See what drove each interaction’s experience: the agent, the bot, or the product

Every interaction gets a rating and a short explanation for each of three Experience Drivers, so you can see whether the agent, the bot or the product itself shaped how it went. Ask Spotlight already uses these ratings, and this release adds them to the interaction view and the Experience Score chip, and lets you filter interactions by them.

What’s new

Interactions are labelled across three Experience Driver categories: Product/Process, Bot Handling and Agent Handling. Each one is rated positive, negative or neutral, with a one-sentence explanation of why. If Bot Handling is negative, for example, you might read that the bot misunderstood the customer’s question, looped them through the menu, and they left without a resolution.

These ratings are already live in Ask Spotlight. It can look up the drivers for a single interaction or aggregate them across a group, so you can ask questions like “What drove negative experiences for billing contacts last week?” or build a report around them.

This release puts the same ratings and explanations into the interface in two places, and adds a filter for each driver:

• The interaction itself: a new Experience Drivers section sits between Summary and Details in the right-hand panel. It’s collapsible, independently of Summary, and lists the three drivers in the same order every time with a coloured rating chip and the explanation underneath.
• The Experience Score chip: hover over it in an interactions table and a popover shows the score with an overall label (Mostly Positive, Mixed or Mostly Negative), plus each driver’s rating and explanation.
• Experience Driver filters: the + filter menu has a new Experience Drivers category with three filters: Product/Process, Bot Handling and Agent Handling. For each one, pick any mix of Negative, Neutral and Positive. You can apply all three at once, so you can pull up, say, every interaction where Bot Handling was negative but Agent Handling was positive, then run Spotlight on that set to see what’s behind it.

Where to find it

Interaction Modal

1. Open any interaction from Explore, Customer Overview or Agent Activity, or open it on the Review screen.
2. Look for the Experience Drivers section in the right-hand panel, below Summary, and expand it.

On Hover

1. To check drivers without opening an interaction, hover over the Experience Score chip in the interactions table on Explore, Agent Activity or Customer Overview

Filter

1. To filter by driver, click + in the filter bar on any page except Reviews, open Experience Drivers and choose Product/Process, Bot Handling or Agent Handling. Move the ratings you want into Selected and click Done. Add the other drivers the same way if you need them.

To look across many interactions at once, use Ask Spotlight, which already supports this. Ask about Experience Drivers directly, for example which driver is behind a drop in CSAT for a contact reason.

There’s no new permission for this. Anyone who can open an interaction can see its Experience Drivers, the filters are there for anyone who can already use the filter bar on those screens, and asking about them in Ask Spotlight uses your existing Ask Spotlight access.

What this looks like in practice

• A QA lead opens an interaction before reading the transcript and sees Bot Handling marked negative because the customer got stuck in a menu loop. She knows where to look before she’s read a message.
• Scanning the interactions table on Explore, a contact centre manager spots a low Experience Score and hovers over it. The popover shows Product/Process as the negative driver, with an explanation pointing to a failed refund, while Agent Handling is positive.
• A support operations manager filters Explore to Bot Handling: Negative for the last fortnight and runs Spotlight on the results. Most of them trace back to the same unrecognised intent, which goes straight onto the bot team’s backlog.
• During a review on the Reviews screen, a team lead uses the Agent Handling rationale as a starting point for coaching feedback.
• After CSAT drops for one contact reason, a CX lead asks Ask Spotlight which driver was most often negative across those interactions. The answer points to Product/Process, so she takes it to the product team.

Good to know

• Driver ratings only exist for interactions from when labelling launched. Older interactions weren’t backfilled, so they show “No Experience Drivers applied to this interaction” and aren’t included when Ask Spotlight answers questions about drivers.
• If labelling fails for an interaction, the section shows an error message in its place.
• Picking more than one rating for a driver matches any of them (Negative or Neutral, for example). Combining drivers only returns interactions that match all of them.
• Ratings and explanations are AI-generated. They’re calculated separately from the Experience Score, though the two usually line up, and other metrics also feed into the overall score.

We’re always improving our AI summaries. If you have questions or feedback about your experience driver summaries and how they could be improved, reach out to your Customer Success contact.