Every revenue leader has a version of the same story.

A six-figure deal at a target account. The rep needs to reach the CFO. There is, sitting in the company's network, a clean path: a board member who worked with that CFO at Cisco for four years. Strength score: 91 out of 100.

The intro never gets asked for.

The deal closes with someone else.

This is not a tooling problem. The path was identified. The relationship was real. The rep had access to the dashboard. The board member would, in fact, have said yes. The deal was lost in a single, silent, three-second moment inside the rep's head: "Can I really go to a board member for this? Will he think I can't close on my own?"

The cold outbound industry exists because the warm channel is socially expensive, not because cold works better. It works worse. By every measurable input, warm introductions convert 3-5x higher and produce 25% higher win rates than cold outreach. The reason the channel is underused is psychological, not statistical.

This is what we call the awkward ask. It is the single largest tax on B2B revenue that no GTM stack currently addresses.

What is the awkward ask?

The awkward ask is the social and psychological friction that prevents three different humans from completing a warm-introduction transaction:

1. The rep, who would have to admit they need help to close.
2. The connector, who would have to decide whether the ask is worth their relationship capital.
3. The champion, who would have to become, momentarily, a salesperson for someone else.

Every warm-intro program in existence fails one or more of these humans. The good news is that none of them mean no. They mean "yes if it were easier, lower-risk, and didn't make me look bad."

The job of relationship software is not to find the path. The path was findable on LinkedIn ten years ago. The job is to take the awkward part off the human's plate.

Persona 1: The rep, in their head

Reps don't ask because asking signals weakness.

The internal monologue is usually some variant of:

• "Will my manager think I can't close on my own?"
• "If I tap a board member for a $200K deal and lose it, that's worse than not having tapped them at all."
• "I asked my CEO once already this quarter. Two is too many."
• "What if the connector doesn't actually know this person well? Then I've burned a chip for nothing."

There's a deeper version of this for senior reps. The longer you've been in sales, the more your identity is built around being someone who makes their own pipeline. Asking for intros, in that frame, feels like admitting you're not really the closer you said you were.

The result: reps systematically under-use the channel that converts highest. They optimize for activity that signals effort (dialing, sequencing, posting on LinkedIn) over activity that signals dependency (asking for help). Activity that protects the rep's ego beats activity that produces revenue.

Most managers know this and complain about it. Almost none of them have changed it, because the fix is not coaching. The fix is removing the moment of asking entirely.

Persona 2: The super connector, in their head

The board member, advisor, or investor sitting on the other side of the path has their own monologue.

• "Do I actually know this person well enough to vouch?"
• "If I introduce them and the company turns out to be wrong for this prospect, I'm the one who looks bad."
• "I'm getting asked for a lot of intros. Each one costs me a small amount of relationship capital. Am I being a good steward of that?"
• "What's the right way to write this email? I keep meaning to and then I don't."

The hidden truth about super connectors is that the constraint is almost never willingness. It's cognitive load. Connecting two parties well takes 15 minutes of context-switching, judgment, and writing. Multiply that by 20 asks in a quarter and you have a 5-hour tax that does not show up on anyone's calendar.

When a super connector "ghosts" an intro request, what they almost never mean is "I don't want to help." What they mean is "I started to think about this, didn't have time to do it well, and didn't want to do it badly."

Most warm-intro programs ask connectors to do more. That's exactly backwards. The right design asks connectors for less: one click of approval on a pre-drafted note, on a request that has already been screened against their preferences (deal size, channel, cadence, type of ask).

Persona 3: The customer champion, in their head

Customer champions are the trickiest of the three, because they're often friends with the rep or the company. Friendship is the asset and the liability.

• "I love these guys, but I don't want to be their salesperson."
• "What if I make this intro and my friend at the other company feels like I'm using them?"
• "Am I going to get asked again next quarter? And the quarter after that?"
• "If I help with this, will it look weird in front of my own team that I'm advocating for a vendor?"

