Today, we’re joined by Vijay Raina, a leading expert in enterprise SaaS technology, to demystify the journey of intelligent automation. While the promise of AI-driven efficiency is compelling, many organizations find themselves stuck in pilot mode, unable to achieve scale. Vijay is here to shed light on what true readiness looks like, moving beyond the technological hype to explore the foundational pillars of success. Our conversation will cover the critical importance of mature processes and clean data, the art of building a compelling business case with the right metrics, and the necessity of fostering cross-functional collaboration. We’ll also delve into the human side of this transformation, discussing how to manage change and empower employees, and how to shift the organizational mindset from treating automation as a one-off project to embedding it as a continuous, strategic capability.
The article notes that process fragmentation is a top barrier to scaling intelligent automation. For an organization just starting out, what are the first practical steps to establish clear process ownership and create standard operating procedures that actually get used by the team?
That’s a fantastic place to start, because it’s the bedrock of this entire endeavor. The reality, as studies from firms like Deloitte have shown, is that fewer than 40 percent of companies have mature, standardized processes. The first step isn’t technology; it’s clarity. You begin with process mapping, but not as a sterile, top-down exercise. You have to get the people who actually do the work in a room and ask them to walk you through it, step-by-step. This uncovers the messy reality of how things truly operate. From that map, you assign a single, unambiguous owner for that end-to-end process. That person is now accountable. The final piece is creating Standard Operating Procedures that are living documents—simple, visual, and stored where people can actually find and use them. When your team can confidently explain not just how a process works, but why it works that way, you’ve built the stable foundation IA needs to amplify, not just automate, the existing chaos.
With over 60% of firms seeing IA ROI within a year, what specific metrics—beyond time saved—should leaders use to build a compelling business case? Could you walk us through creating a consistent value framework to compare different automation initiatives?
It’s encouraging to see reports from places like Automation Anywhere confirming that rapid ROI, often within 12 months, is achievable. However, the most mature organizations move beyond the simple metric of “hours saved.” A robust value framework provides a common language for the entire business. You should be looking at a basket of metrics. First, there’s the cost per transaction—how much does it cost us to process one invoice or onboard one employee? Then you have error and rework rates; what percentage of our work needs to be fixed, and what’s the cost of that? Don’t forget risk and compliance—can automation improve our SLA compliance or reduce regulatory penalties? By creating a simple scorecard that weighs these factors—cost, quality, speed, and risk—you can objectively compare an opportunity in Finance against one in HR. This framework stops the “loudest voice in the room” from winning and ensures you’re always investing your automation dollars where they will deliver the most significant, measurable impact.
The text highlights the critical difference between RPA and a broader IA ecosystem. In your experience, what is the most effective way to explain this distinction to leadership to prevent unrealistic expectations and align investments with end-to-end process automation?
This is one of the most critical conversations to get right. I often tell leaders to think of it like building a house. RPA, or Robotic Process Automation, is like giving your team a power tool—say, a nail gun. It’s incredibly efficient at one specific task: driving nails. You can use it to build walls faster. But it can’t design the blueprint, pour the foundation, or install the plumbing. Intelligent Automation, on the other hand, is the entire construction crew with the architect, the project manager, and a full set of coordinated tools. It’s an ecosystem that uses process mining to find the best place to build, AI to make smart decisions, and analytics to check the quality of the work. IA doesn’t just automate a single task; as PEX Network defines it, it automates the entire end-to-end process of building the house. Framing it this way helps leaders understand they are investing in a capability to transform how work gets done, not just a tool to speed up a small piece of it.
Citing research that shows measurement maturity triples the chance of scaling AI, could you describe the process of establishing baseline metrics? For example, how would you measure the “cost per transaction” or “error rate” for a process before introducing automation?
That IBM research is so powerful because it proves that you can’t manage what you don’t measure. Establishing a baseline isn’t a complex data science project; it’s a gritty, hands-on effort. To measure the “cost per transaction” for something like vendor payments, you sit with the accounts payable team. You calculate their fully-loaded hourly cost, and then you time them. How many invoices can one person process in an hour, on average? You do this for a week to get a reliable baseline. For the “error rate,” you do something similar. You take a sample of 100 completed transactions and manually audit them. How many had incorrect coding? How many were duplicates? How many required a follow-up email? This creates your “before” snapshot. It feels tedious, but having that concrete data—”Our pre-automation error rate is 15%, and it costs us $5 per transaction”—is the only way you can later prove that your IA initiative reduced the error rate to 1% and the cost to $1. Without that baseline, your success is just an anecdote.
Gartner notes that poor data quality is a primary blocker for AI, costing millions. When an organization decides to tackle this, what is the most crucial first step in building a data governance framework, and how do you assign clear data ownership effectively?
That Gartner figure of a $12.9 million annual cost is staggering, and it highlights that this isn’t an IT problem; it’s a fundamental business problem. The most crucial first step is to demystify data governance and make it about accountability, not bureaucracy. You start by identifying the most critical data elements for your key processes—things like “customer address” or “product SKU.” The first step is to assign a named individual from the business, not IT, as the “data owner” for each of those elements. This person isn’t responsible for the database itself, but for the rules and quality of that data. They get to define what a “complete” customer record looks like. They are the go-to person when data quality issues arise. By assigning clear, personal ownership, you transform data from a nebulous, shared problem that no one fixes into a managed asset that someone is responsible for protecting.
