Brian LaManna shows how to fix a company losing 70% of early-stage deals using Gong's AI Theme Spotter and AI Builder. Theme Spotter surfaces the most common objections in lost deals, then AI Builder generates a playbook from won deals with real call examples and discovery questions that reps can actually use.
Chapters
Anonymous customer story here. One of my customers was losing over 70% of early-stage deals before they got to Stage 3. They were losing them between Stage 0 and Stage 1 and Stage 1 and Stage 2. What they tried doing in the past prior to Gong is they asked reps or they dug into CRM. And when they asked reps, they got a myriad of reasons. Everyone had a different opinion. Some reps were very noisy and loud about things like product issues, or they were bad leads from marketing. And when they went into CRM, they noticed the data was even more scarce. It was only what reps decided to input. It was all from their side versus the customer side. So here's how they use Gong to identify why they're losing deals early stage, and then more importantly, creating the playbook around how to fix some of this and get it moving in the right direction. So within Gong here, you're able to come into this Insights tab and AI Theme Spotter. This is where you can ask any question across your entire Gong dataset. So as an example here, you can come into new analysis, you can ask or create any new question. The key is really adding these filters here. So if you're only taking a look at, say, early stage conversion, we might want to use a deal stage filter to say, hey, we only want to take a look at stage Stage 1 and Stage 2. Oops. We'll do Stage 1 and Stage 2, where the deal ended up turning closed lost. So deal stage now, turn closed lost, where during the call it was in Stage 1 or Stage 2. You would hit this save filter here so you can really drill it down. Then you can ask any question across your dataset, right? And this is all calls, all emails, all web meetings, all accounts, all opportunities. It's all that rich data Gong's able to ingest. We're then able to run that across their entire dataset. And they could actually pinpoint which objections were coming up and why they're losing those early-stage deals, as well as the total amount impacted from them. So they could really get to the root of which things happening were impacting the most early-stage deals. From there, they came into AI Builder, right? Now that they know why they're losing those deals early stage, they can create a playbook around how to fix it, right? And what top-performing reps were doing that were a little bit different in the early stage. So what AI Builder allows you to do is to create out any content, things like a playbook or a document based off your Gong data. So instead here, they came in and they hit Add Sources. What they did is they came in and decided, hey, we want to take a look at early stage again, deal stage during call, stage 1, stage 2. But but we want to take a look at deals that actually ended up turning closed won, right? And the different commonalities that we're seeing in those deals versus those that happen closed lost. Down here, they entered their prompts. So now that they're filtering off the right data, early-stage deals that were successful, they could come in here and say, hey, create a playbook around which objections came up, how to handle them, you know, an XYZ. From there, you hit this Go button, and I'll show you what the end output of this actually looks like. Beautiful. So on the left here was the prompt, on the right, early stage discovery objection handling playbook. It's going to break out the core principles, the most common objections that we just learned from that AI Theme Spotter, real examples out in the wild. And then giving you an actual framework on how to overcome that and certain discovery questions to ask to get things back on track after you're able to address the objection. Now, my favorite part with this playbook is this would have taken like probably hundreds of hours to create. It also wouldn't have been very data-backed from like which ones were actually coming up the most. And then within them, it's gonna link out to specific calls and specific snippets. So it has the context of the accounts that happen, and it's gonna give our reps real-life scenarios that they can then listen to and say, hey, I'm getting this objection all the time around compliance and security. Let me hear how the top performers were actually discussing it in step 1, in step 2, and they can hear how the rep positioned it and then voice of customer directly after. So it's a really pointed way of learning. Think about trying to do this without Gong. Again, you'd either try to go into CRM, try to go into reps, maybe you export 5 or 10 calls, you push them to Claude or ChatGPT. They're still going to be missing the actual context, giving you real-life scenarios, real linked to the calls and the context underneath. And you're only gonna be able to export so many call transcripts in an easy manner versus this AI builder, you can run across your entire dataset, think tens or hundreds of calls, just make sure you really dial in those filters here and these sources so that you're running them off the right dataset to create that end output. This AI builder has been immensely helpful for enablement teams and being able to really bridge the gap between strategy an execution.
