You're Losing Deals Somewhere Specific. Most Founders Have No Idea Where.

Every article about "conversion funnel analysis" assumes you have a data team and a dashboard. You don't. Here's how to map your sales funnel, find the exact step where deals are not closed, and fix the one leak that matters.

Most founders can't tell you where their deals actually do not close. Sales problems are almost never spread evenly; they concentrate in one leaky step. Photo by Melanie Deziel/ Unsplash

Ask a founder how their sales are going and you'll usually get one of two answers: "Pretty good!" or "Not great." Both are useless, because both are vibes. Neither one tells you the single most important thing in your entire sales operation:

Why, exactly, are deals not closed?

Not "the funnel needs work." Not "we need more leads." Why, and at what stage of your overall sales activities? At which specific step does a promising conversation turn into silence? Because here's the thing I've learned across 500+ startups: almost every founder's sales problem isn't spread evenly across the whole funnel. It's concentrated in one leaky step. And once you find it and fix it, everything downstream improves at once.

The problem is that every guide on this topic is written for a growth team with a data warehouse with heatmaps, session replays, cohort analysis, and tracking plans. That's great if you're optimizing a checkout flow for 50,000 monthly visitors. It's worthless if you're a founder with 40 leads, a notebook, and a sinking feeling that something's broken.

So this is the no-dashboard version. You can do all of it on a whiteboard in an hour. Let's find your leak.

Step 1: Draw your actual funnel (not the textbook one)

Get a whiteboard or a blank page. Draw the real steps a customer goes through with you, from "never heard of us" to "paying." Not the generic awareness → consideration → decision nonsense. Write down your steps.

For most founder-led B2B sales it looks something like:

Lead → First conversation → Demo/proposal → Follow-up → Closed.

Five steps. Maybe six. If you have more than seven, you're overcomplicating it. If you have fewer than four, you're probably skipping a real stage (most founders forget "follow-up" is its own step, which, spoiler, is often exactly where the problem is).

The rule: each step has to be a moment you can actually count. "How many leads this month?" — countable. "How many had a first conversation?" — countable. "Brand awareness", not countable, doesn't go on the board.

Step 2: Put real numbers on each step

Now, for the last 30 or 60 days, write the number of people at each stage. Be honest. Painfully honest. This is the part founders fudge, and fudging defeats the entire purpose.

It might look like this:

• Leads: 50

• First conversation: 20

• Demo/proposal: 12

• Follow-up engaged: 5

• Closed: 2

  
  

Don't worry yet about whether these numbers are "good." Just get them real and on the board. You cannot fix what you refuse to count, and most founders have simply never written these five numbers down in one place. The act of doing it is, by itself, often the moment the problem becomes obvious.

Step 3: Calculate the drop between each step (and find the biggest leak)

Now the useful part. Look at the percentage that survives each transition — not the whole funnel, each step.

From the numbers above:

• Lead → First conversation: 20/50 = 40% survive (60% drop)

• First conversation → Demo: 12/20 = 60% survive (40% drop)

• Demo → Follow-up engaged: 5/12 = 42% survive (58% drop)

• Follow-up → Closed: 2/5 = 40% survive (60% drop)

  
  

Here's the discipline that separates this from guesswork: find the step with the worst survival rate, that also involves real volume. That's your leak. That's the one step where fixing it changes everything.

One caution the analytics crowd gets right: look at absolute numbers too, not just percentages. A step that loses 60% of 50 people is worse than a step that loses 80% of 5. Prioritize the leak that's losing the most actual humans, not the worst-looking percentage.

In the example above, the lead → first conversation step is the weakest. 30 of 50 people never even have a conversation with you. That's where the money's leaking.

Step 4: Diagnose why. This is where founders have an unfair advantage

The analytics tools stop at "here's where the drop-off is." They can't tell you why, because a heatmap can't get on the phone.

You can. And this is the founder's superpower that no data warehouse can match: you can just ask.

Pick five people who dropped off at your leakiest step. Choose five real humans and Email them. Not a survey, write a human note:

"Hey, you and I talked a few weeks back and it didn't go further, which is totally fine. I'm trying to get better at this. Can I ask, honestly: what made you not move forward? No wrong answers, and I'm not trying to sell you anything."

You'll be amazed how many reply, and how useful the answers are. The leak is almost never what you assumed. The "they didn't have budget" you imagined turns out to be "your pricing page confused me." The "they went with a competitor" turns out to be "you never followed up and I forgot about you."

Five honest conversations beat any dashboard ever built. The drop-off number tells you where. The humans tell you why. You need both, and only you can talk to people.

Step 5: Fix one thing. Just one.

Here's where founders sabotage themselves: they find the leak, get excited, and try to fix the entire funnel at once. New pricing page and new demo script and new follow-up sequence and new lead source. All in the same week.

Then nothing's measurable, because everything changed at once, and you're back to vibes.

Identify the step where most customers drop off. Make a change. Then monitor that specific step for a few weeks and see if the conversion rate changes.

If leads aren't turning into conversations, the fix is probably in your outreach or your qualification, you're either reaching the wrong people or your first message is weak. If demos aren't turning into follow-ups, your demo is probably pitching instead of discovering. If your follow-ups aren't leading to closed deals, you're probably not following up often enough or offering a discount at the first sign of hesitation.

The uncomfortable truth most founders find

When founders actually do this exercise, the leak is usually somewhere they weren't looking, and usually somewhere earlier than they thought.

Everyone obsesses over closing. The dramatic final meeting, the negotiation, the signature. But the most common leak I see isn't at the close at all. It's at the top, founders who have a decent close rate but simply aren't having enough real conversations, because they're hiding behind their product instead of doing outreach. Or it's the follow-up step, the deals that quietly evaporate because nobody sent the third email.

The close gets all the attention. The leak is usually upstream, somewhere boring, somewhere you stopped looking because it didn't feel like "real" sales work.

That's the gift of actually mapping it. It points you at the unglamorous step that's quietly costing you half your revenue, the one you'd never have guessed from vibes.

Most founders can't tell you where their deals actually do not close. Sales problems are almost never spread evenly; they concentrate in one leaky step. You don't need a dashboard to find it. Draw your real funnel (4–6 countable steps), put honest numbers on each, calculate the survival rate of each transition, and find the worst one that involves real volume. Then use your unfair advantage over every analytics tool: call five people who dropped off and ask why. The number tells you where; the humans tell you why. Fix one thing, re-measure, repeat. The leak is usually earlier and more boring than you'd guess, which is exactly why you've been missing it.

Want a second pair of eyes on where your deals are actually leaking — and what to do about it? That's exactly what I do with founders. Book a free 30-minute strategy call →

Or grab the Startup Sales Strategy Workbook → to be able to develop your sales strategy.