How to measure lead quality by marketing source

A high-quality lead is a potential customer whose needs fit what you sell. They have a realistic path to buying. A form submission tells you someone contacted you; it does not establish either of those things.

Start with a short definition your team can use. For a local installation company, a qualified enquiry might mean:

• The person needs a service the company provides.
• The property is inside the service area.
• The requested work meets the company's minimum project size.
• The person is willing to discuss a quote and can help arrange the work.

A software business will use different criteria. It might check the team's use case, company size, and whether the product meets a required feature need.

Keep fit and timing separate. Someone planning a suitable project for next quarter may be worth following up. They are different from someone asking for a service you do not offer.

If your CRM uses MQL and SQL labels, agree on what each means. In HubSpot's lifecycle definitions, marketing qualification and sales qualification are separate stages. You do not need every stage to start, but everyone updating records should use the same definitions.

Avoid judging leads by their source before reviewing them. Marking every referral as good and every social lead as poor would build your assumptions into the report.

A channel report becomes useful when it connects to individual enquiries. You need to see that one lead came through paid search, qualified for your service, and later became a customer.

You can start with a spreadsheet. Keep one row per new sales enquiry, with a stable ID so repeated submissions can be matched.

These are suggested record labels, not required product field names. Keep a review date on your report, too.

Capture available source details when the person visits and pass them with the form submission. Our website form source tracking guide explains that process. TrackLeadSource is built around preserving source and journey context with form leads. Qualification and sales outcomes still need to be recorded in your CRM or spreadsheet.

Use one source rule throughout the report

Someone might discover you through organic search, return through an ad, and then enquire. A report based on first recorded source can give different results from one based on the last recorded source.

Choose a rule and label it. Use that rule for every row. Keep the underlying journey where available so you can investigate later. Our lead attribution guide explains the main choices.

A person's answer to “How did you hear about us?” belongs in a separate column. People can forget the details, while tracking can miss offline recommendations or visits it cannot connect. Keeping both sources gives you more context. Do not overwrite one with the other.

Use a small set of measures you can maintain. Here, a lead means a unique new sales enquiry. Remove confirmed spam, test submissions, and duplicate entries from that count, but keep their counts separately.

Keep real enquiries that do not fit your service in the lead total. Removing them would hide a targeting problem. Existing customer support requests belong in a separate report.

Use counts and percentages together. A channel with one customer from two leads has a 50% customer rate. That is too little evidence to assume it will keep performing that way.

For cost metrics, state what you included. If you use ad spend alone, call it ad spend per qualified lead. Agency fees, creative work, software, and sales time can change the full cost. Do not label an ad-spend-only figure as total customer acquisition cost.

If there are no qualified leads, cost per qualified lead is undefined. Show “No qualified leads” alongside the spend, rather than reporting zero cost.

Imagine a service business reviewing one month's new enquiries. The numbers below are fictional and show the method, not expected performance for these channels.

Each lead has been observed for 90 days after its enquiry date. In this example, all have a qualification decision and a final sales outcome within that window. Each customer is counted once. Revenue means the value of their first completed purchase, excluding tax and net of refunds.

The business uses the same recorded-source rule for all three channels. Paid campaign spend covers the acquisition month. This is a period-level comparison, not a claim that each ad dollar can be matched to one lead.

Paid social brought the most enquiries. It also had the lowest qualification and customer rates in this example.

Looking only at ad spend per lead, social appears cheaper: $2,000 ÷ 100 = $20, compared with $2,400 ÷ 60 = $40 for paid search.

But paid search produced more suitable enquiries. Its ad spend per qualified lead was $2,400 ÷ 30 = $80. Social's was $2,000 ÷ 20 = $100.

Paid search also produced 12 customers from 60 leads: 12 ÷ 60 × 100 = 20%. Social produced four from 100: 4%.

That gives the business a reason to investigate the social campaigns. It does not prove that paid social is always a worse channel. One campaign, offer, or audience may account for the difference.

Organic search brought fewer leads but a higher share of suitable enquiries. Its cost is not zero. Content, SEO work, and website maintenance still take resources. Those costs are not included in this table, so it cannot establish which channel is most profitable.

The revenue figures are not profit or marketing ROI either. Before making a spending decision, account for delivery costs, sales effort, and other relevant costs. Higher revenue can come with a lower margin.

A lead that arrived yesterday has had less time to buy than one from three months ago. Comparing their current outcomes can make the newer group look worse.

Group enquiries by when they arrived. Then compare what happened within the same amount of time after each enquiry. This is often called a cohort comparison: a cohort is simply a group that started during the same period.

Choose a window that fits your sales cycle. The example uses 90 days to make the method clear. A business with longer projects may need longer. When a window is still incomplete, mark the report as provisional.

Keep pending outcomes visible

Suppose 20 qualified leads have reached the same review point. Five became customers, five were lost, and ten are still open.

The customer rate among those qualified leads is 5 ÷ 20 = 25% so far. The win rate among resolved deals is 5 ÷ 10 = 50%. Both calculations can be useful, but they answer different questions. Label the denominator.

The ten open deals are not ten lost sales. Keep them open and update the report later. Similarly, an enquiry nobody has assessed yet is not automatically unqualified.

Keep acquisition reporting separate from this month's sales total. A sale closed in October may belong to a lead acquired in July. For a lead-quality comparison, attach that sale to its original enquiry group.

Start with the records behind a weak result. A percentage tells you where to look; the reason tells you what to change.

Review results by campaign when you have enough leads. A broad channel average can hide one campaign that works and another that attracts the wrong enquiries.

Be careful when sales handling differs. If one group gets a reply within an hour and another waits several days, the outcome reflects more than marketing quality. Check the process before blaming the channel.

Keep unknown sources visible instead of quietly dropping them. If a large share of your customers has no recorded source, the known-source report only tells part of the story.

When a pattern repeats across comparable groups, use it to plan a measured budget change. Watch what happens to qualified leads, customers, and costs afterwards. Recorded attribution helps organise evidence; it does not prove how many sales would disappear if you stopped a channel.

Choose a clear qualification rule and apply it to each new enquiry. Save the source, record the outcome, and give leads a fair amount of time to buy.

Then compare channels using the same definitions. You will have a better basis for deciding where to spend your time and budget than a lead count alone.