Procurement crossed an evolutionary milestone in digital transformation by embracing cognitive technologies such as artificial intelligence, robotic process automation, and natural language processing (NLP). A McKinsey study found that 60% of the hundreds of tasks from source to pay can be largely or completely automated. It is noteworthy that such tasks with automation potential included both routine transactional as well as strategic tasks. From order and invoice processing to vendor selection and management, almost any task held the potential to be automated – including email.
Email Still a Popular Procurement Tool…
Procurement depends to a vast extent on emails to communicate with suppliers, float and receive tenders, evaluate deals, complete compliance requirements while evaluating deals, negotiate, draft, and work on agreements, sign contracts, and meet certification requirements. As many as 80% of procurement organizations used emails to place orders with suppliers in 2017. According to Level Research, in 2018, 53% of US organizations used emails for ordering supplies.
…Despite the Demerits
Email as a procurement tool is fraught with challenges, causing enterprises to contend with:
Lack of visibility into organization spend, compliance issues, inefficiencies, and loss of cost saving
A massive volume of mail inflow – in unstructured and non-standardized formats – received daily to conduct business
Valuable time and effort spent monitoring for spamming and phishing
Manual indexing to segregate work and identify action increasing response time
Missed alerts and lack of dashboards to track SLA breach that delay decision making
It is not surprising that there is an increasing demand for intelligent email classification for faster and better delivery of business services.
Cognitive technology enabled smart email communication can reap multiple advantages for enterprises.
By leveraging Robotic Process Automation to build bots to constantly monitor incoming emails and trigger an action, enterprises can manage high volumes of incoming email requests. Similarly, attachments can be processed using optical character recognition or computer vision software depending on whether they are contract documents or product images. Intent classifier algorithms such as multi-class neural networks or support vector machine subsequently enable identification of the type of procurement requests for further prioritization and categorization when they call for human intervention.
Through NLP, supplemented with the strength of procurement domain ontology, machines can process the natural language within the email body to understand the intent and extract corresponding values, which can be used to fulfill the request. For example, if the request is to find the status of the invoice raised against a purchase order, the invoice number / PO number can be “read” by the machine and used to fetch the status from the backend ERP system. This principle of sensing can also be replicated to respond back to the requester.
What’s more, careful report analysis can help identify the complexity and resolution time of requests, get a big picture on the enterprise’s procurement spend, eliminate agent productivity hurdles, correct process inefficiencies, and ultimately, understand the satisfaction level of email requesters through sentiment analysis.
Machine FirstTM Advantage
The above Machine FirstTM approach brings great benefits as it allows machines the first right of refusal against any email that is received. Machine FirstTM approach to email automation frees employees’ time to focus on more value-add activities because of it:
Reduces routine and redundant manual work by at least 60% in the first six months
Enables automated monitoring and classification of emails and alignment to relevant departments/teams
Picks out escalations and prioritizes those emails for immediate action
Accelerates problem identification and enables reduced TAT and improved SLA from days to hours or even minutes
Equips the support teams with the necessary information to execute optimal actions and more time to resolve complex requests
Increases categorization accuracy from 75% to anywhere above 94% over time by aiding continual learning through machine learning.
About the author(s)
Krishnan Ramanujam is President and Head of Business & Technology Services at Tata Consultancy Services (TCS). He leads Consulting and Services Integration, Cognitive Business Operations, Cloud Platform Services and Digital Transformation Services globally. Krishnan drives forward the vision, direction and go-to-market strategy for TCS’ Services Lines. In addition to fostering the development of new services and solutions, he also guides the complex global transformation initiatives for the world’s leading enterprises.
He also drives growth and profitability for companies by spearheading and leading their evolution to next generation, agile operating models and transforms business functions. Krishnan has successfully positioned TCS as the industry’s leading expert in enterprise transformations by developing and leveraging best-in-class experts and offerings in Consulting, Cloud, IoT, AI, Analytics, Enterprise Applications and Interactive Services.
In addition to helping TCS’ clients transform their businesses, Krishnan is focused on transitioning TCS to be fully agile by 2020 – upskilling and reskilling thousands of employees, building collaborative workspaces, enhancing the management of contracts and partnerships and improving customer service.
Krishnan has been a part of TCS for the past 28 years and has rich experience in business and technology consulting. He has held several key leadership positions such as the global Head of Consulting Enterprise Solutions, Chief Operating Officer of TCS Financial Solutions, Director for the State Bank of India Group Core Banking Program, and head of TCS’s Global e-Commerce & Enterprise Application Integration practice. He also had a brief stint with Tata Internet Services as its Chief Technology Officer.
Krishnan joined TCS in 1991 after completing his Master’s degree in Electrical and Computer Engineering from the Rose-Hulman Institute of Technology, Indiana, USA. He also holds a Bachelor’s degree in Instrumentation and Control Engineering from the Government College of Technology, Coimbatore, India. Krishnan is an excellent speaker and as a thought leader, he speaks in global conferences and actively interacts and shapes opinion among industry analysts.
Krishnan lives in Mumbai, India with his wife and two daughters. An avid reader, he enjoys non-fiction books, music, movies and tennis, and is passionate about promoting education in India’s rural communities.