The majority of new technology implementation projects fail to meet their objectives. But they’re not good ideas poorly executed. They’re poorly thought through from the start. Most fail before the vendor is even selected. The decision gets made at the executive level, the software gets purchased, and then someone realizes the process it’s supposed to fix was never clearly defined. But even those that are backed by good intentions all too often fall into the same trap. The roll-out plan accounts for user training, usually as an afterthought in the broader change management process. But when expectations aren’t clear, understanding a few new buttons won’t make a difference. Over time, your team stops using the new tool. It might be a great tool. But the workflow nobody bothered to reassess remains the biggest bottleneck.
Don’t digitize a broken process
An analogous concept in process improvement circles is known as “paving the cow path”. This means taking the informal, inefficient route that people have used and worn down over time and making it permanent by building infrastructure around it. This is what occurs when companies automate a manual process without first questioning whether the process itself made sense.
Before you even think about evaluating software, document what is actually happening – not what your SOPs say should be happening. Run through it with your users. You will find workarounds, shortcut spreadsheets, steps that only exist because they were built into a system that was configured poorly three product versions ago. Business process mapping at this stage is not exciting, but it will quickly tell you whether you have a process problem or a tool problem. They are not the same thing. A process problem needs management and redesign. A tool problem needs technology. Mix the two up, and you are how companies end up spending six figures on a piece of software no one uses.
Run a waste audit before you go looking for solutions
Lean methodology can be a helpful guide in this context. It encourages you to scrutinize your processes in terms of common sources of waste: waiting time, over-processing, unnecessary handoffs, defects and rework, and excess motion. Identify the presence of each waste category in the work of the team, and the frequency of their occurrence.
This simple process has a couple of useful effects. First, it reveals the actual bottlenecks in your operations – not the ones you thought were blocking you. Second, it guides you as to whether technology offers the right solution to the bottleneck. For instance, waiting time due to a work flow that requires four signatures for any decision you make is not a technology issue. Nor is potential over-processing due to a redundant report that is filed. This is not a data issue: it is a governance issue. The waste audit will help you distinguish between the noise and any signal, so that any technology you introduce addresses a real issue.
Map the data journey before you build anything
In order to truly thrive, new technology needs clean, coherent information as its lifeblood. This is often the unexpected roadblock that grinds new tech implementations to a halt. The ai implementation challenges that rack up the most technical debt have nothing to do with tech – they’re all about your organization and how it currently handles information. Data is walled off between departments. Every team or office has its own convention for tracking the same process. The details you need in order to make analytical decisions are scattered through email threads and comments on a shared spreadsheet.
These are the red flags to look for: if you don’t recognize the description above, where are you finding value in your current software stack? What niche solutions have teams taken up using that your procurement officer would think are unnecessary? Drop every official line of inquiry and take a good long look at what your teams are really doing – especially the ones you’re counting on to evaluate options with an open mind.
Watch it happen
Walk the entire path that a piece of data follows from the time some external party first sends or enters it until it comes back into your technology suite as a factor in a decision. Where does it enter your business? Where does it get written down, saved, and handed off to the people interfacing with your clients or governing bodies? Where does the client ask you a question then wait while you tab through a few windows looking for the answer? Not surprisingly, wherever that happens is most likely the same place where competitors are gaining an internal edge.
Calculate the real cost of your current workflow
You cannot begin to justify any new investment unless you have a realistic number for what your current process costs, including time spent on manual steps, error rates and rework costs, the staff hours consumed by coordination and status updates, and any compounding inefficiencies from legacy systems that don’t communicate.
Then compare that against the total cost of ownership for the new workflow – not just the software license but the implementation time, training, lost productivity during transition, and maintenance. If the math doesn’t clearly favor the change, or if the break-even timeline is two years out, that’s information worth having before you sign anything.
An audit done well won’t kill your technology initiative. It will make it better. It gives you a precise lay of the land on which a new tool will be asked to perform and the demands it must meet to deliver the ROI you’re promising.
