# The Anatomy of a Broken B2B Sales Pipeline: Where Yours Is Bleeding

By Vara Tech · Sales Operations · November 9, 2026 · 10 min read

Your dashboard says the pipeline is worth ₹4 crore. Your bank balance says the quarter is going to close short. Both are telling the truth.

That's the strange part. The CRM isn't lying. Every deal is in there with a stage, an amount and a close date. The money just isn't arriving, and every person on your team gives you a different reason why.

When two numbers that should agree don't, something is happening between them that nobody is looking at. That is what a broken B2B sales pipeline looks like from the CEO's chair, and it's where this investigation starts. Your CRM is the evidence. By the end you'll know which of three conditions you're dealing with, which numbers prove it, and which break to fix first.

No detectives, no drama. Just your own data.

## The suspect lineup

A pipeline isn't one thing. It's a system with stages, and it can fail at a different stage for a different reason. Most B2B sales pipeline problems fall into one of three conditions: starved, leaking or clogged. Each has its own symptom, its own cause and its own fix. And one pipeline can suffer from all three at once.

Start with what you hear inside your company.

- **Starved:** "We never have enough meetings." Not enough of the right people are going in.
- **Leaking:** "People reply, then nothing happens." Interested prospects drop out after the first reply.
- **Clogged:** "We have loads of deals, but nothing closes." Deals sit in the pipeline without moving.

Starved doesn't always mean too few leads. It can mean plenty of activity aimed at the wrong people. Leaking means interest arrives and then disappears. Clogged means deals look alive in the CRM but aren't moving.

### Three-question quick check

Work down the list and stop at the first yes.

1. Are you generating plenty of outreach but getting very few meaningful responses? That points to a **starved** pipeline.
2. Are prospects responding with interest but failing to progress? That points to a **leaking** pipeline.
3. Do you have many opportunities sitting in your CRM without meaningful progress? That points to a **clogged** pipeline.

This is a starting point, not an audit. Several yeses are common. The reason to stop at the first one is that the earliest break matters most: fixing a later problem may not solve an earlier one. The next three case files take each suspect in turn, and the five-number X-ray after them gives you evidence instead of impressions.

## Case file 01: Starved

**The symptom.** "We never have enough meetings."

**What it looks like.** The team is busy. Sequences are running, calls are being made, the activity report looks healthy. But the conversations that come back don't sound like your best customers. Replies say "not relevant." The person who answers can't buy. The company is the wrong size.

**The underlying problem.** A starved pipeline isn't always short of leads. It's short of the right ones. The usual cause is simple: nobody has written down who can actually become a customer. The ideal customer lives in the founder's head, shifts a little whenever someone new joins, and outreach goes out without a clear reason to reach out now. More volume doesn't fix that. It gives you more activity pointed at the same wrong targets.

**The evidence.** One example from our own work:

> **Evidence file: deep-tech IoT campaign (Vara Tech).** We ran a 30-day outbound campaign for a deep-tech IoT platform with no outbound history. It reached 72 founders and CEOs with 711 emails, held a 41.4% open rate, and booked 3 qualified meetings averaging INR 20 lakhs each.

Limits: one campaign, no comparison group, and open rates are a soft signal, so the 3 meetings matter more than the 41.4%. It shows what a small, carefully chosen list can do. It doesn't prove that narrower always beats wider.

**Our read.** Seventy-two people is not a big list. But a short list chosen for a reason is one a team can follow up properly, which matters in the next case file.

**Run the test.** Pull up your last ten wins. Write down what they share: the industry, the company size, what was happening in the business when they bought, the job title of the person who said yes, and the problem they were trying to fix. Then look at the last twenty people your team contacted. How many fit the pattern? If writing the pattern down takes more than five minutes, or your team would give you five different answers, you've found the starvation. Our guide to [building a B2B ICP that actually works](/blog/build-b2b-icp-signal-based-targeting) goes deeper.

**The fix.** Go narrower before you go louder. Choose one segment, one role and one reason to reach out now, and rebuild the list around it. Raise volume only after that list produces conversations that sound like your best customers.

*Next suspect: suppose targeting is fine and the right people do reply. That's where the next problem hides.*

## Case file 02: Leaking

**The symptom.** "People reply, then nothing happens."

**What it looks like.** On Thursday a prospect writes back: "Interested, please send details." Your salesperson is on a call, then travelling, then it's Monday. The follow-up goes out late, or not at all. By then the prospect's attention has moved elsewhere. Nobody decided to drop it. There was no alarm and no error message. It just didn't happen. This is sales pipeline leakage at its simplest.

**The underlying problem.** Interest alone doesn't move a deal. It has to be turned into a conversation, and that takes a process: someone who owns the reply, a clear next action, and a way to see what hasn't been answered. When follow-up depends on someone remembering, it fails the first week things get busy.

