How to Find Where Your Leads Are Getting Lost (90-Day Leak Checklist)

If you are asking where are leads getting lost, do not buy more traffic first. Start with the last 90 days of data and trace the path from visit to revenue. More clicks at the top of a leaky funnel only create more expensive work for the people downstream.

Our guide to where your leads are getting lost across the full lead generation system explains the larger picture. This post turns one part of that system into a runbook you can use in a spreadsheet, CRM report, or analytics dashboard.

Short answer: compare traffic, leads, qualified leads, sales conversations, opportunities, customers, and revenue over the same 90-day period. Segment the numbers by source and offer, then start with the largest commercially meaningful drop. The biggest percentage drop is not always the most expensive problem.

A lead leak is a measurable falloff between two stages of your acquisition and sales process. The stage with the largest useful gap is where you investigate first. “Useful” matters because a low-volume channel can show a dramatic percentage change without affecting the business, while a smaller loss on a high-value channel can be costing far more.

This checklist is for B2B marketers, founders, and growth teams who need to decide what to fix before launching another campaign. It is not a generic list of ways to generate leads. It is a way to find the weak handoff already inside the system.

Start with a 90-day snapshot, not a traffic hunch

Traffic is easy to see, so it often becomes the explanation for every lead problem. A traffic chart can be down because demand changed, a campaign ended, tracking broke, or the right people stopped arriving. It can also be flat while lead quality quietly gets worse. The chart tells you something, but it does not tell you where the process stopped working.

A 90-day window gives you enough activity to see patterns without mixing in years of changes. It is a practical review period, not a universal benchmark. If your sales cycle is longer than three months, keep the acquisition window at 90 days but extend the customer and revenue view far enough to capture the eventual outcome.

Export or record these numbers before changing anything:

• Sessions or qualified visits by traffic source.
• Lead submissions or other meaningful conversion events.
• Qualified leads, using the definition your sales team actually uses.
• Sales conversations, meetings, or completed discovery calls.
• Opportunities created.
• New customers and closed revenue.
• Marketing spend, if you are comparing paid channels.

Keep the dates, filters, and definitions beside the numbers. A report that says “leads” without explaining whether it includes spam, duplicates, existing customers, and job applicants is not a useful baseline.

For search traffic, use the Google Search Console Performance report to see queries, pages, impressions, clicks, and average position. For site behavior, use your analytics platform. For outcomes after the form, use the CRM. No single report sees the whole path.

If you want a quick planning aid before building the spreadsheet, use the lead generation calculator to frame the relationship between traffic, conversion rate, leads, and downstream goals. Treat the result as a planning scenario, not as a substitute for your actual CRM outcomes.

Build one funnel with six measurable stages

Before calculating a conversion rate, decide what each stage means. A simple lead-generation funnel has six stages:

1. Traffic: a person arrives from a source you can identify.
2. Lead: the person takes the action that starts a sales or marketing follow-up process.
3. Qualified lead: the inquiry meets agreed criteria such as fit, need, timing, or buying authority.
4. Conversation: the prospect has a real sales interaction, not just an automated confirmation.
5. Opportunity: there is a defined business problem and a plausible path to a purchase.
6. Customer and revenue: the opportunity becomes a customer and produces recorded revenue.

These labels are not interchangeable. A form completion is not a qualified lead. A booked meeting is not an opportunity. An opportunity is not revenue. If your team collapses every stage into “conversions,” you lose the information needed to choose a fix.

Use one row for each source and offer combination where the volume supports it. For example, “paid search / demo request” should not automatically be blended with “organic search / downloadable checklist.” The people, expectations, and next steps may be different.

The full-funnel view also protects the hub article from becoming another vague “get more leads” page. The point of the lead-loss system map is that every handoff matters. This checklist gives you a repeatable way to inspect those handoffs with actual business data.

Step 1: Check the traffic-to-lead handoff

The first question is not “How much traffic did we get?” It is “Did the right visitors have a clear next step?” Compare visits with leads by source, device, landing page, campaign, and offer. A traffic source that looks weak in aggregate may contain one page or audience that works well. A source with impressive volume may be filling the site with people who were never a fit.

Use this basic calculation for each useful segment:

Visitor-to-lead rate = leads ÷ relevant visits × 100

For an illustrative example, imagine a landing page received 4,000 relevant visits and 80 form submissions. The visitor-to-lead rate is 2%. That number is not automatically good or bad. It becomes useful when you compare it with another page serving the same audience, or with the same page before a known change.

