The demand for data is expanding exponentially. With that growth comes a broadening range of data uses, as found in an array of Artificial Intelligence, Predictive Analytics and Machine Learning (PAML), and the Internet of Things applications.
Creating and analyzing that data can lead to very powerful insights and actions. However, this work can also be very stressful on the people charged with managing it all.
A recent SAP report by Forrester Consulting highlights the opportunities and tensions found in this data explosion. It probes into the central issue of increased adoption and strained resources. In so doing, Forrester conducted a survey of 353 global decision makers who are involved with PAML and familiar with enterprise applications.
Their research found that companies struggle to keep up with demand for PAML. However, these firms are making strides to remedy those challenges through greater investment and priority on PAML capabilities.
More people to manage more data
Three key findings stand out in this report.
Rising demand and burdens: Of the respondents, 79% of companies see growing demand for machine learning models, with over 90% of companies reporting that PAML models are integral to the ongoing success of their business. However, rising demand puts a big burden on data science teams to create these models: 64% of companies surveyed said it’s hard for their internal teams to keep up with the demand for PAML models.
Struggles to embed: Companies are challenged to infuse machine learning models in their production applications. This is caused and compounded by the lack of skilled personnel (e.g., data scientists) to support demand.
Expand data analytics: New PAML solutions put analytics into the hands of non-data scientists, such as business analysts, domain experts, engineers, marketing professionals. This expansion reflects the fact that more than 80% of companies expect to increase their PAML investments over the next 12 months.
In other words, companies need more people managing more data.
Barriers are tactical, not strategic
The C-Suite sees value adopting PAML solutions to help drive growth. Yet this report shows there are tactical barriers to going full steam down the PAML path.
What’s holding them back?
Privacy, for a start. Over half of respondents (55%) see new data privacy and compliance as a barrier to adopting PAML solutions. This underscores the need for real-time data anonymization capabilities to enable compliance with regulations such as the General Data Protection Regulation (GDPR).
Lack of data talent is another barrier. Of respondents, 48% of companies struggle to deploy and manage PAML models used in production applications. Forrester says this stems from a lack of skilled personnel (e.g., data scientists) to support demand, a challenge for 42% of companies.
Take an Incremental Approach
On the flipside of these challenges are opportunities for companies to use prebuilt solutions from technology partners, enabling them to implement PAML capabilities without overly taxing their existing – or finding more – data-science expertise.
To that end, they can take incremental steps in defining and achieving their PAML goals: For example, firms can:
- seek intelligent products with embedded PAML capabilities, thus reducing burdens on data science teams.
- explore data management platforms to help manage data and queries across all cloud types (e.g., local, private, and public).
- invest in better data technology to support development teams.
Enhanced PAML = better CX
Improving automation through machine learning generates a range of benefits, not the least of which are more efficient processes and operations, more timely data usage, and adoption of AI applications – all implemented in the name of better customer experience and more revenue.
Indeed, if this were an equation, you could think of it as Enhanced PAML = Improved CX.
Over time, this equation adds up to growth.
