12/28/2023
There's no one-size-fits-all way to analyze intent performance…
But with the right framework, you can better figure out what's working and build a plan to improve your bot.
Jay Athia from Wysdom shares 3 insights that offer a fresh perspective on measuring intent performance:
1️⃣ Analyzing intent performance isn’t enough
Intent is how businesses view the conversation journey, but customers might see it differently.
For an effective bot improvement plan, look at real customer conversations. Think of contact reason as an audit for your intent – it helps identify gaps between what's programmed and what happens.
2️⃣ Containment is a false measure of bot performance
While containment might seem like it shows how well your bot handles conversations, it can be misleading. It only reveals what happened within the bot and doesn't indicate if the customer switched to a more expensive channel.
Look at other signals like automation rates and sentiment, and importantly, consider the live agent call handle time for the same topic to get a full picture of intent performance.
3️⃣ Signs to look for when planning your bot improvement roadmap
When planning to improve your bot, you'll have many ideas and input from different parts of the business. To ensure your changes have the impact you want, it's crucial to consider performance from various angles.
For example, if containment is falling in a specific topic, your first instinct might be to fix it. However, checking experience metrics could show that the bot is doing well. On the flip side, if containment rates seem good, a closer look at automation or sentiment might reveal a decline in customer experience.
💡 Measuring performance across a suite of metrics gives you a more holistic understanding of what’s working or not, so you can make sure your roadmap is addressing functionality that will have a positive impact.