Then read the rate beside lead quality. A page that produces 80 inquiries and 8 qualified leads may be more valuable than a page that produces 120 inquiries and 3 qualified leads. If you only inspect the first conversion, you may send more budget toward the weaker source.

Separate demand problems from handoff problems

Low traffic and low lead volume can be a demand or distribution problem. Healthy traffic with low lead volume points more strongly toward message, offer, page, or tracking. High lead volume with low qualification points toward audience or promise quality. These are different diagnoses, and each needs a different next action.

• Low visits, healthy lead rate: investigate reach, rankings, ad delivery, and channel coverage.
• Healthy visits, low lead rate: inspect intent match, offer clarity, page content, and form friction.
• Healthy visits and lead rate, low qualification: inspect targeting, claims, qualification questions, and sales criteria.

Check tracking before making a copy judgment. Compare the analytics event count with the thank-you page, form platform, and CRM. A missing event can look like a conversion problem. Duplicate events can make a healthy page look better than it is.

Step 2: Find friction on the page and in the form

If qualified people reach the page but do not become leads, inspect the conversation between the source and the form. The headline should make it clear that the visitor is in the right place. The offer should explain what they get. The proof should support the claim. The form should make the next step feel safe and understandable.

Start with message match. A visitor who clicks an ad about a lead-generation audit should not land on a generic page about digital marketing. The first screen should carry the same problem, audience, and promised next step. If the visitor has to translate the page back into the reason they clicked, you have added friction before the form begins.

Review the page in this order:

1. Expectation: does the first heading confirm why the visitor arrived?
2. Value: can the visitor explain what they will receive in one sentence?
3. Proof: is there evidence that fits this audience and problem?
4. Risk: does the page explain what happens after submission?
5. Action: is there one obvious next step?

Then look at the form itself. Count the fields, but do not assume the shortest form wins. A field can add friction and improve the quality of the conversation. The right question is which information is necessary now and which information can be collected later.

For example, a request for an audit may need a work email, website, and a short description of the problem. It may not need a full company history, a phone number, a budget range, and ten required dropdowns before anyone has explained the next step. Test the smallest form that gives the team enough context to respond well.

If you need a broader framework for prioritizing page friction, use where to begin with conversion optimization. Keep this 90-day review focused: identify the evidence for the leak, then choose one page or form change rather than redesigning the whole site from instinct.

Step 3: Inspect the lead-to-qualified-lead drop

A lead-to-qualified-lead drop tells you how well the promise, audience, and sales definition fit together. Calculate it as:

Lead qualification rate = qualified leads ÷ total leads × 100

Do not compare this number until the sales team agrees on what “qualified” means. One person may count every inquiry with a company email. Another may require a specific need, service fit, budget, and buying timeline. Both can produce a number, but neither can guide a marketing decision until the rule is written down.

Read the raw leads. Sample a manageable set from the 90-day window and label each one for fit, problem, timing, source, and outcome. You are looking for patterns such as:

• The ad attracts a different industry than the page describes.
• The offer promises a service the business does not actually sell.
• The form is too open and invites requests outside the service area.
• The page uses a broad term that hides a narrow sales process.
• The lead is a duplicate, spam submission, vendor, student, or existing customer.

This is where quantity can become a distraction. A low qualification rate does not always mean the form is broken. It can mean the campaign is reaching the wrong people, or that the message is attracting people who want a different outcome. The fix may be a negative keyword, a tighter audience, a clearer service boundary, or a more specific offer.

Use sales notes as marketing evidence

Ask sales to record a short reason code when a lead is disqualified. Keep the list small enough that people will use it: poor fit, no need, no timing, no response, duplicate, outside scope, or other. After 90 days, sort those reasons by source and campaign. The pattern will often be more useful than another pageview report.

Do not change the qualification definition just to make the rate look healthier. That destroys the comparison. If the definition needs to change, record the old rule, the new rule, and the date of the change.

Step 4: Find the handoff gap between qualification and conversation

A qualified lead that never has a meaningful conversation is usually an operations problem, a timing problem, or a trust problem. It is not automatically a traffic problem. Check how quickly the lead is assigned, whether the owner receives a notification, whether the contact details are correct, and whether the next step is clear.

