September 16 by Conga
As I shared in my last blog, machine learning is not a new concept. However, in the past decade we have seen an exponential uptick in sales intelligence capabilities and achievements. This is driven by the economics of cloud computing, the expansion of cloud storage, the proliferation of sensors driving Internet of Things (IoT) connected devices, and the pervasive use of mobile devices that consume gigabytes of data in minutes. While the drivers for machine learning are clear, applying sales intelligence and machine learning to maximize business outcomes – whether tinkering with a machine learning library and platform, or using “opaque” approaches – remains a challenge.
Sales Intelligence Maturity: Harnessing machine learning to drive business outcomes
Historically, analytics in the enterprise was centered on reports and dashboards. These tools, created by business analysts to serve managers, were used to measure past performance. Their ability to improve business outcomes was limited by the amount of time (weeks or longer) required to create the report, the relatively narrow audience with access to the information, and the lack of value added insight derived from the data – i.e., no predictive or prescriptive intelligence is provided to decision makers.
The Apttus Intelligence Capability PyramidTM summarizes the levels that enterprises need to scale to gain greater insights and to drive more profitable business outcomes
First let me go over some definition:
- Descriptive analytics. Descriptive analytics is the foundation level that helps users understand what has already occurred. This is done by laying out relevant summaries and supporting data into formats that are easy consumable, both by end-user staff and management. Analytics at this level is a “rearview” exercise.
- Predictive analytics. Predictive analytics helps users recognize patterns and detect meaningful trends. More significantly, you’ll be able to generate projections of different developments for different time horizons based on the output of the analysis.
- Prescriptive analytics. Prescriptive analytics on the other hand delivers granular insights and forecasts showcasing what is likely to occur, accompanied by relevant, system-driven recommendations on next best actions and tactics to adopt.
- Cognitive analytics (machine driven). Cognitive analytics exploits machine learning to refine trend and pattern analyses in an on-going, unsupervised basis in order to constantly evaluate processes and associate data. It also leads to automatically initiating specific, suitable policies, actions and workflows.
Today, the vast majority of enterprises have needs for descriptive analytics, which are necessary for effective management, but not sufficient to accelerate business performance. In order to scale to a higher level of responsiveness, enterprise organizations need to move beyond descriptive analytics and climb up what we call the Apttus Intelligence Capability PyramidTM. Think of this pyramid as a capability maturity model – the higher up you go, the more refined and sophisticated your organization is in using machine learning to drive better business outcomes.
I would argue that prescriptive analytics, when feasible, is the minimum threshold for applied machine learning in the B2B enterprise as insights are accompanied by recommendations at the point of decision making. This is critical because most users are not subject matter experts or data scientists. We need to codify the collective enterprise expertise and make the recommendations relevant, trusted, seamless to the user, accessible in context, and timely. However, as the old saying goes, “you can lead a horse to water, but you can’t make it drink”.
Making Your Horse Drink with Sales Intelligence
Deploying sales intelligence into a complex, enterprise-wide process like Quote-to-Cash begins with domain expertise, followed by time for the machine to prove its value and trustworthiness. Without Quote-to-Cash domain expertise, deploying machine learning through a platform will not yield optimal results or produce the well thought out user experience for continued usage (adoption is one thing, but continued usage is a more telling metric). Quote-to-Cash affects users across the gamut, including sales, sales ops, marketing, legal, finance, operations, and IT. If any one of these users decide that the prescriptive analytics are not the proper action to take at the proper time, he or she will ignore the guidance provided by the system. Even as the model is corrected and optimized and continues to improve, the user will hold doubts, defeating the purpose of integrating machine learning into the Quote-to-Cash process.
Some systems use an “opaque” box approach – one in which the deployment team cannot look under the hood to see how things like price optimization are calculated and prescribed. While some may see value in keeping that information confidential, it poses a challenge: that if you can’t see the model, how can you judge its accuracy? In today’s open source world with freely available machine learning algorithms, maintaining proprietary models is legacy thinking. Contemporary thinking argues toward having a transparent box approach with prescriptive insights guiding the users.
Trust is earned through proven, repeatable results. And while at Apttus we are extremely proud of our pre-built algorithms that come as part of our Quote-to-Cash product, we would never presume that your data scientist wouldn’t be able to build a better model more suitable to your business. Our strength is in delivering a platform that automates the creation of intelligence, into which your models and algorithms can be applied. Apttus will utilize them to seamlessly present Quote-to-Cash recommendations to your users all across the organization so that these insights are trusted and continuously used.
Apttus has applied machine learning today
At Apttus we have made it easy to get started using machine learning today. Since we’re the leader in Quote-to-Cash software, with over a decade of domain expertise, we are uniquely positioned to help you apply machine learning to the area of your business that will have the greatest impact – winning deals.
You can apply machine learning for:
- Quote scoring that predicts the probability of winning the deal, while also giving sales reps recommendations to improve the quote and increase the win probability. By learning from historical quote data, quote scoring helps all sales reps sell like the top sales reps.
- Cross-sell and up-sell recommendations for products and services that customers are most likely to purchase based on analysis of past purchases and similar customers. Cross-sell and up-sell recommendations help sales reps maximize deal size and increase account penetration with machine learning insights.
Pricing intelligence produces optimal price or discount levels for each deal based on machine learning analysis of the deal characteristics and sales history. Pricing intelligence helps sales reps and sales operations deal desks through the quoting and negotiation process with guidance for initial, target, and walk-away price levels for every deal.
This is not vaporware, some “opaque” box approach, or a machine learning library. Intelligent Quote-to-Cash with machine learning packages the data science and the machine learning in an intuitive, easy to use interface that is seamless to the user while providing prescriptive analytics at the point of a user decision.