Measure and improve

Tracking business impact: cost, revenue, and CSAT

Tracking business impact means proving the agent is not just busy, but valuable. Unless separates this into two clear conversations: one with operations about performance, and one with the CFO about money. The Analyze phase of the Engine brings both together, along with customer satisfaction, so you can see the full picture of what the agent is worth.

The three pillars of business impact

The Analyze dashboard organizes impact into three pillars: performance, business impact, and AI maturity. Performance covers resolution rate, deflection rate, latency, and CSAT, broken down per skill, per role, and per language. Business impact translates those numbers into euros, showing cost reduction and revenue generation. AI maturity shows where you sit on the ladder from Experimental to AI-First, with suggested next steps.

Each pillar serves a different audience. Performance answers the question “is the agent working?” for operations. Business impact answers “is it worth it?” for the CFO. AI maturity gives your DPO documented proof for procurement and audit.

Cost reduction: the savings side

The cost reduction tab shows tickets the agent deflected, team hours saved, and the ROI calculation that ties both together. Ticket deflection counts conversations the agent resolved without escalating to your helpdesk. The dashboard shows the deflection count, the rate as a percentage of total conversations, and the euro value based on your cost-per-ticket setting.

The AI resolution rate is the share of deflected conversations where the agent fully resolved the question. A high rate means the agent is finishing conversations, not just catching them. The escalation rate is the share handed off to a human, and a number that keeps dropping is better.

Costs have two components: your Unless subscription and AI management hours, the time your team spends configuring and reviewing the agent. Savings also have two components: money saved by deflecting tickets and money saved by team efficiency, which comes from hours the Team Assistant saved by drafting replies, summarizing context, and surfacing next-best actions.

ROI is net savings divided by total costs, shown as a percentage and as a euro amount per month. You can switch between reporting periods, and every number traces back to a Customer Agent action and a downstream event. Nothing is invented. Clicking into a number shows the conversations behind it.

Revenue generation: the growth side

The revenue side of business impact shows uptake by moment. Acquisition conversions, expansion plays, and retention saves are tracked separately, so you can see where the agent is capturing revenue across the journey. Roughly 60% of net revenue retention sits in Retention and Expansion, the moments most platforms ignore. Unless works all four moments, so it protects and grows revenue, not just deflects tickets.

Running AI across the full lifecycle unlocks 3.8x the annual value of support-only AI. Living Knowledge can also replace your help center vendor, saving EUR 40-120K a year. The biggest returns sit beyond the ticket queue.

Customer satisfaction: the quality check

The customer satisfaction dashboard shows CSAT scores, resolution time, customer ratings, and sentiment trends. Deflection without satisfaction is just abandonment with extra steps. This tab shows whether customers are actually happy with the answers the agent gives.

The CSAT score is the average rating customers give after a conversation, shown as a number out of 5 and as a percentage of positive ratings. Always read it alongside the response rate, the share of resolved conversations where the customer actually left a rating. A high CSAT on a low response rate is less reliable than a moderate CSAT on a high response rate.

Resolution time is the median time from first message to conversation resolved, shown as a trend. Lower is better, but not at the cost of satisfaction. A fast wrong answer is worse than a slower right one. Sentiment trend reads the emotional tone of each conversation, and it moves faster than survey responses, so it catches problems before they show up in CSAT.

The baseline comparison is a before-and-after view. Set the date you deployed Unless, and the dashboard shows the delta in CSAT, resolution time, and sentiment between the two periods. You can break down all metrics by Moment, and click into any metric to see the contributing conversations.

How the loop closes

Each phase feeds the next. Train builds the foundation, Analyze proves value and identifies improvements, and insights from Analyze feed back into Train to start the next cycle. The same dashboard tells your CFO and your DPO what they each need to hear, with the numbers backed by the audit trail. Clear metrics for your CFO and compliance documentation for your DPO, both from the same view.