Calculate:

Conversation rate from qualified leads = completed conversations ÷ qualified leads × 100

Then add time. A monthly total can hide a simple pattern: leads arriving during business hours get a response, while evening and weekend leads wait until the next workday. If your process promises a call, review the actual elapsed time from submission to first human attempt.

Trace one lead from the form to the CRM and then through the sales record. You should be able to answer these questions without asking three different people to search their inboxes:

1. Did the lead create a CRM record?
2. Was the original source and campaign preserved?
3. Was an owner assigned?
4. Did the owner receive a task or alert?
5. Was a confirmation sent to the prospect?
6. Was the next action recorded after each attempt?

If the answer is no at any point, mark that handoff as a leak. Automation can send a confirmation, but it cannot take responsibility for a conversation. Someone still needs to own the lead and record what happens next.

This stage is often overlooked because marketing dashboards stop at form submission. Add the CRM outcome to your reporting before deciding which campaign wins. A campaign that produces fewer leads but more completed conversations may deserve more attention than the apparent volume leader.

Step 5: Follow the opportunity to customer and revenue

The final part of the checklist asks whether the opportunities created by marketing become customers. This is the stage that protects you from optimizing for cheap activity. A low-cost lead can still be expensive if it consumes sales time and never creates a viable opportunity.

Track these transitions for each meaningful source:

• Qualified lead to conversation.
• Conversation to opportunity.
• Opportunity to customer.
• Customer to recorded revenue.

Use both counts and values. Ten opportunities with a small average contract may be less valuable than three opportunities with a larger commercial fit. Do not hide that difference in an overall conversion rate.

For a clearly labeled example, Campaign A spends $4,000 and produces 100 leads. Ten become opportunities and one becomes a $12,000 customer. Campaign B spends the same amount and produces 40 leads. Eight become opportunities and three become $30,000 customers. Campaign A has the lower cost per lead. Campaign B produces the stronger business result. The example is illustrative, not a benchmark.

Make sure revenue is attributed consistently. If sales updates an opportunity months after the original campaign, the source should not disappear. If a deal has multiple touches, decide whether your report uses first touch, last touch, or a multi-touch model. A simple consistent model is more useful than a sophisticated model nobody trusts.

When the CRM does not contain reliable source data, treat that as a measurement leak. You may still have a marketing problem, but you cannot tell which one until the outcome data is connected. Google also documents ways to send offline lead outcomes back into advertising systems, such as importing offline conversions into Google Ads. The exact setup depends on your CRM and consent process.

Use the biggest-drop rule to choose where to start

Once the six stages are filled in, calculate the drop between each pair. A simple stage loss is:

Stage loss = previous stage volume − next stage volume

A conversion rate is:

Stage conversion rate = next stage volume ÷ previous stage volume × 100

Use the percentage to spot a weak handoff, then use business value to rank the work. The largest percentage drop may be normal for an early stage. The largest revenue loss may be later in the funnel. You need both views.

For an illustrative funnel, 10,000 visits become 200 leads, 50 qualified leads, 20 conversations, 8 opportunities, and 2 customers. The traffic-to-lead rate is 2%, the lead qualification rate is 25%, and the opportunity-to-customer rate is 25%. If improving the landing page could double leads but the sales team cannot follow up, the extra volume may not create extra customers. If improving qualification produces fewer but better opportunities, that may be the better first move.

Score each leak on three questions:

1. Size: how much volume or revenue passes through this handoff?
2. Confidence: do the data and lead records support the diagnosis?
3. Fixability: can the team make and measure a change in the next 30 days?

Start with the issue that has a strong combination of size, evidence, and fixability. That rule keeps the review from turning into a debate about whichever chart looks most dramatic.

Run the 90-day lead leak checklist

Use this sequence in a working session with marketing, sales, and whoever owns the website or CRM. Assign one person to record definitions and decisions.

Before the review

• Set the 90-day start and end dates.
• List every active traffic source and campaign.
• List the offers, landing pages, and forms used in that period.
• Agree on the definitions for lead, qualified lead, conversation, opportunity, and customer.
• Confirm that time zones, duplicate rules, and test submissions are handled consistently.

During the review

• Export traffic and conversion events by source, page, device, and offer.
• Reconcile event totals with form submissions and CRM records.
• Sample lead records and label quality and disqualification reasons.
• Check assignment, response time, meetings, opportunities, and outcomes.
• Calculate stage conversion rates and record the denominator for every rate.
• Rank leaks by commercial impact, evidence, and ability to test a fix.