The champion's currency is reputation, and reputation compounds. They will go above and beyond for the right ask at the right moment. They will absolutely not be your default outbound channel.

The right ask, for a customer champion, has three properties:

1. It is specific. Not "anyone you know at Snowflake," but "Sarah Chen, VP Eng at Snowflake, who you worked with at Looker."
2. It is timed to a positive trigger. The week after a great QBR. The day after a product win. Not in the middle of an open support ticket.
3. It makes the champion look good to the person being introduced to. The pitch is framed so that the champion is sharing something valuable, not pushing a vendor.

Most CRM workflows are blind to all three. They surface contacts based on relationship strength scores and let the rep figure out the rest. The rep, predictably, asks badly or doesn't ask at all.

The cost of the awkward ask, quantified

We've started to track this across customers. The numbers are striking.

In a typical B2B company:

• The relationship paths exist for roughly 60-70% of target accounts in the ICP. (Run a network analysis on any 500-person company and this is what you find.)
• Reps initiate a warm-intro ask on fewer than 10% of those addressable accounts. The other 50-60% of paths sit unused.
• Of the asks that do get made, roughly half stall at the connector stage. The connector either ghosts, replies "let me think about it," or asks for more context that never comes.
• Of the asks that complete, conversion to a booked meeting is in the 60-80% range, often higher when the connector is a board member or customer.

Multiply those numbers out. The pipeline you have access to is roughly 5-10x larger than the pipeline you are using. The blocker is not your CRM. The blocker is the awkward ask.

This is the part most "relationship intelligence" tools miss. They are designed for the 10% of reps who would have asked anyway. They make those reps slightly faster. They do not change the behavior of the 90% who don't ask.

Why an agent does this better than a human

The reason agents are the right architecture for this problem (and the reason "more dashboards" is the wrong one) comes down to a single property: agents don't experience the awkward ask.

An agent will:

• Read the call transcripts and CRM history and surface the right path without the rep having to admit they need help.
• Compose the intro request in a tone the connector will actually forward, without the rep having to find the words.
• Time the ask to a positive trigger (a QBR, a press hit, a product launch) the rep would not have spotted.
• Frame the ask to make the connector look good, not to make the rep look needy.
• Escalate quietly to the manager when a deal has a clean path and the rep has gone silent for five days, without the rep losing face.
• Close the loop with the connector when the intro produces revenue, so the next ask is easier.

None of those are technically hard. The human equivalents are not technically hard either. The reason humans don't do them consistently is that each one is socially expensive. Agents pay that cost in compute, not in ego.

This is also why the future of this category does not look like a better CRM view. It looks like a teammate sitting in Slack and Salesforce who runs the entire motion, end to end. We call ours Rudy. Whatever you call yours, the design principle is the same: the human should never have to do the part that makes them flinch.

What this means for revenue leaders

If you're running a GTM team, three implications follow.

One. Stop coaching reps to ask more. It doesn't work. You've tried. The behavior is too well-defended by the rep's identity. Build a system that removes the moment of asking.

Two. Stop measuring "intros requested" as the activity metric. That metric rewards the reps who were already going to ask. The leading indicator you actually want is intros that should have been requested and weren't, surfaced before the deal stalls. That number is invisible without instrumentation, and it is almost always large.

Three. Treat super-connector experience as a product surface. Your board members, advisors, and customer champions interact with your GTM motion. That interaction is currently designed by reps, ad hoc, badly. Design it on purpose. Give connectors one channel, clear preferences, pre-drafted asks, and closure messages when their intros produce revenue. The goal is that being a connector for your company feels good, not extractive.

The companies that get this right in the next two years will quietly compound an unfair advantage. Their network compounds. Their reps stop dialing strangers. Their CACs collapse. Their win rates climb. And their competitors keep trying to fix outbound.

The warm channel was never broken. The asking was.

Want to see what removing the awkward ask looks like in practice? See Rudy in action from intro request to meeting booked, in under 15 minutes.

‍