We know that co-creation between business and IT is key to success. Could you share an example of how a successful cross-functional team makes decisions? How do they typically navigate disagreements when prioritizing processes for automation or selecting new technology?
The most successful teams I’ve seen operate less like a project team and more like a permanent business function. Picture their weekly prioritization meeting. The finance process owner presents a business case for automating invoice reconciliation, highlighting high error rates. The IT architect immediately chimes in on technical feasibility, noting that it will require an API to a legacy system. A frontline AP clerk then adds crucial context: “The biggest delay isn’t data entry; it’s waiting for manager approvals on non-PO invoices.” This is where the magic happens. Instead of a disagreement, it becomes a richer problem-solving session. They navigate these moments by always referring back to their shared value framework. The decision isn’t based on what’s easiest for IT or what the finance director wants most; it’s based on what will move the needle on their agreed-upon metrics of cost, quality, and speed. This co-creation model, which Deloitte’s research shows makes you up to 70 percent more likely to scale, turns potential conflicts into better, more holistic solutions.
The article points out that fear of job loss is a major hurdle. What does a successful communication plan look like in practice? How can leaders frame automation as a tool that supports employees, and what does a concrete skill transition plan involve?
This is where leadership truly earns its keep. A successful communication plan is not a one-time email; it’s an ongoing campaign of transparency. It starts months before the first bot is deployed, with leaders openly acknowledging that roles will change. They frame it honestly: “We are automating the repetitive, draining parts of your job—the copying and pasting, the manual data checks—so you can focus on the parts that require your expertise: analyzing exceptions, talking to customers, and solving complex problems.” This is then backed by a concrete skill transition plan. This means identifying the future skills needed—like data analysis or process improvement—and providing specific, funded training pathways. It involves workshops on “working with your digital teammate” and redesigning job descriptions to reflect higher-value work. When employees see a clear, tangible path from their current role to a more engaging future role, fear is replaced by opportunity.
Process mining is described as the bridge between process excellence and intelligent automation. Can you walk us through a real-world scenario where a company used process mining to uncover a non-obvious automation opportunity that delivered far greater value than their initial, intuition-based ideas?
Absolutely. I worked with a company that was convinced their biggest problem in order-to-cash was manual order entry. They were ready to invest heavily in an OCR solution to automate it. Before they did, we convinced them to run a process mining analysis on their system logs. The software created a visual map of their actual process, and the result was stunning. Manual order entry was a minor issue. The real bottleneck, invisible to everyone, was in the credit check process. Orders from established customers were being routed through the same lengthy, manual credit approval as brand-new clients. The process map showed these orders sitting in a queue for days, creating massive delays. By implementing a simple automated rule—”if the customer has been with us for over two years and has a perfect payment history, bypass the manual credit check”—they shaved days off their cycle time. It was a far less expensive and exponentially more valuable automation than what their intuition had told them to do.
McKinsey’s research indicates that the highest value comes from deploying automation across functions, not in silos. What strategies can an organization use to encourage this? How can a success story in finance, for example, be leveraged to inspire adoption in HR or operations?
Breaking down those silos is essential for moving from isolated wins to enterprise-level transformation. The most effective strategy is to create a formal “Center of Excellence” or CoE, whose job is not just to build automations but to be an evangelist for them. When the finance team has a huge win—let’s say they automate their entire account reconciliation process and reduce the month-end close by three days—the CoE’s job is to shout it from the rooftops. They don’t just send an email; they create a professional case study with hard numbers. They host a “lunch-and-learn” where the finance process owner presents their journey. Then, they proactively go to leaders in HR and Operations and say, “We did this for Finance. We saw a 40% reduction in manual effort. What is your equivalent of account reconciliation? Let’s find it and do it again.” By actively marketing internal successes and providing a clear path for other departments to replicate them, you create a flywheel of adoption that naturally crosses functional lines.
The final sign of readiness is treating IA as a long-term capability. What does this look like from an operational standpoint? How do mature organizations structure their teams and governance to ensure continuous improvement, rather than having automation efforts fizzle out after a project ends?
This is the ultimate shift from doing projects to building a muscle. Operationally, it means automation is no longer funded by one-off project budgets. Instead, there’s a permanent, dedicated budget for the Automation CoE, just like any other business department. The team structure is stable, with roles for developers, business analysts, and continuous improvement experts. Governance becomes embedded in the normal operating rhythm of the business; quarterly business reviews, for example, will have a standing agenda item on the automation pipeline and the performance of existing bots. Most importantly, the mindset changes. High-performing organizations, as the PEX Network report shows, don’t ask, “When is the automation initiative over?” They ask, “Our bots saved us 20,000 hours last year; what’s our target for next year, and which processes will get us there?” It becomes an endless cycle of identifying, automating, and improving—an integral part of how the company operates and competes.
Do you have any advice for our readers?
My single biggest piece of advice is to resist the temptation to lead with technology. The success of your intelligent automation journey will be determined long before you select a vendor. Focus first on your foundations. Get obsessed with understanding and standardizing your core processes. Invest in the unglamorous but essential work of improving your data quality. Most importantly, bring your people along on the journey from day one. Communicate openly, invest in their skills, and celebrate them as the heroes who are driving this transformation. If you build a strong foundation of process, data, and people, the technology will be an incredible accelerator. If you don’t, it will just be an expensive disappointment.