**The evidence.** Two studies are worth knowing.

> **Evidence file: response time (Harvard Business Review, 2011).** Researchers James Oldroyd, Kristina McElheran and David Elkington sent a web-generated test lead to 2,241 US companies and timed the reply. 37% responded within an hour, 16% within one to 24 hours, and 24% took longer than a day. 23% never responded. Among companies that did reply within 30 days, the average response took 42 hours. A companion analysis of 1.25 million leads from 29 B2C and 13 B2B US companies found that contacting a lead within an hour made a company about seven times more likely to have a meaningful conversation with a decision-maker than waiting longer than an hour.

| Response time | Share of companies |
| --- | --- |
| Within an hour | 37% |
| Within one to 24 hours | 16% |
| Longer than a day | 24% |
| Never responded | 23% |

Limits: the data is more than a decade old, covers US companies and web enquiries rather than replies to outbound email, and measures the first response to a new lead. Treat it as a direction, not a rule.

> **Evidence file: selling time (Salesforce State of Sales, 7th edition, 2026).** Salesforce surveyed 4,050 sales professionals in 22 countries, including India, in August and September 2025. On average they said about 40% of their week goes to selling and 60% to everything else, including planning (16%) and manual data entry (13%). The report counts customer meetings and prospecting as selling.

Limits: these are self-reported estimates from a panel survey, and about 70% of respondents work at companies with more than 200 employees, so a smaller team's week may look different.

**Our read.** Speed matters, and a salesperson's week leaves little spare room. So a follow-up that relies on someone finding the time is fragile. That is our interpretation of the research, not something either study measured directly.

**Run the test.** Pull every interested reply from the last 30 days. For each one, note when it arrived and when your first real answer went out. A real answer moves things forward with a specific next step, not "will revert shortly." Count how many got one inside a business day.

**The fix.** Four parts:

1. **A response rule.** Every interested reply gets a real answer within one business day.
2. **A named owner.** One person, with a backup, not "the team."
3. **A daily unanswered-reply list.** Anyone can see what's waiting.
4. **A follow-up process.** Decide in advance what a good reply contains (the information requested, one question, a proposed next step) and what happens if the prospect goes quiet.

Some teams use AI to draft and queue these replies for a person to approve. That can help, but the rule, the owner and the list work with a spreadsheet too. See [what to automate in Zoho CRM and what should stay human](/blog/zoho-crm-sme-automation-human-balance).

*Next suspect: now suppose replies get answered quickly and meetings happen. The last problem is the hardest to spot, because it looks like success.*

## Case file 03: Clogged

**The symptom.** "We have loads of deals, but nothing closes."

**What it looks like.** Picture a proposal that went out 45 days ago, and nobody has replied. The close date has moved from this month to next month, then to the month after. The CRM says "negotiation." The buyer says nothing. No competitor has won the deal, so it still counts as pipeline.

**The underlying problem.** A deal can be open without being alive. Buyers postpone, champions go quiet, and without a defined next step nothing forces the deal to move or to end. The result is a big pipeline number made partly of opportunities that aren't realistically progressing.

**The evidence.**

> **Evidence file: no-decision losses (Dixon and McKenna, The JOLT Effect).** In spring 2020, Matthew Dixon and Ted McKenna analysed 2.5 million recorded sales conversations. They report that around 40% to 60% of lost deals ended in "no decision," that medium to high indecision showed up in 87% of deals, and that win rates were about 30% in calls with medium indecision against 6% with high indecision. Their explanation is that buyers fear making the wrong choice more than they fear missing out.

Limits: the findings come through a book and through Challenger, a sales-training firm, rather than a peer-reviewed journal, so sample details are less transparent than in academic work. They describe lost deals, not all deals, and the data comes from recorded sales calls.

> **Evidence file: slower decisions (Salesforce, 2026).** In the same survey, 57% of sales professionals said customers take longer to decide than they used to.

Limits: self-reported opinion from the same sample as above.

**Our read.** A deal with no competitor in sight isn't safe. Doing nothing can win. Every stalled deal is a loss that hasn't been recorded yet, and it inflates the number on your dashboard until someone looks.

**Run the test.** Open your CRM and sort the open deals by how long they've sat in their current stage. Then check five things:

- **Deal age:** how long has it been open?
- **Time in current stage:** has it moved recently?
- **Close-date changes:** how many times has the date been pushed?
- **Next step:** is there one, with a date?
- **Your normal sales cycle:** is the deal older than a typical win? (Your last ten wins will tell you the cycle.)

**The fix.** Good sales pipeline management here comes down to three habits:

1. **Every open deal gets a next step with a date.** No date, no pipeline. Move it to a parked list instead of deleting it.
2. **Review stale deals every week.** Decide: revive it, park it or close it. A forecast built on dead deals can't be trusted.
3. **Make the first commitment smaller.** If buyers fear choosing wrong, a pilot, a trial or a small first project can lower the risk of saying yes. It isn't guaranteed to improve closing, and it often means a smaller first deal, so test it on a few deals before changing your offer.

