Telecom & Technology Consulting
When it comes to launching new technology, ROI is always the ultimate measuring stick. That’s especially true with AI initiatives, which carry a special sense of urgency. The question is, when is it realistic to know if the investment paid off?
Despite the promise that AI will automatically deliver positive results, it’s common for companies to be challenged to locate ROI after launch. In our experience, tech leaders would be wise to not expect those results too soon. Improved customer experience, employee productivity, and new revenue streams often need one to three years to be realized.
The problem: When organizations are too focused on quick ROI, they may pause the AI project before it has time to properly run its course. Instead, they need patience – to allow time for the initiative to unfold and see what opportunities for adjustment may arise.
Measuring ROI prematurely isn’t merely due to lack of patience or unrealistic expectations – in many cases, it’s the result of poor planning. In order to plot a prudent path to assessing ROI, tech leaders must first establish sound expectations and metrics.
CIOs also have to recognize that all AI projects can’t be measured in the same way; it’s not an apples to apples comparison. For example, AI projects that are deployed for general tasks can take longer to deliver ROI than more specific AI projects.
However, in both cases, ROI can only be truly measured in separate increments over time rather than a landslide that appears in one fell swoop. Again, patience is the key.
It’s also important to consider if an AI deployment is even the right way to solve a specific business issue. Unfortunately, enterprises have been jumping on the AI bandwagon not necessarily because it’s the right technological approach, but because they fear falling behind the competition.
Blame for this type of reaction often goes to boards and consultants who, without conducting due diligence, pressure CIOs to adopt the current hyped technology. So in turn, the company invests in the wrong initiative, which leads them on an endless, fruitless effort to measure RIO when there’s in fact nothing to measure. Ironically, a supposed solution (AI) to a problem creates an even bigger problem.
To be successful, and to avoid wasting resources, launch your AI projects with a small window for evaluation. This way, you can leverage the solution with a cost-effective option. Develop a proof of concept, then a working model, and when you’re ready expand as necessary. They key is to avoid launching with a massive, company-wide project. Implementing in bite-size chunks allows you to accurately measure results and KPIs.
As for measuring ROI, it’s essentially impossible to do so immediately. In fact, many AI initiatives generally require 18 to 24 months to generate any level of ROI. But whatever you do, don’t get stuck on a specific timeframe. Be patient.
