Hello Reader

You’ve filled the pipeline, but the deals aren’t closing. Now the CRO is on your case about lead quality, and you’re back defending marketing’s contribution.

Do you change the targeting?

Pour more budget into the top of the funnel?

Before you do either, find out where those buyers are getting stuck.

You need to know what happens between them showing interest and deciding whether to buy.

​Lauren McHugh, Head of Marketing at Docket, says that surfacing buyer objections has helped her team build more effective offers and move pipeline faster.

"I spent 15+ years in product marketing, so ICP and understanding the buyer are basically in my DNA,” she says. “What's changed in the past 10 months though is how I get the information. I stopped digging into buyer objections in surveys and rep anecdotes, and instead rely on Docket's AI agent on our website.”

The Docket agent engages prospects in real time with a live conversation to qualify their intent.

“Buyers spend an average of 18 minutes talking to it and they tell our agent things they'd never say to a salesperson on a first call,” Lauren says. “So I use AI to sort through those conversations and compare them with what's coming up on sales calls. That's how we figure out which objections are just noise and which ones actually stall deals, then we build offers around the real ones.”

If you don’t have your own website agent you can also use call transcripts, feedback surveys and CRM notes for the analysis. AI can help you work through that material and identify the questions buyers still need answered.

I ran this exercise on my own sales conversations and it gave me a much more specific set of problems to address than “we need better leads.”

Help buyers make a decision

As marketers, we don’t always control the product or the price, but we can change how buyers evaluate what we sell, and get the information they need to make a decision.

That starts with understanding the decision from their point of view:

1. What prompted the buyer to consider a change?
2. What problem are they trying to solve, and why does it matter now?
3. What are they comparing?
4. What will determine their choice?
5. Who needs to approve the decision?
6. What is preventing the next step?

Take a hypothetical reporting-software company. A prospect likes the demo, but needs to know whether the product works with their data and how much implementation will demand from their team.

The solutions consultant shows them the general dashboard. Their questions about connecting their CRM and the work required from IT remain unanswered.

Marketing could work with the solutions consultant to create a walkthrough of the CRM connection and an implementation plan showing the tasks, responsibilities and estimated time required from the buyer’s team.

The product is the same, but the buyer gets a clearer picture for deciding whether to proceed.

Understanding how a team works can also help you ask more specific questions about implementation.

​Rachel Fields, VP Marketing at Section, described how her team uses agents to unearth buyer pain points:

“Instead of the typical process, which is time and cost-intensive, we built agents that can develop a workflow map of any team within a few days,” she says. “We pressure test to identify the highest pain points which helps us create a better program for the client.”

What showed up in my own sales process

In my recent homepage review, I looked at whether people understood what I was selling.

Spoiler: they didn’t.

With that fixed, I turned my attention to the sales cycle.

I asked AI to examine 12 months of call transcripts and prospect feedback to identify where buyers hesitated, what they still needed to decide and what happened next.

Four patterns were affecting my closed-won deals:

1. Substantial scoping or evaluation work remained after the sales call. Deals took longer to progress because the buyer still needed to define and assess the solution.
2. Budget or price constraints. Even when there was interest, limited or uncertain budget made commitment harder.
3. Internal approval was required. Additional stakeholders and approval steps were slowing momentum.
4. IT, governance, or tool-readiness issues. Implementation concerns created friction and required additional preparation.

​Charlie Treadwell, CMO at Elisity, used Claude Code to conduct a similar review. He investigated Salesforce opportunities marked “closed, no decision.” His evaluation showed how the recorded reason for a stalled deal can leave out important details.

“The CRM entry might say ‘no budget,’ but the underlying conversations can tell a different story,” he says. “The sales team never reached the decision-maker or economic buyer.”

Because Gong, Slack, Salesforce and SharePoint are connected to the Elisity Claude environment, Charlie’s team can analyze deals at a level of detail that would be difficult to reconstruct manually.

“Claude reviews everything and then builds a timeline showing what happened, which concerns surfaced and where the process broke down.”

Charlie’s team is using the analysis to involve decision-makers earlier, check whether budget concerns have been confirmed and connect technical validation to a buying decision.

Run the same assessment on your sales history

Okay, so you want to run this analysis on your own pipeline. Where do you start?

First, choose the product or offer and buyer group you want to examine.

If you’re using one of the AI tools (like ChatGPT or Claude) you can connect it to your outside platforms. That means connecting your sales call transcriber (like Gong) and CRM (like Hubspot) and ask it to retrieve the relevant records.

Make sure you specify the sources and date range.

If you don’t have connected platforms you’ll need to attach the call transcripts and any other buyer feedback you’ve got handy.

Second, include the proposals, product descriptions and pricing information. Include won deals alongside lost, open and no-decision opportunities.

Third, ask ChatGPT or Claude to run the analysis and report on any obvious patterns.

I always ask for this in a table so it’s easy to scan:

Once you have your analysis:

1. Check the diagnosis. Read the source quotes behind each conclusion. Confirm deal stages against the CRM and check whether ChatGPT has treated a salesperson’s interpretation as something the buyer said.
2. Choose one recurring obstacle. Look for a problem that appears across several relevant deals and affects the next step. Separate issues marketing can address from budget constraints, product gaps or procurement delays.
3. Confirm it with sales and buyers. Ask sales whether the pattern matches their experience. If your evidence is shaky, see if you can follow up with actual buyers. Ask them what they need to know before you can move forward.
4. Change how buyers evaluate the product. If buyers can’t explain the business case for your product internally, create a business case for them.
5. Test whether more buyers progress. Use the revised material with a set group of prospects and track their progress.

Let’s say your analysis shows that buyers repeatedly ask about price.

