Hey everyone! 👋 I've been experimenting with Read AI's API for the last few weeks, and I wanted to share a practical use case I built: a performance leaderboard for my sales development reps (SDRs). We were already using Read for meeting analytics, but I knew the data could be put to work for motivation and coaching.
The goal was to pull key metrics from our recorded sales calls and demos, then display them on a simple internal dashboard. I focused on a few data points Read provides that are gold for SDRs:
* **Talk/Listen Ratio:** Encourages balanced conversations.
* **Questions Per Minute:** A good proxy for engagement.
* **Sentiment Score:** Helps flag potentially frustrating calls for follow-up.
* **Overall Meeting Score:** A nice, quick summary metric.
I used the `/analytics/summary` endpoint as my main data source, scheduling a nightly fetch for all new meetings tagged with "SDR" or "Qualification." The API response is pretty comprehensive, so I had to map and transform a few fields to fit our scoring model. Here's a quick look at the logic I applied for the leaderboard score:
`Leaderboard Score = (Meeting Score * 0.4) + (Questions Per Minute * 20) + (Positive Sentiment % * 0.3)`
The biggest lift was connecting this data to our CRM (HubSpot) to tie each meeting to a specific rep and opportunity stage. Once that was linked, I used a simple front-end to display a ranked list, along with weekly trends for each rep. The impact was immediate—it sparked some healthy competition and gave managers concrete talking points.
A couple of things to watch out for:
* The API has rate limits, so batching your calls is a must.
* You'll need a unique meeting identifier to join data with your own systems.
* The sentiment labels are helpful, but I'd recommend combining them with manual spot-checks for nuance.
Has anyone else built something similar with Read or another conversation intelligence platform? I'm curious how you're weighting different metrics for sales roles. I'm thinking of adding "Client Speaking Time" as a new KPI next.
Cheers,
Henry
Cheers, Henry