Navigating AI agent monetization: the KPIs you can't ignore
AI agents are the hot new thing, and you shouldn’t treat them like a SaaS. They have agency, they can do stuff!
It’s new, but does that mean it is different?
Agents are not just a service. They’re like a coworker. They CAN perform actions on your behalf.
Because they’re different, monetizing these new AI agents effectively requires a strategic approach that goes beyond traditional SaaS pricing models.
It’s pretty crucial that you track the right metrics and use them to inform your pricing and go-to-market strategies. By monitoring their KPIs, you can gain valuable insights into your AI agents' performance, identify areas for improvement, and make data-driven decisions to get the most of your efforts.
In this article, we'll explore the essential metrics you need to track for effective AI agent monetization.
Understanding AI agent monetization
AI agent monetization is a new field that requires a different approach compared to traditional SaaS businesses.
Unlike SaaS models, where pricing is often based on a per-seat or per-user basis, AI agents provide value through their interactions and outcomes rather than individual seats or "credits."
To effectively monetize AI agents, you should think about their metrics that capture the true value of these intelligent systems.
You want to be agile, and adapt quickly, right? Well, this agility is crucial in the much much faster-paced world of AI, where things change at a higher pace.
Common KPIs for AI agents
Think about what your agent actually does.
Is it an SDR? If so, you want to look at how many meetings it books, how many leads it provided (ones that didn’t suck).
Is it a CS agent? If so, you want to see if it not only closed off issues, but also if it actually resolved it, and with customer satisfaction achieved.
Is it a content writer agent? If so, is the content it wrote actually being read?
To be truly valuable, these metrics need to be not only interesting, but also actionable. That means you change your business and pricing based on them.
They’ll expose your agents' strengths and weaknesses.
We can think of these KPIs in three main categories:
Financial metrics: These KPIs focus on the financial impact of AI agents, such as revenue generation, cost savings, and return on investment (ROI).
Quality engagement metrics: These indicators measure how effectively AI agents interact with the customer in mind, and that can be customers and the quality of the user experience they provide.
Operational efficiency metrics: These KPIs assess the performance and productivity of AI agents in terms of task completion, resolution rates, and handling time.
You should balance your KPIs across these categories.
Financial metrics to track
Revenue generation
Monitoring the revenue generated directly through AI agent interactions is a crucial financial metric for businesses. This includes tracking sales, subscription sign-ups, upsells, or leads created as a result of customer interactions with AI agents.
Basically, what was your agents’ contribution to overall revenue?
It's important to track the percentage of total revenue attributed to AI agent-assisted transactions.
Cost savings
In addition to revenue generation, AI agents can also drive significant cost savings by automating tasks and reducing the need for human intervention. Measuring the reduction in operational costs due to AI agent automation, such as decreased customer support expenses, is a key financial metric to track.
To evaluate the ROI of AI agent implementation, you obviously compare the cost savings achieved to the investment made on it.
Quality engagement Metrics
Interaction volume
Tracking the number of customer interactions handled by AI agents over time provides valuable insights into customer adoption and usage patterns, but also informs your pricing.
When you monitor trends in interaction volume, you identify peak periods and adjust your resources accordingly, but you can also introduce “surge pricing” based on it.
This metric also helps assess the scalability of their AI agents and plan for future growth. As interaction volume increases, you can proactively optimize your agents’ activities to handle the higher loads.
Customer Satisfaction Score (CSAT) and Resolution Rates
Measuring customer satisfaction ratings following interactions with AI agents is a very common metric for evaluating the quality of the customer experience.
CSAT scores provide insights into how well AI agents are meeting customer needs and expectations, beyond simply resolving their issues. You can probably charge more for high CSAT than lower CSAT.
Net Promoter Score (NPS)
While NPS is a flawed metric, it can still provide valuable insights into customer loyalty and the likelihood of customers recommending a business based on their experience with AI agents.
A low NPS score often indicates that the AI agent did not meet customer expectations or failed to deliver a satisfactory experience.
By monitoring NPS scores and bringing in your customer feedback, you can identify pain points and areas for improvement in their AI agent interactions, but you can also inform your pricing.
Should the customer be paying you for bad interactions, or those that caused a customer to churn?
Lead velocity rate (LVR)
While not perfect either, this is the growth rate of qualified leads. It provides you insights into future revenue, and tells you if your agents are providing good leads.
LVR is (qualified leads in that month - qualified leads last month) ÷ qualified leads last month.
So if you had 10 leads this month and 5 last month, the LVR is 1 (or 100%).
It’s generally considered a good indicator of growth (as opposed to MRR or ARR).
Your customers will want to increase the lead generation if they don’t have a high enough LVR - and you can charge for that.
Operational Efficiency Metrics
Resolution Rate
The resolution rate measures the percentage of customer inquiries successfully resolved by AI agents without the need for human intervention. It’s an indicator of the effectiveness and efficiency of AI agents in handling customer support tasks.
If your agent can’t resolve things, should the customer be paying you for it, and if so how much?
Average Handling Time (AHT)
Average Handling Time (AHT) measures the average time taken by AI agents to resolve customer interactions. This metric is crucial for assessing the efficiency of AI agents in providing support and resolving customer issues.
By optimizing AHT, businesses can deliver more efficient support experiences, reduce customer wait times, and increase overall satisfaction. Continuously monitoring and improving AHT can help businesses strike a balance between providing thorough support and maintaining operational efficiency.
First Contact Resolution (FCR)
First Contact Resolution (FCR) tracks the percentage of customer inquiries resolved by AI agents within a single interaction.
FCR is defined as the number of issues resolved on first contact ÷ total issues.
So if you solved 10 issues on first contact out of 25 total, the FCR rate is 0,4 or 40%.
A high FCR rate indicates that AI agents are addressing customer needs and minimizing the effort required to resolve issues.
Of course, that means you can charge for a higher FCR.
Taking these metrics to monetization
I have given a few very basic examples of what you can use the metrics with your pricing.
The point is you want to be dynamic and adapatable.
Examination of performance data reveals not only current capabilities but also future potential, where businesses adjust their buying strategies to match the value.
This is not just the A/B testing of yore. You want to find the best configuration of metrics to price and monetize.
The tl;dr is that by tracking the right metrics and continuously optimizing your AI agents, you can get more out of your AI Agents. Don’t settle for seat-based or credits.