Ask ChatGPT to examine those conversations for more detail.

• What do buyers say about the cost?
• Does the concern appear across several deals?
• What happens after they raise it?

Ask for supporting quotes and explanations supported by the transcripts.

Suppose several buyers say they need to justify the purchase to their CFO, but struggle to explain the financial return. Check the interpretation with sales and, if you can actual buyers.

If that confirms the obstacle, create a business case template that helps buyers estimate the costs and expected return using their own numbers.

Work with sales to use the template in opportunities where financial approval is the next step. Track whether buyers get CFO approval.

You now have a specific marketing task and a way to assess whether it helps buyers proceed.

Here’s one of the insights from my own analysis:

At Amplitude, SVP of Marketing, Angela Ferrante, is using AI to get precise about buyer concerns and blockers.

"We test the buying experience, find patterns in buyer behavior, and pressure testing our content and collateral," she says."I have a few loops that run automated reviews, and it found points of friction in meeting booking."

Angela's team also built tools that analyze sites and content from different buyer personas, flagging things like unclear value, missing proof, and unanswered objections.

What I changed in my own sales process

After reviewing the analysis, I made four changes:

1. I asked prospects who was on the buying committee.
2. I defined the audience, scope, and expected outcomes.
3. I gave internal champions evidence and talking points.
4. I asked which requirements were essential for proceeding.

This gave my conversations more structure and made the unanswered questions easier to identify.

Tracking those questions alongside deal outcomes will help me assess whether the changes improve progression.

Work with sales, product and delivery teams

Lauren emphasizes that addressing these friction points requires breaking out of the marketing silo.

“I talk to sales to identify if they're hearing the same thing, dig in with customer success to hear how it’s working for our customers, and product gets all of it, because half the time a buyer's question is really a feature request.”

Rachel has a great example of cross-department collaboration. The Section team built an agent that joins calls—with permission—to record and score conversations using a BANT (budget, authority, need, timing) rubric.

“The agent then recommends follow-up questions and next steps to keep the deal moving forward,” Rachel says. “It made our calls much more efficient and valuable for the client.”

To bridge the gap between marketing, sales, and product, consider these steps:

• Ask sales to review the buyer quotes in the analysis. Do they agree that you’ve identified a current obstacle?
• Ask product to verify any capability claims you make.
• Work with implementation teams to confirm the steps, responsibilities and effort required from the buyer.

Test any changes you make with prospects who face the obstacle you identified. Check whether it helps them take the next step.

If the same conversations continue to stall, review the diagnosis and the response with sales and keep iterating.

Every improvement you make will drive pipeline faster.

Use this prompt to assess your buyer journey

Attach your sales cycle material or identify the connected sources you want AI to use.

Include the product or offer documents, and specify the buyer group and date range.

Here's the prompt:

Analyze my call transcripts and prospect feedback from the last 12 months to identify what is preventing buyers from taking the next step.

This is a diagnosis exercise. Do not design a new offer or write promotional copy yet. Do not assume that a different offer would solve the problem.

SCOPE

Use the offer, buyer group and date range I specify alongside this prompt. If I have not specified them, inspect the materials first and ask up to three questions only where the missing information would materially change the analysis.

1. ESTABLISH WHAT YOU HAVE

Read the available material and create a source inventory. Record the source type, date, offer discussed, buyer role, opportunity identifier, stage and outcome where available. Use “unknown” for missing details.

Separate:

- Buyer statements in sales conversations.

- Seller interpretations, CRM notes and AI-generated summaries.

- Prospect research or survey responses.

- Feedback from customers after purchase.

- Product, service and commercial documentation.

Group repeated calls and notes about the same buying opportunity. Count an obstacle once per opportunity when reporting frequency, while preserving the different stakeholders’ perspectives.

2. TRACE THE BUYING DECISION

For each opportunity, identify what the buyer wanted to accomplish, the next step being considered, their questions or concerns, my response and any documented action afterward.

Identify decision-makers or approvers mentioned but absent from the conversations. Do not assume the contact speaks for them.

Keep the stage and chronology clear.

Distinguish:

- A question the buyer asked.

- A concern the buyer explicitly connected to proceeding.

- An observed delay or action.

- Your interpretation of why it happened.

Compare concerns across won, lost, no-decision and still-open opportunities when outcomes are known. Note concerns that successful buyers also raised, whether they were resolved and what evidence supports that conclusion. Do not infer that a particular answer caused a win.

3. ASSESS WHAT KIND OF PROBLEM IT IS

Build the categories from the evidence. Consider whether the issue involves understanding, proof, evaluation effort, implementation, internal agreement, commercial terms, budget, timing, seller follow-through or genuine product/service fit.

Use product or offer documents to check what was promised and what could be delivered (attached).

Look for contradictory evidence and cases where your leading explanation does not fit.

4. RETURN THE ANALYSIS

A. Scope and limitations

Summarize sources reviewed, distinct opportunities, date range, outcome coverage and important gaps. Explain which groups can reasonably be compared.

B. Pattern table

Show up to five supported buying obstacles. For each, give:

- Buyer group and stage.

- Number of distinct opportunities mentioning it, with a clear denominator.

- Known outcome mix.

- Supporting source references.

- Evidence strength: strong, moderate or weak, with a brief reason.

Before returning the answer, check that every major conclusion has source support, counts refer to distinct opportunities and uncertainty remains visible.

I'll walk through the settings, connectors and skills that get Claude working from your brand and audience.

Then I'll set up workflows that take over your weekly research and reporting, and show you how to build a market intelligence agent that tracks your competitors.

Join me on Wednesday, Oct 21 at 12pm PT for the free 60-minute session.

Can't make it live? Register anyway and the recording comes to you.

​Save my free seat​