Why I Ask Clients for Their Worst Lead First

Why do I ask clients for their worst lead before anything else?

Because the worst lead tells me more about a go-to-market system than the best one. A great lead usually arrives despite the system. A bad lead shows exactly where targeting, qualification, routing, or messaging broke. So early in discovery, I ask a client to show me one lead they regret spending time on.

Most discovery calls start with goals. More pipeline, better conversion, a faster site. Those goals are real, but they are also vague, and almost every company says the same things. When I ask for the worst lead, the conversation gets specific almost immediately. People remember the lead that wasted their week.

I have delivered 100+ projects for 25+ clients over 6+ years, most of them websites and the automations behind them for lead generation. I lean on this question because it skips the polished version of a company's funnel and shows me the real one.

What exactly do I mean by the worst lead?

I mean a real lead, from the last few months, that took time from someone on the team and turned out to be a poor fit. Not spam a filter could catch. A lead that looked plausible, got a reply, maybe a call, and then went nowhere for a reason that, in hindsight, was visible early.

The distinction matters. Obvious junk tells you about your spam filter. A plausible bad lead tells you about your ideal customer profile, your form, your messaging, and your qualification rules. It is the lead that fooled the system, which makes it the most useful one to study.

I ask for the actual record if possible. The form submission, the CRM entry, the emails, the call notes. Memory smooths over details. The record shows what the team knew at each step and when they could have spotted the problem.

What does the worst lead reveal that a dashboard cannot?

It reveals the specific decision points where the system let a bad fit through. Dashboards show totals and averages, like conversion rate by source. One bad lead, traced end to end, shows the actual mechanism: which field was missing, which rule was too loose, which page attracted the wrong buyer.

When I trace a bad lead, I usually walk through five questions. Where did they come from? What did the page they landed on promise? What did the form ask, and what did it fail to ask? Who received the lead, and why that person? What did the first reply say, and what should it have said?

Somewhere in those five answers there is almost always a fixable gap. Sometimes it is copy that attracts the wrong company size. Sometimes the form never asks a question that would have disqualified the lead in one click. Sometimes the routing sends everything to the founder, who is too busy to qualify properly.

Why not start with the best customers instead?

I do study the best customers, but second. Best customers are where you want to go. The worst lead is where you are leaking right now. Fixing the leak first protects the team's time, and clean time is what lets them work the good leads properly.

There is also a bias problem with best customers. Teams tell a flattering story about why those deals closed, and the story often credits the wrong thing. The worst lead has no flattering story. Nobody is defending it. That makes the discussion more honest and the lessons more reliable.

After the worst lead, I look at the best customer and ask a simple question: what would have to be true on the website and in the system for ten more like this to arrive and be handled well? The two conversations together give me the shape of the work.

How does this question change what I build?

It points the build at the real constraint. If the worst lead came from misleading copy, the first fix is on the website. If it came from a weak form, the fix is qualification. If it sat untouched for days, the fix is routing. I would rather fix the actual leak than build what was requested.

This connects to a rule I hold firmly: I will not automate a process that is already broken. Automating a broken qualification step just sends bad leads faster. I wrote about that in why I won't automate a process that is already broken, and the worst-lead question is often how I find the broken part.

It also keeps the work proportionate. A small B2B team does not always need a big enrichment stack. Sometimes the worst lead shows that one extra form question and one routing rule would remove most of the waste. That is a small, fixed-fee piece of work, and it pays back quickly.

What should you expect when you ask this question?

Expect a specific story rather than a summary. The question gives people permission to talk about what frustrates them without sounding negative about their own company. It also signals that you care about the daily reality of the sales team, not just the look of the website or the size of the pipeline.

Expect some hesitation too, especially if the worst lead came from a channel or campaign someone in the room is proud of. That hesitation is useful. It tells you where internal politics might shape the project, and it is far better to know that in the first week than to discover it at launch.

And if someone says there are no bad leads, take that seriously as well. It usually means there are very few leads at all, or that nobody tracks what happens to them after the form. Either answer changes the scope of the work, so the question still did its job.

How can you run this exercise on your own?

Pick your three worst leads from the last quarter. For each, trace the source, the page, the form answers, the routing, and the first reply. Write one sentence on where the problem could have been caught earliest. Then fix the earliest point that shows up more than once.

Do it as a team, not alone. The marketer sees the page and the campaign. The rep sees the call. The person who built the automations sees the routing. Each one spots a different part of the failure.

Write your definition of a qualified lead down afterward. It will be sharper than before, because it now includes what a bad fit looks like in practice. My notes on what should count as a qualified lead for a two-person team may help you frame it.

What has this question taught me about my own work?

It taught me that most lead problems are system problems, not people problems. Reps who waste time on bad leads are usually following the system they were given. The fix lives in targeting, the website, the form, and the automations, which is exactly where GTM engineering does its work.

It is also part of why I describe my work as GTM engineering first and Webflow second. The website is often where a bad lead starts, but the full answer runs through the CRM, the routing, and the reply. I explained that shift in why I lead with GTM engineering instead of Webflow.

What should you do next?

Before your next pipeline meeting, ask each person to bring one lead they regret working. Trace each one end to end, find the earliest point it could have been caught, and fix the point that repeats. You will learn more in an hour than from a month of dashboards.

The same habit shapes how I start redesigns, and I wrote about why I ask for recent lost deals before a website redesign.

I build websites and the automations behind them for lead generation. If you want an outside view on where your worst leads are coming from, bring one to a call. Reach out and let's chat.