How to consider automation for anything but not everything

Download a guide to implementing the best Customer Experience Strategy

Free Download

Download our latest white paper on how to deliver an excellent customer experience (CX) with ease and quickly.


Get your FREE whitepaper

Organisations are now embracing Automation by investing time, cost and efforts to ensure they are not left behind their peers in the industry. Automation and more specifically Robotic Process Automation (RPA) implementations are increasing dramatically day-by-day. All industry verticals including BFSI (Banking, financial services and insurance), Pharma & Healthcare, Retail & Consumer Goods, IT (Information Technology) & Telecom, Communication & Media & Education, Manufacturing, Logistics & Energy & Utilities and the rest are investing and implementing RPA initiatives.

Grand View Research Report

This report published by Grand View Research shows the

  • Market share spread across for each industry vertical, where BFSI has taking huge marketshare compared to other verticals.
  • Market size valuation year-on-year and projected to be increased in fast pace from 2021 to 2027.

This clearly states how organisations are exponentially increasing the investments towards RPA. So it is very critical for organisations to understand various principles and methodologies to differentiate between the right and wrong candidates of business process for automation. Any RPA implementation can fall into one of these 5 categories :-

API Integration : Integration using API’s enable single entry and single source of truth.

Legacy Applications : Applications that cannot be immediately migrated while still running with slow performance.

Data Extraction : Extracting, analysing and process of data is best suited for Data Management Platform.

Dashboards and Reporting : MI and BI reports generated using standard reporting tools.

Decision Making : Learn and understand decision making process by monitoring and recording.

All of these categories have to be looked at to understand when to automate and when not to automate. The diagram depicted below explains where your business process falls into which category and what are the chances of implementing with specific technology.

Auomate Anything vs Automate Everything

Automate Anything

API Integration : Anything (BOT’s implementation – 20%). Implement for what is only necessary leading to an API-led integration.

Legacy Applications : Anything (Manual processes run by Humans – 30%). More simpler activities or unstable functionality should be left out of automation avoiding increased maintenance. Always looks for optimising existing automations before implementing new ones.

Data Extraction : Anything (Screen scraping to extract data – 20%). Only extracting data with scraping where an application does not provide any other means such as export data into a file, read data using API or access to backend database.

Dashboards and Reporting : Anything (BOT’s implementation – 20%). Dashboard and report generation with BOT’s should be encouraged only if it is not possible with standard data platforms and reporting tools.

Decision Making : Anything (Human understand for making decisions – 40%). Decision making sometimes is very complicated which might need either too much information or enough experience or in some cases multiple members of the staff.

Do not Automate Everything

API Integration : Everything (API implementation – 80%). Aligning with Integrations strategy promote and implement only API based integration.

Legacy Applications : Everything (BOT implementation – 70%), RPA should be the first port of call as any other integrations or automations might add be very expensive and harder to implement. Always ensure all of these automations are governed during the lifecycle of the application.

Data Extraction : Everything (Use of either ETL or API – 80%). Always use the best practice methods to extract data such use ETL for directly extracting the data from the database, use API to read data or use automated file exports to extract the data required.

Dashboards and Reporting : Everything (Industry standard dashboard and reporting tools – 80%). All standard dashboard and reports can be defined with the tools using the data produced by the data management platform.

Decision Making : Everything (Intelligent automation BOT’s implementation – 60%). All RPA softwares that are available in the market now support IA Bot’s and the technology is getting mature day-by-day. Deep understanding of a bot can learn and make decisions over time is key for this implementation.


RPA always seems to be the right technology for automating any business process which has proven completely wrong based on the explanation provided above. RPA can be applied for anything but keep in mind that it should not be applied for everything. The era of monolithic application has now moved to microservices based applications which shown us the amount of pain incurred to move from one to another. In the similar fashion RPA implementation should be considered as a microbot implementations that is self sufficient and easy to maintain, most importantly easy to dispose if the application is no longer required. RPA implementations should follow a scalable, sustainable and failsafe architecture.

ACS has helped organisations in setting up Automation as a practice. Complete the form and you can receive a FREE no obligation consultation.

Suggested reading

Subscribe to receive the latest insights to your inbox

Please see our Privacy policy for more information.