29 hours of talk time

Your cold call recordings are a goldmine.

And there’s way more there than you have investigated.

I have long dreamed of building a system for “call science” or “conversation science” that analyzes cold calls for true top of funnel insight.

Why?

• Conversations with customers and demo requests/pipeline contacts are a biased sample: they’re already interested. Everyone has those tools - but still fail at creating pipeline.
• A weird amount of messaging strategy and product insights are locked up in these brief-but-poignant conversations.
• The best reps get prospects talking (not too much, just enough) which means they share emotional motivations and meaningful objections.

All of this fuels a robust nurture strategy in both sales follow-up and marketing retargeting.

Thanks to AI and calling tech - this is now really easy to do.

For instance, here’s the pitch that worked most often in September:

You need to analyze your cold calls. Now more than ever.

29 hours of talk time later…

Last month we had our two new SDRs start out here at TitanX.

Across 772 conversations in September with 502 companies, they accrued 29 hours of live conversation with our ICP.

It is hard to articulate the full scale of the opportunity that presents.

First, the dispositions themselves fuel our SDRs’ ability to follow up and nurture the right contacts into pipeline.

Second, the conversations and their dispositions serve as triggers for marketing. I’m building re-engagement, retargeting, and nurture workflows off of a live feed of call results.

Third, the conversations represent a dataset ripe for analysis. See the list of objections we got below (thanks Attive.io for pulling this in 3 minutes).

#3 confirms something we’ve seen other places in our marketing: many people still confuse TitanX with a dialer. This informs what actions I need to take as a marketer to publicly counter that confusion (e.g. a tech stack diagram where TitanX is squarely not in the dialer spot).

Make Conversations Your Prime Signal

Conversations with target accounts are the single best signal you could possibly invest in.

No amount of scraping and AI will replace a few key questions to a couple key people in your target accounts.

So:

Present the case for call science to make the case for your headcount.

Argue for human-driven research in an era where most are relying solely on AI. We use AI to research accounts, quite a lot in fact. But we make sure to get a conversation with our ICP.

Nothing beats talk time.

After all, B2B lives or dies off of conversations. Beyond the tech, the conversations themselves ARE the system you’ve been needing this whole time.

However you can create more conversations and analyze them for insight, this is how you’ll win.

This includes AE cold call talk time too, an additional 29 hours

You’re Probably Only ⅓ of the Way to Analyzing Your Calls

There are 3 layers of analysis, and almost everyone is only partially doing #1:

1. Disposition Science - where the breakdown of call results leads to further action within the calling team
2. Marketing Triggers - where the disposition leads to certain types of marketing activity. Think of this like Marketing Nurture but better.
3. Call Science - where you analyze groups of calls from all time, or specific windows of time, or lists/campaigns, or reps, and determine things like rep coaching opportunities, marketing content, and product strategy.

Going deeper into a couple of those Call Science tactics:

• Marketing Content: I actively analyze our cold calls for objections and ideas for social content. And there’s so much more that we will be doing, especially around per-account custom assets.
• Product Strategy: every cold call and demo call feeds a Google Sheet with product feature flags and insights requested by our prospects.

Here’s a visual breakdown of the three layers.

Hit reply if you want to know how to get this going in your stack (I read every reply). I mentioned Attive.io which is a great solution if your calls are in Gong.

For the tech-curious: Last December I was building connectors from Phoneburner to Clay to Pipedream to ChatGPT to analyze calls and produce CRM-ready values about competition, product insights, and account research.

Then this Spring I was building it with AWS Bedrock to automatically “vectorize” cold call recordings and layer an LLM on top. (Seriously, overkill, but it was that or turn each transcript into a PDF and load it into a pre-built LLM tool that never scanned all of them - sadly, thousands of cold calls quickly use up all the context on a GPT or Gem or Claude Project).

Fortunately you shouldn’t have to do anything drastic. The trick is to internalize this holistic “conversation science” as an ongoing discipline. Something we’re launching here and are very serious about standardizing for the industry.

Check out Joey’s recent post - a deeper look at Disposition Science:

Thanks for reading,

Evan Dunn (LinkedIn)