The Business Process Automation Journey & Practical Examples

Kristen Daihes Partner
Read Time: 6 minutes apprx.

Why all the hype around business process automation? According to market research firm IDC, companies lose 20-30% in revenue every year due to inefficient business processes. That’s a lot to leave on the table. Using machine learning and optimization to automate some of these processes has a high ROI that can be achieved in a matter of months. This blog will attempt to demystify business process automation, layout the journey of a business process automation initiative, and provide some practical examples. Hopefully some of this content helps you kick-start your journey to go after some of that lost revenue.

Demystifying Business Process Automation

We constantly hear people talk about business process automation – but what is it exactly? It’s technology-enabled automation of activities or services. This is everything you go through from the steps you take to pay an invoice, process customer orders, set inventory levels, ship product to customers, and more.

Let’s take order processing as an example. A lot has been written about achieving “zero touch” status, but the reality is most companies struggle with significant customer order workflow issues. In fact, in some companies I have seen them touching more than 70% of their orders to deal with mismatches in pricing, UPC codes, item codes in transitions, or changes in customer delivery. Think about the benefits in both efficiency and effectiveness by finding a way to drive more automation into this process!

My colleague Mike Watson wrote an interesting blog on What Toyota, Schneider National, Paypal and Palantir got right. Toyota’s secret was autonomation – “automation with a human touch”. Schneider National used automation as a stretch goal, designing optimization systems knowing they will be augmenting people. His blog has lots of great nuggets to consider like figuring out how to use technology to do what it does best and use people to do what they do best. The key to success is figuring out the right mid-point for your business.

The Journey of a Business Process Automation Initiative

How do you go after a business process automation initiative? Here are four components to getting started.

  1. Value Proposition – clearly defined & visible, laying out the size of the prize
  2. Capability – ability to unlock value
  3. Quick Wins – building out forward momentum
  4. Buy-In – ability to go farther, achieved through continued quick wins and forward momentum

A methodology we have found success with at Opex Analytics contains 3-steps. First, surface exploration focuses on business discovery through interviews, group forums, and working sessions across a broad set of individuals supporting a process. The goal is to determine whether any immediate insights surface that merit digging deeper into areas such as process bottlenecks, data flow, insights, or business and geographic distinction. Next, research sprints are conducted. This is basically experimentation with different data sources to test hypotheses and determine possible value contribution. The objective is to identify potential improvement opportunities that can be explored as a proof of concept within 4-6 weeks. The final step is evaluation of the proof of concept, the decision to implement the change or instead to fail fast and choose another avenue of exploration.

Opportunity sizing for process automation is important. Find a way to map out your opportunities visually – we like to plot opportunities out using the y-axis for “potential for advanced analytics” and the x-axis for automation time savings. Then you have something tangible with identified size of the prize to use in gaining alignment within the organization.

Practical Examples of Business Process Automation

  1. Data Products. Data products are great examples of solutions that are easy to develop while automating key manual, repetitive tasks. These products can often be developed using data engineering applications such as Alteryx, or creating optimization algorithms using open source software (such as Python PuLP) and using a platform to deploy as an application for a business user to use. Some practical examples many organizations can go after very quickly often include:
    • Load building
    • Anomaly order detection
    • Data refresh for network design
    • Carrier assignments
    • Routing
    • Box selection
    • Retailer back room box unloading
    • Priority unloading of drop trailers in “the yard”

    If you would like to understand more detail or see example output of the above list of ideas, you can access a replay of the December 2016 Analytics Academy webinar I delivered here.

  2. Intelligent Dashboards. In addition to data products, intelligent dashboards can add value in improving decision quality, speed in decision-making, automation, broader visibility in real-time driving speed to insight.

Intelligent dashboards and master tables can significantly reduce manual intervention. Let’s take the example of a VMI planner preparing to write an order. Often times these planners must touch multiple screens across different applications to gather the information necessary to drive action. Time to action can be significantly reduced by bringing all the relevant information into one place. What if a tailored dashboard was at the planners fingertips with visibility to historic demand, historic promotion history and lift, real-time inventory and trends, trending coupons from manufacturer and retailer, UPC change and item discontinuity visibility, or promotion order quantity?

Intelligent applications can also be developed by collecting data shown to planners and tracking decisions planners actually made, thereby improving intelligent app performance through reinforcement learning.

Going Beyond Business Process Automation

Automation is just one lever in driving value. As an organization’s analytical maturity increases, there is increasing value creation at an enterprise level.

Let’s use inventory risk as an example. Pure business process automation may drive more efficiency in safety stock target setting for products within a portfolio. Moving into descriptive analytics you can start to understand what your inventory looks like. Diagnostic analytics can then help to understand the root causes behind stock-outs. Predictive analytics can help identify which SKUs are likely to go out of stock in the next 3 weeks. And prescriptive analytics can help identify which SKUs to produce, where to stock them, how much to stock and when.

In Summary

Business process automation can add enterprise value. You should explore building and deploying data products to automate manual, repetitive tasks. Enhance your journey with intelligent dashboards to help drive speed to insight reducing time to action and improving decision-making, and even begin to explore deploying intelligent applications. Most importantly, figure out how to use technology to do what it does best and use people to do what they do best. Find the right mid-point for your business.