey in chat-based sales support channels. Virtual assistants like the CenturyLink-backed Angie are now being used for B2B email marketing, interpreting 99% of client responses and passing hot leads to human sales reps for action. On the B2C front, Amazon, Apple and Google each have their own AI present in millions of consumers’ pockets and homes, recording shopping patterns, analyzing preferences and generating tailored product recommendations.
Each of these represents not only a step forward in automating the sales sector in its entirety, but also an additional source of data to tackle the more difficult portions that yet remain.
Last Man Standing
A recent study by the Harvard Business Review found that 40% of sales activities can be automated at the current level of AI technology, rising up to 85% when it comes to parts selling.
According to the authors of the study, the implication is that while AI continues to develop near-human intelligence, so will sales leaders need to develop “machine intelligence” to stay on top of their field. This means identifying tasks for AI delegation, delineating which processes require human intervention, and implementing proper protocols for escalation. Gone will be the traditional executive with a penchant for slick talk but an aversion to computers — to climb the ladder, leaders will need to be as comfortable with apps and programs as they are with people.
Even sales managers themselves will need to adapt to keep pace with their AI counterparts. While relationships and soft skills will still be crucial, the successful reps of the future will be part salesperson and part data analyst: being able to connect with people and their colorful personalities, while working with virtual assistants and cold hard data. As AI dominates the middle-of-the-funnel process, reps can devote more time to promising leads and focus on closing the sale.
Thankfully, the human touch is something that can’t be fully extinguished from the sales process… yet. Until artificial intelligence becomes sufficiently advanced to include emotional intelligence, traits like empathy and people skills will still be the exclusive domain of human sales managers. In the end, the relentless march of AI technology does not necessarily mean the death of the salesman, but rather its evolution.