Why Networked Intelligence Wins Every Time
Do you remember a time when purchasing a product online meant waiting for days or weeks for it to arrive? Before Amazon figured out fluid delivery of digital purchases, retailers were driving customers to in person shopping if they wanted products in hand quickly.
But the Amazon Effect has had implications far beyond retail — helping companies large and small anticipate customer decisions and automate tasks that typically require human intelligence. Companies like Facebook, Google, and Netflix have been using neural networks and other machine learning approaches to not just speed up their business processes and forecasting approaches, but to change ways of working. Their success and increased focus on continual tech optimization have made artificial intelligence an increasingly common buzzword in tech and business headlines. And when it comes to sales forecasting, the consistent use of AI can have a real impact on businesses’ bottom line.
But many sales and less technical fields often use less advanced AI (or AI-lite) — focusing AI integrations on automation versus true machine learning and collective intelligence.
What’s the difference between AI-lite and true networked intelligence?
Amazon was using something called “anticipatory shipping” all the way back in 2013, helping the retail giant get orders as close as possible to customers before they actually purchased them.
How are they able to anticipate what consumers will order before they add items to their cart? Because they are pulling from multiple data sets to make probabilistic predictions of what will be purchased and moving products to last mile warehouses before purchase are even placed.
This was revolutionary at the time, but more and more companies have taken on the approach. And though sales teams have been slower to move towards automation than supply chain and marketing teams, today 37% of sales teams report using artificial intelligence in their day-to-day work.
That’s a 76% increase since 2018, according to Salesforce. But often the technology is being used as a way to automate the documentation and sharing of internal information, not to better understand customer and market trends.
AI-enabled sales technology helps many teams capture, share, and find historical sales activities in their CRM, which provides more transparency and gives sales reps precious time back in their day, but doesn’t realize the full capabilities of artificial intelligence and machine learning.
Transparency and historical data is a huge step forward for sales. But it’s not enough. That’s why Collective[i] adds market intelligence to its Intelligent Forecast tool.
Historic data can’t predict future outcomes
Customers today — both consumers and B2B audiences — have increasingly high expectations. They’re doing most of their research online. They expect seamless, personalized brand experiences, from start to finish. Capturing and synthesizing internal historic actions to predict future success simply won’t work — especially when it comes to complex, B2B sales needs.
In a typical firm with 100 to 500 employees, Gartner estimates “the average buying group now includes 11 active members and up to seven occasional participants,” with 77% of B2B buyers saying their last purchase was very complex or difficult.
Enterprise sales teams are struggling to keep up. Sales managers and revenue operations teams spend a lot of time tracking down selling data, while sales professionals are often wasting precious time on data entry and other administrative tasks (while only 25% a rep’s time is spent actually selling, according to Sirius Perspectives) when they could be out selling and researching buyers.
Training sales professionals to optimize their actions to past sales success is going to miss present reality and future, shifting trends. During the pandemic, many companies relying on historic and human forecast predictions, made forecasts based on previous pandemic factors. They were forced into layoffs based on assumptions and responses to prior crises. Many of those same companies missed opportunities as a result and were left scrambling to rehire and retrain when revenue stabilized.
Automation and augmentation of existing processes can look like digital transformation. The reality is that automating a process built on replicating past performance or training a machine to do it faster will continue to make companies miss the mark — even if they get there faster.
While opening communications between sales teams and other internal departments like marketing and legal is a huge bonus for companies that have not had that sort of transparency in the past, it’s only a small piece of the puzzle.
The reality is that Yogi Bera was right: “It’s tough to make predictions, especially about the future.”
Why an accurate forecast isn’t enough
Accurate sales data has been elusive for so long that the prospect of accurate info in CRM has become something of a holy grail for sales operations leaders. One CRO we spoke with said that he’d kill for accurate CRM data: “I feel like Christopher Columbus trying to find the right info from my teams.”
That’s why the traditional measure of a good forecast has been accuracy — specifically how close the prediction came to the outcome. The traditional approach is now being sped up and automated by many players in the forecasting space. While automation is an improvement (especially when it comes to team productivity), flaws in this approach remain.
Because accuracy is a byproduct of real AI, not the goal.
A few players in the sales technology space capture sales activities, connecting to email, appointment calendars, invoices, video conferences, and recorded phone conversations to collect and synthesize historic data to help forecasts achieve better accuracy.
But that approach implies that by better understanding past behaviors, future sales can be predicted.
“The absolute outcome of a sales pursuit is not knowable in advance nor is it exclusively in the control of a sales professional, process or organization,” says Collective[i] cofounder Stephen Messer.
The power of networked data
Networked intelligence is — simply put — a kind of enterprise-level crowdsourcing.
Neural networks are a subset of machine learning at the heart of deep learning algorithms. They identify underlying relationships in data sets through a process that mimics the way the human brain operates, allowing computer programs to recognize patterns and solve common problems, training data to learn and improve accuracy over time.
Collective[i] uses its neural network to not just to automate and accelerate mundane tasks for clients, but to understand individual buyer behavior and market trends, offering up real-time insights and recommendations to sellers about next and best actions to take in their selling process. As new information enters the system, the recommendations get smarter, giving teams faster and more accurate predictions — effectively super powering their forecasting processes.
That’s one reason the company was recently named Data Solution of the Year by Data Breakthrough. “Collective[i]'s difference is that it goes beyond the confines of an individual company’s sales history,” writes ZDNET. “And by doing so, captures a fuller picture of the buyers who are being targeted, such as what other suppliers they work with, and what other products they have bought or are about to buy.”
When you accept that the future is dynamic, forecasting becomes a yardstick for the impact of activities (conversations, emails, meetings, sales skill, etc.) — a scorecard, not the goal.
In other industries, adapting to new data sources as they arise is considered a bonus, not a weakness. The shipping industry uses weather forecasting to adapt travel routes that change as new information emerges, while stock traders make educated guesses on market trends, and readily adapt when market volatility erupts.
Basing forecasts off the prior accuracy of commits implies that the future is fixed and knowable. It leaves businesses open to operating reactively and at risk of missing threats and opportunities that could make or break quarterly numbers.
“The traditional way of working is process driven (focused on creating consistent approaches for every buyer) and activity based (more is more),” says Messer. “But buyers now expect sellers to know their preferences. A cookie-cutter approach to sales just won’t cut it. Not to mention that enforcing consistency often comes at the expense of productivity.”
Modern B2B buyers expect their sales professionals not to simply broadcast sales soundbites, but become trusted advisors. And modern sales organizations realize that if they do not automate the busy work as much as possible, the seller will never have the time to provide that value to their clients. Automation of sales forecasting, logging of activity, and digital networking means the seller is spending time understanding problems, proposing solutions, and building relationships.
Gartner research shows that 80% of the sellers who used a sense-making approach closed high-quality, low-regret deals. True artificial intelligence and machine learning entered into the sales process enable sales teams to step away from data entry and focus on work that matters: learning about buyers, understanding trends in the market, and strategizing solutions they can help with.
That’s why Collective[i] technology focuses on automating the tasks that can be done by computers so that sellers can focus on strategy and personal outreach.
“The objective isn’t to replace humans,” says Messer. “Automation is more about moving up the value chain, getting smarter and setting sellers up for success.”Explore Collective[i]