Be where your buyers are headed: An argument for sales automation

I think companies that win do so because they have core capabilities they understand and use well.   One of these key capabilities that I admire the most goes like this:

Be where your audience is going to be

Understand in great detail who your audience is, how to reach them on their terms and be hyper-relevant to their lives.  Always be asking yourself who your buyers really are and where they are headed. And wherever your buyers are headed, get there ahead of them and set up a lemonade stand.

This principle sounds great, but it's hard to put into practice; it's hard to make into a corporate capability.  And nowhere is this more problematic than in the sales process. 

For a large part of the buyer's journey, most sales professionals are flying blind: they have no idea how sales-ready the prospect is, what the prospect is actually interested in and what the next most valuable thing they could say to them should be.  And this is where you blow it. It does not matter if you are a market leader or disruptive new entry: your prospects like to do their own research now. And you are only going to win because they perceive and can articulate exactly how you stack up better than your competition.  You win because of your comparative advantage. 

In sales, every touch point must be valuable to the client to win that (often subjective) comparative assessment.  Customers compare alternatives using information they gather at different speeds and along different paths. If you want to be part of this process, you have to find ways to help those individual buyers take the right steps and see the right information. You have to already be where they are about to go.  If you do all of this well, you win more often because you win the comparative contest: you were either perceived as more relevant to the needs of that individual buyer or buyer team, or you manifested the clearest view as to a better use of budget over a competing project.  If you miss those moments, or if you run the wrong play or offer insight that is not immediately meaningful or simply mis-construe the sales-readiness, you lose. And losing means the sales and marketing investment leading up to that point was an expensive waste.

Good salespeople have always had good instincts that help them meet the buyer on the buyer’s terms and move together optimally through a sales process.  But scaling this is nearly impossible. Even if you could train a global sales team to be that good, the buyer's needs and behaviours evolve too fast for the program to keep up with. There is actually plenty of evidence to suggest that sales efficiency across B2B is static or more likely in decline in the last 5 years. I think this is a symptom that sales (as a process) is in a state of failure. 

A case for automation

In addition to meeting the needs of the buyers, sales performance lies in the ability to be consistent, continuously improving and to do so at scale.  This is the kind of problem that begs for a comprehensive sales automation solution.  Yet the prevailing models for automating business processes like sales have been pretty static for some time now.   Evolution of the sales and marketing tech-stacks heavily favoured automating only those tasks that would have been otherwise repetitive or labour intensive. The support for workflows requiring comprehensive decision-making remains impractically thin.  To be clear, the traditional tools provide huge value in terms of speed, reduction in effort and reach. However, these tools all provide value from essentially the same paradigm that they did 20 years ago. They are systems-of-record feeding applications that are force-multipliers - e.g., instead of connecting with one person at a time, you can connect with thousands; instead of updating customer information in 210 places, you now only have to do it in one; instead of waiting days or weeks to see how things are going, you can generate real time reports in seconds, etc.  These are great, and you can't live without them, and the success of the companies providing the best-in-class tools here is evidence of that.  But none of these force-multipliers do much to add value for the audiences; in particular they don't really help the seller be where the individual buyers are trying to go.   

 A big vision for sales automation is exciting but also something relatively few companies are capable of delivering to enterprises at scale. The tech-stack would have to evolve from today’s CRM-centric model to something that draws intelligence from buyer's prior choices and consumption of content as well as a host of other signals. The system would then need to make predictions and actively suggest and deliver next steps, hyper-relevant content and offers while getting smarter as it goes. Contextually aware machine-learning approaches can drive recommendations capable of predicting next-best steps in a sales process along these lines. But practically integrating this kind of power into a fully automated sales process that encompasses both human and digital interactions fluidly is not quite a reality yet.

So why don’t we have it yet?

One of the reasons we might be a bit stuck on this today is because the current recommendation schemes and approaches don't do a good job supporting this class of decision making.   We think we have recommender systems all nailed down, but we don’t. Despite what many people think, the intelligence and structured data needed to automate sales is considerably different than what we use today to personalize social ads or music recommendations on streaming services, etc.  For one thing, if you want to really know what buyers are likely to be interested in, you have to comprehensively interrogate content they have previously spent time with. And it’s not just enough to know they clicked on a presentation that was tagged as being about accounting and software. You have to have a reliable way to extract a far more granular topic-model from everything they spend actual time with.  It doesn’t end there. You also need elegant solutions for acquiring the data, structuring it, and dealing with its dimensionality and context in the machine learning. There are issues with visitor identification and the decision waterfall where data is sparse. Finally, producing and delivering a useful prediction and activating it wherever it needs to be in real-time presents a range of latency and integration challenges.

I'm pretty interested in this problem and have been for a long time.  The way I see it, there remains an unfathomable level of wasted time and effort on both sides of the B2B selling and buying process.  Solving for this is the future of how B2B products and considered purchases will come to market. This certainly does not mean that sales and marketing professionals will be replaced by AI; quite the contrary. Human creativity, understanding, empathy and high-touch communication are paramount to the evolution of sales and buying process. Automation now needs to do more to enable both the buyer and the seller to get on the same page so the human stuff can become a bigger part of why companies win.

Something I am genuinely proud of

I'm proud to see that PathFactory - a company I co-founded and built  with my good friend Nick Edouard (now being run by the amazing Dev Ganesan) - has taken a big bite out this problem by solving for the tricky data-structuring, ML modelling and contextualization issues in a way that works in practice.  It's a worthy pursuit because I think it serves both the buyers and the companies that sell to them.  Helping buyers make the best possible comparative choices between options with the least amount of friction is good.  Ensuring the companies with the best comparative advantage win those deals, rewards both innovation and good corporate citizenship.  I think sales will become automated to produce the outcomes I describe here and I hope that the years of work the team at PathFactory put into our end of the solution will help make it real.

A waitress welcomes new guests at the picturesque teashop along the Tokaido road from Edo to Kyoto. Tokaido gojusan-tsugi no uchi 東海道五拾三次之内 (Fifty-Three Stations of the Tokaido Highway), Utagawa Hiroshige (歌川広重) 1833