After the review

• Choose one primary leak and one supporting metric.
• Write a testable hypothesis about the cause.
• Assign an owner and a date for the change.
• Keep the original 90-day baseline unchanged.
• Schedule a follow-up review using the same definitions.

The last point is where many audits fail. Teams write down a recommendation, then launch three unrelated changes. When the next report improves or declines, nobody knows why. A useful checklist ends with a decision and an owner, not a longer backlog.

Turn the diagnosis into one measurable test

Once you identify the likely leak, state the hypothesis in a way the team can test. “The page needs work” is not a hypothesis. “Visitors from the audit ad are not submitting because the page headline describes our agency instead of the audit they expected” is specific enough to investigate.

A practical test brief has five parts:

1. Observed pattern: what the 90-day data shows.
2. Likely cause: what the lead records, recordings, or sales notes suggest.
3. Change: the one thing you will alter.
4. Primary measure: the next stage you expect to improve.
5. Guardrail: the quality or revenue measure that must not get worse.

Examples make the difference clear. If the leak is traffic to lead, change the offer or page message and watch lead rate with qualified-lead rate as the guardrail. If the leak is lead to qualified lead, tighten the audience or add a clear fit statement and watch qualified volume, not just total submissions. If the leak is qualified lead to conversation, fix routing and ownership, then watch response time and completed conversations.

Do not start with button color unless the 90-day evidence points there. Large changes in audience, offer, promise, routing, or follow-up are more likely to explain a meaningful stage drop. Small visual details can matter, but they should earn their place after the larger questions have been answered.

For a more complete way to prioritize conversion work, pair this checklist with the conversion optimization starting framework. The goal is not to run a perfect experiment. The goal is to make one change whose outcome teaches you something useful.

Watch for these false leak signals

Not every gap is a broken marketing step. Before assigning a team to fix it, rule out the common measurement and context problems.

• Seasonality: a 90-day period may include a holiday, event, or predictable demand change. Compare with a relevant prior period before calling it a leak.
• Small samples: a channel with a handful of leads can swing from zero to one customer without proving a trend.
• Mixed intent: brand, research, job-seeking, support, and buying queries should not always share one conversion benchmark.
• Broken attribution: redirects, missing UTMs, cookie restrictions, or CRM field changes can make a source look weaker than it is.
• Changing definitions: if sales changed the qualification rule halfway through the period, mark the break instead of comparing the rates as if nothing changed.
• Duplicate records: repeated form fills and multiple contacts from the same account can inflate lead volume.
• Lagging outcomes: an opportunity created near the end of the window may not have had time to become a customer.

Write down these limitations beside the finding. An honest “we cannot judge revenue yet because the sales cycle is longer than the review window” is better than a confident but wrong conclusion.

Keep the leak from opening again

After the first fix, preserve the baseline and create a small operating record. Every lead should have a source, campaign or offer, stage, owner, next action, and outcome. Every form should have a defined event and a way to reconcile submissions with the CRM. Every campaign should have a clear primary conversion and a downstream quality measure.

Give the report a regular owner. A monthly review can catch tracking breaks and routing failures. A quarterly 90-day review can identify changes in source quality, sales conversion, and revenue. The schedule matters less than using the same definitions so the trend means something.

Also record what you chose not to fix. If the team decides that a low-volume source is not a priority, write that down with the reason. This prevents the same debate from returning every month and makes it easier to revisit the decision when volume or strategy changes.

The best outcome of a leak review is not a dashboard with more rows. It is a shared answer to three questions: where the drop occurs, what evidence supports the diagnosis, and who will change what next.

Find the leak before you buy more traffic

When you know where leads are getting lost, your next marketing decision gets simpler. You may need more qualified traffic. You may need a clearer offer. You may need a page that continues the conversation from the ad. You may need better routing, faster follow-up, or a CRM field that preserves the source.

Start with the 90-day path from traffic to revenue. Measure each handoff, check the records behind the totals, and rank the biggest commercially meaningful drop. Then make one change with one owner and one guardrail.

If your team has traffic, forms, and sales activity but cannot explain where the system breaks, we can help map the path and identify what deserves attention first. Learn more about our conversion rate optimization services, or request a lead generation audit when you are ready to turn the diagnosis into a focused plan.