There’s a question we’re hearing more and more often: “How is artificial intelligence transforming pricing, and what can we expect in the future?” But it’s really a question of how soon changes caused by AI will arrive, because their seeds have already been sown.
As you’ll see, AI will adopt various guises in the not-too-distant future as it impacts pricing and negotiation at multiple points, in multiple ways, along the commercial sales journey. The one thing to understand and accept? How those changes are very much inevitable but will be profoundly positive.
Price is a measure of value
One of the definitions of ‘price’ is ‘a measure of value (perceived/real, communicated/delivered)’. Value, as multi-dimensional concept, is captured in a single number: price.
Important value drivers include the product or service offered, but equally the context of the transaction. What are delivery and payment terms, any rebates agreements in place, how is the product offered used: as critical component in the buyers end product or a one-off low risk purchase, what are ‘next best alternatives’? Just to name a few. Hence the critical importance of segmentation, awareness of the dynamics of the commercial environment, the ability to process new information efficiently, and data quality.
Awareness Versus Predictions
AI can help to identify the most important drivers (‘features’) that determine commercial success. Or better: that have defined commercial success.
Not a Crystal Ball: AI can provide predictions inferred from patterns in historical data, but AI is not a silver bullet nor a crystal ball. New patterns, such as COVID-19 supply chain disruptions are unique for most of us humans and AI alike.
Augmented Intelligence: Whilst AI can help predicting outcomes based on historical detected patterns, dealing with partial information and new events and trends will for some time remain the remit of true intelligent problem solving, still unique to humans, who can apply thinking patterns and solutions from one domain to another.
Especially in this phase of emerging AI and incomplete or low-quality data (more about that later), successful outcomes in commercial processes are best secured by a hybrid, a.k.a. Centaur, approach. As AI is exposed to more patterns, which is a matter of time and reliable data, the balance will shift toward more machine and less human.
Accelerating Technology Trends
AI, and machine learning in particular, is evolving at high speed. I want to highlight a few aspects of this.
Our imagination, not the sky, is the limit. Once the domain of highly specialized data scientists and high-performance (and expensive hardware), AI is now increasingly democratized, with an ever-growing collection of libraries and platforms that have – and will continue to – lower the entry barrier for applying affordable and value-adding AI/ML to any business process.
Conversational AI, leveraging computational linguistics, is getting better at exponential speed. Speech-to-text, text-to-meaning and sentiment, have become mainstream solution components which application scope covers the major languages. Documentation and reporting were always an activity that sales like to minimize if not avoid. Unfortunate as sales are an important antenna in the marketplace. Conversational AI can summarize voice memos, chat and email messages in market/client knowledge.
Just like Bloomberg writes analyst reports or Wimbledon tennis matches are summarized by IBM Watson’s AI using sensor data from rackets and shoes and video analysis, we can fast forward a little and think of how negotiations will be conducted as a 4-way conversation: Both parties will be supported by their own smart devices that listen to the conversation and provide relevant suggestions that augment the quality of negotiation. Conversational AI has a great potential to augment segmentation models with dynamic signals that help in describing the commercial context and determining the winning offer. A win-win.
Knowledge systems will be the backbone of smart devices and will help organizations and individuals to cope with information overload and bring out relevant detail when and where it matters.
Knowledge systems using graph models and supporting technology have seen a rapid surge in the last 5 years. These systems could help answering product application questions, detect critical knowledge gaps from unanswered questions. Think how relevant signals can be processed and converted in relevant information. Knowledge systems will help with building situational awareness and apply the relevant information in a commercial process.
With Michael Porter’s 5 Forces model in mind, think next best alternatives, competitive context, and upstream and downstream supply chain impacts of the value chain under consideration. Combined with conversational AI, new and powerful applications emerge, such as interpreting industry reports and distilling relevant trend information; crowdsourcing the selling success of products, and competitive intelligence from deal justifications drafted by sales reps for low-priced deal approval requests.
Knowledge systems will enable with fast market entry and deploying a new sales team in a new sales area and at the same time make organization less dependent on fact-based knowledge that is only the heads of people.
Recommendation systems use knowledge systems and can suggest products the buyer might be interested in, based on simple or sophisticated definitions of similarity. A small step further is providing options by suggesting alternative offers with different products and T&Cs the seller or buyer can choose from.
Digital commerce will continue to augment and replace traditional, human supported sales processes. Both seller and buyer need timely and adequate information that helps making the buying decision. Information that is increasingly provided by smart devices and systems. Depending on the risk and impact (value) of the buying decision, we can expect more and more traditional sales processes being fully or near fully replaced by digital commerce, being faster, convenient and probably more satisfactory. We could say: sales evolution Darwin style. Evolve & adapt or extinct.
Explainable AI has and will become a must-have capability. Like any important decision, the motivation, objective, and process need to be auditable. Gone are the days when a data scientist can deploy a magical algorithm that’s a “black box” with limited or no information about how it was trained and what rules (hyperparameters) it uses. Explainability reduces the risk of misalignment between model users and model builders and helps managing liability while promoting adoption. If we humans understand the process and the data, we are likely to accept the outcome. Think how car navigation apps with real-time traffic information provide evidence to back up the suggestion to take an unusual exit, displaying the reason (a car accident) and alternatives (time and distance options).
Data is the new oil
AI pricing solutions are accelerating the speed whereby buyers and sellers can adopt to changes that influence the context of the sales transaction. With increasing speed, the tolerance for low data quality and slow manual data quality improvement processes reduces exponentially. Would you trust your life to an autonomous vehicle with a faulty proximity radar driving 60 MPH in congested traffic?
Many large enterprises seem to have just woken up to the reality of their true data quality and convoluted manual data management processes that allowed them to survive in a world with relatively low dynamics. Developing and feeding smart models with incomplete, low, or unknown data quality will yield very predictably poor outcomes. The investment in adequate data quality and external data antennas is high, overdue, and unavoidable.
The future is now. AI is often over-hyped by marketing, which might result in a third “AI winter” due to costly damages to reputation and financial success of enterprises. At the same time, we see AI included in – and sometimes fully taking over – many everyday tasks. Mostly gradually, almost by stealth, though sometimes more disruptively. Think about how most of us ‘obey’ the car navigator’s instructions and accept the ETA without challenge, driving without knowing what route. It’s a system that delivers value by dynamically reacting to changing context, providing information on deviations and abnormal events.
Commercial processes, on the other hand, can be significantly more complex, which calls for selective adoption of AI in the form of augmented intelligence, where AI pricing solutions assist the people involved. It can help sellers and buyers in developing options, assessing risks, optimizing outcomes, and alleviating administrative burdens of reporting and informing others.
As the complexity and risk threshold continues to decrease, more and more traditional buyer-seller processes will be replaced with smart systems. Not only because the 24/7 availability or cost, but increasingly because of the superior value-added execution and outcomes. This is digital transformation: Sometimes evolutionary, more often disruptive in nature. Time to prepare for change!