Your Head of Sales asks: “What’s the average deal size for accounts that use Feature X?”
Reasonable question. Should take 30 seconds to answer. Instead, it takes three days.
Because answering requires pulling product usage from Mixpanel, exporting deal data from Salesforce, matching accounts between systems (hoping the company names align), calculating averages in a spreadsheet, and sending the findings over Slack.
Or, you could ask your data team to build a dashboard for it. Which they’ll add to their backlog. Behind 47 other dashboard requests. Available in two to four weeks.
By then, nobody remembers why they needed the answer.
The Dashboard Trap
We built dashboards to democratize data. Make it accessible to non-technical people. Enable “data-driven decisions.”
It worked, for a narrow set of questions. The 12 specific things someone thought to ask when the dashboard was built? Those are covered. Everything else requires either a new dashboard or a data request.
The promise was self-service analytics. The reality is self-service for the 12 things you pre-built. Everything else is gated behind your data team’s calendar.
And the kicker: the questions that matter most are rarely the ones you anticipated. The most valuable customer insights come from ad-hoc questions. From following a thread. From someone noticing something weird and asking “why?”
Dashboards are great at answering the same question repeatedly. They’re terrible at answering the question nobody thought to ask.
What Conversational Analytics Looks Like
Let’s replay that original question.
Your Head of Sales asks: “What’s the average deal size for accounts that use Feature X?”
She types it into your AI customer workspace. Eight seconds later, the answer appears: $47,200 average deal size for Feature X users, compared to $31,800 for non-users. A 48% premium.
She pauses. That’s interesting. She asks a follow-up: “Is that because bigger companies use Feature X, or does Feature X actually drive larger deals?”
The AI customer workspace breaks it down by company size. Turns out, even controlling for company size, Feature X users close deals 23% larger. Something about the feature changes the sales conversation.
She asks one more: “What percentage of our pipeline right now has Feature X enabled in their trial?” Answer: 34%. She makes a note to push Feature X enablement in all active trials.
