Successful Automation Projects – Advice from the Trenches

by Dan Rotelli | March 10, 2020

Let’s face it, automation rules the day. Give me a problem, and I’ll show you how automation is the answer. However, with the dizzying array of automation solutions available, finding the right solution the first time isn’t so easy. I always say “the first time is the hardest time” so I’m going to share some tips from personal experience that should make life much easier for you.

Take a Step Back

You’ve decided you need an automation tool (or tools). That’s great. But before you start scheduling demos and compiling RFI’s, take a step back and focus on the key business outcome you’re concerned about.

You need to start with the critical business problem and then work your way back from there. Your tech stack needs to align with your desired business outcomes, not the other way around. That seems obvious, but what we’ve seen happen so often is that automation projects are approached from a technology perspective first.

And I’ll tell you right now, that by focusing on the business-case first, you will end up with an orchestrated portfolio of solutions. Unfortunately, with enterprise automation there’s no silver bullet and no single-point solution.

Your desired business outcome should include ideas like new ways of working, additional services you’ll be providing, new levels of engagement (both inside and outside your business), and new business outcomes that improve existing operations.

In my opinion, the healthcare payer industry has some of the toughest challenges to solve. And it makes sense because modernizing these systems provides massive return on investment through efficiency gains.

After you’ve clearly defined what problems need to be solved, and the desired outcomes, you can focus on automating the right process steps (rather than automating bad processes).

Focus on the Right Processes

What's the most ideal process required to solve the business objective? Starting with the process ensures you focus on the right things first. What process steps need to be enhanced, added, or eliminated?

Here’s where automation begins to take stage. If you know early-on in a process that you’re dealing with paper-based claims, for example, you know you’ll need a solution automate optical character recognition. Let the process dictate the needs.

The best way to analyze processes is with the help of a business analyst (BA). Most seasoned vendors provide BA services to augment your own. They’ll ensure front-line workers are involved and help key business stakeholders define what success looks like.

Many times, in process redesign, you’ll discover ways of integrating internal and external data systems to completely remove steps that human workers used to perform. In healthcare, for example payers can streamline eligibility processes by automating manual data entry look-ups, or entirely remove steps in financial reconciliations by automating data extraction and verification steps.

How Choose the Appropriate Tools for the Job

As a vendor in intelligent automation, I’ll be first to tell you there’s a lot of shady sales propaganda out there. And we’ve seen it all from our customers’ horror stories. From sweeping claims about A.I. that can solve it all, to easy robotic process automation that tackles any manual task.

Automation, done right, will involve multiple solutions and technologies all working in harmony.

Don’t be too quick to pull the trigger on a one-size-fits-all solution. They don’t exist. You really need to partner with vendors who are willing to go on a journey with you. While outcomes may be similar within industries, the path to get there is always unique. I’ve found that the best outcomes come from a unified vision between the vendor and the customer.

And the worst outcomes are certainly from mis-matched expectations and lack of adhering to a master data model. When multiple departments deploy solutions without communicating outside their own lines of business, it almost always spells trouble.

The bottom line is that your organization’s executive team has got to be involved because automation almost always requires redesign of core processes. What is the cost of failure? And what’s the definition of success in terms of revenue? You’ve got to be clear.

In order to choose the best tools, you’ve got to understand a few key things.

3 Keys to Choosing the Right Automation Solutions


What’s the intended workflow going to look like? You need a solid understanding of overall volume and the sources of information. All too often we see technology "fail" because of something simple like a change in a form’s layout. Another big hang-up is workflow that changes based on logic or human decision-making. You’ve always got to factor in the human element.


This is a big one. It is easy to assume that data is easy to get to. In reality, it most often isn’t. There’s always going to be the problem of how to handle unstructured or variable (changing) data sets. This includes data in the form of paper, video, and audio. You might have an ongoing digitization problem that needs to be solved. Different automation tools have different strengths and weaknesses. By understanding your data needs first, you’ll reduce the risk of surprises later on.


How will you integrate or enhance legacy systems? Is a pre-built solution a better choice than building something using existing tools? I think the most overlooked considerations in system decision-making are the costs to train and implement the solution. Is training going to be painful? Will you need new specialist skill-sets to ensure success?  If the solution uses machine learning, for example, will you be able to build a single rules-based model that doesn’t change, or will you need to build new models on the fly?

4 Steps to Achieving Wisdom You can Use at Work Today

4 Steps to Achieving Wisdom You can Use at Work Today

How to create an Information as a Second Language program. [Free Guide]

4 Steps to Achieving Wisdom You can Use at Work Today

4 Steps to Achieving Wisdom You can Use at Work Today