Machine Learning Value Propositions for the Midmarket

Investments in Machine Learning (ML) and Artificial Intelligence (AI) are nowadays surging, as enterprises attempt to leverage these technologies in order to improve their competitiveness. However, a closer look at the market shows that the vast majority of these investments concern large enterprises and high-tech startups. In particular, large corporations take advantage of their size and wealth in order to deploy data-driven intelligence and stay ahead of their competitors. On the other hand, many startups are producing disruptive AI and ML related innovations, which makes them very attractive to investors like venture capitalists and innovation funds. Contrary to large enterprises and startups, companies in the midmarket seem to be more reluctant to adopt ML and AI. This is however bound to change, as most midmarket enterprises that have invested in ML are already reporting very positive and very promising results. The latter concern improvements in internal business processes, as well as the creation and roll out of novel products and services. Moreover, feedback from early adopters shows that ML can become an innovation enabler for the Midmarket in sectors like retail and finance.

Machine Learning Opportunities for the Midmarket

Here are some of the most representative ways in which Midmarket enterprises can benefit from Machine Learning:

• Gaining Scalability Benefits: Midmarket enterprises have in most cases hard time competing with larger enterprises, given that large corporations have various scale-related advantages. For example, Midmarket companies can hardly achieve the economies of scale that are offered by bigger enterprises. Machine Learning can help them offset the scale advantages of large corporations, through processing very large amounts of unstructured and heterogenous data in very efficient and cost-effective ways.
• Create an Innovation-Related Competitive Advantage: Machine learning is an innovation vehicle for Midmarket companies. As an example, in the retail sector, machine learning applications enable intelligent product recommendations and dynamic pricing predictions, which Midmarket companies can use to provide value both to their suppliers and to their customers. In general, Midmarket enterprises can nowadays leverage machine learning in order to gain a competitive advantage against their competitors.
• Take advantage of Sensors and other Data Sources: Several Midmarket companies have access to data from sensors, internet connected devices and other sources like social media. However, without machine learning these data remain unexploited. Machine Learning can provide Midmarket enterprises with opportunities to exploit these data in tangible business cases that can generate new revenue streams.
• Provide high quality Professional Services: As already outlined, a very large number of enterprises of all sizes are currently investing in machine learning and AI. In most cases these enterprises need support in their machine learning endeavors, as they typically lack the knowledge and skills required in order to deploy and fully leverage state of the art machine learning solutions. This provides a great opportunity for Midmarket companies to offer professional services around machine learning, including training, consulting and business support services. MidMarket enterprises can therefore excel in providing high quality assurance and advisory opportunities around machine learning platforms and applications.
• Get Rid of Manual Error Prone Operations and Boost Automation: Machine learning can help Midmarket companies improve their productivity through automating business processes and reducing human mediated, error prone steps. Hence, the Midmarket can exploit ML as a vehicle for automating both internal and external business processes.

Value Creation Guidelines

With so many exciting opportunities at hand, Midmarket enterprises should get prepared to pursue them. In this direction, they had better take into account the following best practices and guidelines:

• Seek Opportunities in all Functional Areas of an Enterprise: Machine learning is not confined to a single functional area of an enterprise. In the near future most business processes will be data-driven, including business processes in the areas of production, marketing, finance, accounting, human resources and more. Therefore, Midmarket enterprises should look for opportunities in all these functional areas, prior to deciding their area of focus i.e. the area where they will develop their core competency.
• Employ Machine Learning Experts: There is a proclaimed talent gap in machine learning expertise, which is one of the main inhibiting factors for companies that wish to be become active in the ML space. It’s therefore essential for all enterprises to on board machine learning experts that will help them start the development and the offering of their machine learning services on the right foot. Machine learning experts will also help them educate other employees, but also in developing a data-driven culture in the company.
• Educate Existing Employees: In addition to hiring machine learning experts, it’s important for Midmarket companies to educate and upskill existing employees. Given the skills gap in machine learning, this is a key prerequisite for creating a competent team that can deliver efficiently internal and external machine learning projects. The training investment can be significant, but it shall produce a great asset for the company i.e. employees that can leverage machine learning.
• Start Small and Focus on Customer Needs: Machine learning is a cool and trendy technology, but it should be used in a way that improves business results. Midmarket companies should avoid technology driven approaches that develop products for the sake of technology. Rather they should be driven by customer problems and how machine learning could address them. Furthermore, as machine learning technologies are associated with implementation risks, solutions should be developed based on small solid steps without however ignoring the big picture. Implementation can be gradually scaled up as more data become available and as machine learning models are fine tuned.
• Create a Data Management Infrastructure: Much as models are important in machine learning, the most important asset is probably the datasets that are used to train the models. Therefore, Midmarket companies entering the machine learning market should have a strategic plan for the collection and management of very large amounts of data. The data collection effort should be continuous as more data can enable improved training of models and ultimately better outcomes. In this direction, Midmarket companies should have in mind that the amount and the quality of data that they possess could set them apart from competitors.

In the next couple of years, it’s likely to see an increased number of Midmarket companies entering the machine learning game. There are certainly plenty of opportunities for them. However, these opportunities will come along with very stiff competition from other enterprises, including both large corporation and Small Medium Businesses (SMBs). In order to sustain this competition, Midmarket enterprises will have to excel in terms of data assets quality, business expertise and technical competencies. While this is a very hard and very demanding task, following the above listed guidelines could provide an essential boost in confronting the challenges.