Crunch Network

The Next Wave Of Enterprise Software Powered By Machine Learning

Next Story

Smithsonian’s Giant Leap Into Kickstarter Pays Off

Enterprise software is about to undergo radical transformation — a substantial change that will make the shift to software as a service (SaaS) look like a simple facelift. This transformation is being powered by machine learning.

With machine learning, computers can process and mine data in real time to automatically discover insights and generate predictive models. Companies can find patterns and foresee what will happen in the future based on real-time analysis of their data. The possibilities fueled by machine learning are endless.

Rather than having humans enter data using a web form sitting on a database, machine learning solutions automatically mine data from both inside and outside the enterprise. This data may reside in historically unstructured systems like emails and calendars or from call centers or legacy voicemail systems. The software will not only aggregate and organize this data, it will mine this information to generate insights and predictions, which can then be used by business people to make decisions.

The Problems With Traditional Enterprise Software

Machine learning eliminates many problems with traditional enterprise software solutions. First, in traditional software, the data residing in the “system of record” is only as good as the quality of the human input. It’s a well-known reality that most sales people do a poor job of updating CRM info, resulting in weekly sales calls between sales managers and account execs and the myriad spreadsheets used to track the “actual” pipeline data.

Also, traditional enterprise systems built on relational databases are not good at presenting longitudinal views of information and, therefore, are impaired at generating patterns or insights. Hence, most corporations rely on large data warehouses that receive data dumps from enterprise systems. In this approach, business people must wait weeks, if not months, before the data teams can produce useful insights.

Lastly, traditional enterprise systems rely on thousands upon thousands of human-created and curated rules — rules that are static and, by definition, become obsolete as businesses evolve.

Why The Incumbents Won’t Win

What makes this next revolution so exciting is that it likely won’t emerge from within the existing software leaders. While the SaaS wave has largely been driven by enterprise software veterans (Marc Benioff, formerly of Oracle, Dave Duffield and Aneel Bhusri from PeopleSoft), this new wave is being pioneered by founders and teams largely outside the traditional stomping grounds.

Why? The skill sets and technical competencies required for successful machine learning software companies are different than the database, forms and workflow expertise found at traditional incumbents. Instead, the founders of this next-gen set of companies are more likely to come from Facebook, Google and Twitter than from Oracle or SAP.

The consumer Internet players have been using machine techniques for years to analyze and act on large volumes of data — dictated by the impossibility of using humans to create and update rules on a real-time basis.

Emerging Leaders In Machine Learning Apps

This new wave of enterprise software, powered by machine learning, is already finding its way into core business disciplines:

Sales: Gainsight and Clari (both BCV portfolio companies), Lattice, Insidesales.com and Gainsight are using data science and machine learning to automatically detect sales opportunities and renewals at risk and to generate more accurate sales forecasts.

The possibilities fueled by machine learning are endless.

With these solutions, sales and account managers get alerted ahead of time by the “machines” to specific customers or deals that are at risk, thus allowing management to take corrective action. Whereas legacy sales solutions generally just provide a snapshot in time, these newer companies are offering actionable, forward-looking insights, thereby driving enormous value to the enterprise.

Marketing: Captora and Persado, which are both BCV portfolio companies, are using data science to rapidly personalize and evolve content to meet the shifting needs of prospective customers. With solutions like these, traditional legacy marketing vendors simply become commodity-like delivery vehicles and repositories, while the intelligence layer provided by these vendors will increasingly occupy the strategic mindshare of the CMO.

Human Resources: Entelo, Gild and Concept Node are using machine learning models to identify and recruit talent and to make internal teams work more effectively. Legacy applicant tracking systems will become replaceable back-end systems, while the actual “work” of recruiting talent will take place in these newer solutions.

Finance: Anaplan, Adaptive Planning and Trufa are part of a new generation of finance-oriented companies that are using predictive analytics to transform the financial planning function.

Traditional finance software is all about recording every historical financial transaction. In contrast, these predictive analytics systems provide dynamic views into opportunities to drive more profit, grow faster and generate efficiencies — many of which may arise unexpectedly as business and market conditions change.

Decade-Long Opportunity Ahead

For entrepreneurs and investors, this is an exciting time to innovate and place new bets in enterprise software. BCC Research predicts the machine learning market will reach $15.3 billion by 2019, with an average annual growth rate of 19.7 percent. One of the early growth categories is predictive analytics software, which is expected to reach $6.5 billion worldwide in 2019, up from $2 billion in 2012, according to Transparency Market Research.

As we look forward, machine learning will be the defining characteristic that distinguishes “legacy” from “modern” enterprise applications. Every software category will be impacted — and the winners are all up for grabs.

Featured Image: agsandrew/Shutterstock (IMAGE HAS BEEN MODIFIED)