Hey everyone! 👋 I've been noticing a lot of questions lately from folks who are just starting to look at new agent tech (think dialers, conversational AI, copilots, etc.) and are totally overwhelmed by the ROI piece. I get it—it can feel like you need a finance degree just to get started!
Here’s how I usually break it down for a first-pass analysis. You don't need anything fancy—just a spreadsheet and some solid estimates.
**First, map out your core costs (the TCO side):**
* **Software Licenses:** The obvious one. List price per user/per month.
* **Implementation & Onboarding:** Often a hidden initial cost. Get a quote.
* **Integration Costs:** What will it take to connect with your CRM (like Salesforce!)? Factor in dev time or consultant fees.
* **Ongoing Admin & Maintenance:** Will you need a dedicated person? Even 10% of an FTE's time has a cost.
* **Training & Change Management:** Getting your team to actually use it is a cost.
**Then, model the tangible benefits (the ROI side):**
This is where you need to get concrete. Think about what the tech promises and how to measure it.
* **Productivity Gains:** If an AI copilot saves an agent 30 minutes per day, what's the value of that recovered time? Can they handle more contacts?
* **Increased Conversion/Upsell:** If the tool provides better insights or scripting, can you attribute a lift in win rates or average deal size?
* **Reduced Ramp Time:** For new hires, can they become proficient faster?
* **Deflection/Handle Time:** For support tech, can you handle more chats per hour or deflect tier-1 tickets?
My biggest tip? **Start with a single, key metric you know you can track.** For sales teams, it's often "increased revenue per rep." For support, maybe "contacts handled per hour." Build your model around that one driver first. It keeps you from getting lost in the weeds.
Would love to hear from others—what was the first metric you focused on when you built your first agent tech ROI model? What surprised you the most when you looked back after 12 months?
—Amy
You're missing the biggest TCO line for cloud-based agent tech. Data egress and API call costs will blow your spreadsheet up if you don't model them.
For productivity gains, you need to factor in the cost of the time saved. If an agent saves 30 minutes but that time isn't used for revenue-generating work, it's not a benefit, it's just idle capacity. Model the actual outcome, like increased call volume or sales, not just hours.
Show me the bill
That's a really good point about the time saved needing to be productive time. I've seen teams assume savings automatically turn into revenue, which isn't always true.
What would you recommend for measuring that outcome? Like, do you track the new call volume directly in Jira or your CRM to prove the link?