How Do You Calculate Gen AI’s ROI?How Do You Calculate Gen AI’s ROI?
Business leaders should identify specific opportunities for productivity improvement.
July 29, 2024
As vendors have rolled out generative AI copilots and virtual assistants, most are doing so by paying an additional license fee to gain access to generative AI features. This additional cost leads IT and business leaders to question the value of generative AI and if sufficient ROI exists to justify the additional licensing cost. Even those with access to free generative AI virtual assistants may incur some costs such as user training and additional management overhead to set appropriate security, compliance, and privacy policies. Thus, the million-dollar (or $30 per month per user) question is “How do I determine ROI on my AI spend?”
Determining ROI
At Metrigy, we’ve typically looked at three measurements to define ROI for investment in any collaboration technology (not just generative AI). These are:
Increased revenue through the leveraging of improved collaboration capabilities to support sales activities or to deliver new services. For example, a company investing in high quality video endpoints may find that their virtual sales teams achieve a higher close rate, or that the company is better able to deliver services such as telehealth or distance learning.
Cost savings by reducing operational costs. Here, examples include reduction in licensing costs compared to legacy systems, reduction in staff time needed to manage platforms, and the use of new collaboration capabilities to reduce travel costs. Additional cost savings may come through being able to do “more with less” and reduce the need for headcount to support specific activities.
Productivity improvements through time savings. This is most easily measured by looking at repeatable processes such as employee and customer on-boarding time, average Scrum time, or average time required for product updates. Productivity improvements may contribute to both revenue gains and cost savings.
In all these areas, generative AI tools offer potential ROI benefits as shown in the examples below. Note that these examples all free up human time:
Revenue Increases | - Sales assistants to coach salespeople during live calls - Automated creation and optimization of sales and marketing collateral - Reporting and analytics to provide greater insight into opportunities and pipeline |
Cost Savings | - Operational management assistants to speed troubleshooting to reduce downtime - Meeting room analytics to identify usage patterns to enable more effective provisioning - Improving contact center agent efficiency by automating call summarization and action item processing, reducing headcount requirements |
Productivity Gains | - Automating meeting transcription and summarization to save time spent on post-meeting tasks - Summarizing chats and emails to save time spent reading messages - Content creation tools to speed time spent creating documents |
Obviously this is a short list, and many potential ROI benefits will be organization or industry specific.
Measuring the Benefit
Unfortunately, most IT and business leaders lack the time and resources to accurately identify and track ROI for collaboration spend. For example, in Metrigy’s most recent Workplace Collaboration and Contact Center Compliance and Security study, an average of only 12% of participants track ROI for their collaboration spend, typically through examining changes in spending, evaluating productivity improvements in repeatable processes, or looking at how improve collaboration benefits sales and customer support activities. ROI analysis is higher among those looking at how AI benefits the contact center, but there is still a long way to go before companies have true visibility into the benefits of their collaboration applications.
To determine generative AI benefit, IT and business leaders should start with an identified set of processes or roles and then measure the impact of the roll-out of generative AI. In addition, it is important to measure the utilization of generative AI tools to ensure that users are taking advantage of the capabilities made available to them. In both cases, companies can leverage management tools that may be available to them (e.g., Microsoft Viva Insights), task times tracked through project management tools, or simply through end-user surveys.
Converting Time to Money
One challenge with measuring generative AI ROI is translating time savings to money. One common justification for purchasing generative AI licenses is the scenario that virtual assistants or copilots will save employees three to four hours per week, and thus will pay for itself.
That time savings may be real, but the monetary benefit is only true if one uses those three to four hours per week for other activities that aid the bottom line. If I use those three to four hours a week of time savings to surf the web or play games on my phone, I’m not likely providing any ROI to my employer (unless that extra time improves well-being and job satisfaction, as my colleague Robin Gareiss noted.
Again, here the need is for IT and business leaders to identify tangible improvements linked to the way(s) the generative AI tool changed workflow.
Bottom Line
Not every company will need, or want, to identify specific ROI for generative AI. Some may be fine with rolling it out on the idea that it provides employee and/or competitive benefit. But for those struggling to justify the additional licensing cost, take the time to identify specific, measurable benefits that contribute to improvements in revenue or decreases in costs, potentially through gains in employee productivity. It takes a bit of work, but identifying measurable ROI will justify wide-scale generative AI investment.
About Metrigy: Metrigy is an innovative research and advisory firm focusing on the rapidly changing areas of workplace collaboration, digital workplace, digital transformation, customer experience and employee experience—along with several related technologies. Metrigy delivers strategic guidance and informative content, backed by primary research metrics and analysis, for technology providers and enterprise organizations