*Three suspects, one pipeline.* Now stop reading about other people's evidence and collect your own.

## The five-number X-ray

You can't investigate a pipeline from memory. Pull the last 90 days from your CRM and count five things. It's a simple sales pipeline analysis you can do in an afternoon.

1. **People you contacted.** Shows how much outreach went out, and whether those people look like your best customers.
2. **Interested replies.** Shows whether that outreach is producing real interest.
3. **Interested replies followed up within one business day.** Shows whether interest is being caught.
4. **Meetings held.** Shows whether interest is turning into conversations.
5. **Open deals that have sat in one stage longer than your normal sales cycle.** Shows whether deals are progressing.

Two cautions. These five counts aren't a funnel, so don't expect each to be smaller than the last. The fifth is a separate check on deals already in the pipeline. And a low number 2 can mean wrong targets, a weak message or emails landing in spam, so run the last-ten-wins test before you decide it's a targeting problem.

### How to read them

- **A targeting problem:** compare 1 and 2, then check whether the people who replied look like your last ten wins.
- **A follow-up problem:** compare 3 with 2. The gap is interested replies that didn't get an answer inside your rule.
- **A progression problem:** look at 5 against your total number of open deals, and at 4 to see whether interest becomes conversations. Twelve stuck deals means something different out of 14 open deals than out of 140.

### Illustrative example

These figures are an example, not a real company.

In this example, the team contacted 300 people, got 24 interested replies, followed up on 15 of them within one business day, held 9 meetings, and has 12 deals stuck beyond the normal sales cycle.

Here's the finding that matters: 15 of the 24 interested replies got a follow-up within one business day. That means 9 interested replies didn't meet the one-day rule. (The meeting count is also 9. That's a coincidence; the two numbers aren't related.) It's the earliest visible process break in this example, so it's where this team should start. The 12 stuck deals are a second problem, and the total open-deal count would show how big. The targeting question can't be answered from counts at all. It needs the last-ten-wins test.

## The pipeline autopsy

Now turn the numbers into a verdict. This is the report for the example above. It's illustrative, not a real client diagnosis.

- **Primary condition:** Leaking.
- **First point of failure:** between the interested reply and the follow-up.
- **Evidence:** 9 of 24 interested replies (about 38%) had no follow-up within one business day.
- **Secondary condition:** possibly clogged. 12 deals are past the normal sales cycle, and the total open-deal count is needed to judge.
- **Still unknown:** whether it's also starved. The counts can't say, so run the last-ten-wins test.
- **Recommended action:** set the one-day rule, a named owner and a daily unanswered-reply list. Then give every stuck deal a next step with a date, or park it. Then run the last-ten-wins test.

Write one for your own pipeline. Start with the earliest break you can prove, then work down.

## The verdict

Back to the question that started this: why does the pipeline look healthy while revenue isn't coming in? Because a pipeline can look full and still be starved of the right prospects, leaking interest, or clogged with deals that stopped moving, sometimes all at once.

Sales pipeline health isn't the size of the number on the dashboard. It's whether prospects, replies and deals keep moving. A pipeline doesn't necessarily need more leads. It may need better targeting. It may need faster follow-up. It may need firmer control of the deals already in it. Real sales pipeline optimization starts by finding the earliest break, because fixing a later problem may not solve an earlier one.

So: count the five numbers. Find the earliest break. Fix that first.

If you'd like help investigating your own pipeline, Vara Tech offers a free 30-minute Sales Diagnostic. We'll review your last 90 days, run the five-number X-ray, show where the pipeline is losing momentum, and recommend what to fix first. [Book your free Sales Diagnostic](https://wa.me/918484857426?text=I%20would%20like%20to%20book%20the%20free%2030-minute%20Sales%20Diagnostic.) or see [how it works](/how-it-works).

## Sources

- Oldroyd, J., McElheran, K. and Elkington, D. (2011). "The Short Life of Online Sales Leads." Harvard Business Review 89(3). Record: [scholarsarchive.byu.edu/facpub/9711](https://scholarsarchive.byu.edu/facpub/9711). Response-time figures and the 1.25-million-lead companion study as summarised by [Harvard Business School Working Knowledge](https://www.library.hbs.edu/working-knowledge/why-business-it-innovation-is-so-difficult).
- Salesforce (2026). State of Sales, 7th edition. 4,050 sales professionals, 22 countries, surveyed August to September 2025. [ANZ edition PDF](https://www.salesforce.com/en-au/wp-content/uploads/sites/12/documents/pdf/sales/ss_state-of-sales-7th-edition-anz-localisation-v3.pdf).
- Challenger (2024). "[How to Overcome Customer Indecision Using the JOLT Effect Approach](https://challengerinc.com/losing-to-customer-indecision/)," summarising Dixon and McKenna's 2020 research.
- Vara Tech deep-tech IoT campaign figures: Vara Tech case study.