How to Track Which YouTube Videos Bring Leads

Define the action you want to count

A lead is a completed action that meets your written business rule. It might be a completed booking, a submitted inquiry, or an email signup. Choose the action before you count results.

A click shows that someone used a link. It does not show that they completed your form. A completed form does not show that the inquiry fits your business. A lead does not show revenue.

The supplied reader questions ask about these separate needs. One asks how to track video link clicks, booked calls, and email signups. Another asks which videos generate leads. These are direct public questions, not verified customer results. See the link-tracking question, published September 26, 2025; checked October 25, 2026; question evidence only and the video-to-lead question, published June 24, 2025; checked October 25, 2026; question evidence only.

Write two rules:

1. Lead rule: State the completed action that enters your report. For example, count a booking only after your website records its completion.
2. Quality rule: State what makes that lead useful. For example, count an inquiry as qualified after the sales owner accepts it as relevant.

A qualified lead is a lead that passes the quality rule. Keep “not reviewed” separate from “reviewed and did not qualify.” A missing review is not a failed review.

If your website accepts bookings and email signups, report each action separately first. Do not add them together and call the result a count of people. One person could complete both actions.

Build a video register before you add links

A video register is a table that connects each video to its website link. Use the stable YouTube video ID as the video key. A key is a value used to match records.

Save these fields:

Keep older link versions. Replacing a link must not erase the meaning of tags already saved on lead records.

You can fill the register manually for a small video list. The YouTube Metrics Scraper can help collect public identity and context. An Actor is a hosted data collection program. A Task is a saved input preset for that program.

The documented Actor fields include `postId`, `postUrl`, `publishedAt`, `channelId`, `views`, `postMetricStatus`, `status`, and `scrapedAt`. These support video identity and public metric records. They do not provide website leads or private YouTube Studio analytics. This fit comes from the public default build 1.1.2, checked October 25, 2026; contract inspection only, no paid run.

Start with one known video

The supplied small input is valid against the checked public input schema:

{"maxItems":1,"startUrls":["https://www.youtube.com/watch?v=dQw4w9WgXcQ"]}

The URL is a sample. Replace it with your intended known video URL. `startUrls` holds the input URLs. `maxItems` limits the number of returned metric rows. Here, the limit is one row, not one lead or one viewer.

You can start from the YouTube Video Performance Tracker Task, checked October 25, 2026; public preset presence only. Inspect its saved settings and replace its sample inputs before use. The supplied evidence verifies the public page and preset identity. It does not verify the saved input through authenticated access or show that the preset was run.

If you run the collection, check the returned video ID and URL against your register. Save the collection time and status with public metrics. Keep missing values as missing. Do not replace an unavailable view count with zero.

Give each video a distinct tagged destination link

UTM parameters are labels added to a destination URL to describe referred traffic. They help connect a website visit with a saved source label. Google documents `utm_content` as a way to distinguish a creative or link. See Google campaign URL guidance, checked October 25, 2026; traffic-label facts only.

Choose consistent values for these fields:

For example, you could use `youtube` for the source and `video` for the medium. These are proposed naming choices, not required platform values.

Make `utm_content` distinct for each video link. If you use a link-version label, save its exact match to the video key in the register. The label must have one clear meaning.

Follow these steps:

1. Choose the website page that serves the video's offer.
2. Create the tagged link with your chosen values.
3. Save the full link and its video key in the register.
4. Check that the link label has no conflicting register entry.
5. Save each later link version before you place it.

Do not put names, email addresses, or other personal data in URL parameters. Use a content label that identifies the link.

Decide whether your question concerns the whole video or a specific link placement. One shared link label can group traffic from that video's eligible placements. Distinct placement labels require separate register entries that still point to the same video key. Choose only the detail you need.

A tag is a saved label. It does not identify an individual YouTube viewer. It also does not prove why someone completed a form.

Check the actual link location

A correct destination URL is useful only if viewers can use it from the intended location.

YouTube says URLs in Shorts descriptions and Shorts comments are not clickable. External links on long-form videos require advanced features. Check YouTube link guidance, checked October 25, 2026; format and eligibility limits.

For each video:

1. Confirm that the intended location supports a clickable external link.
2. Open the video as a viewer.
3. Use the actual placed link.
4. Check the destination page and retained tag values.
5. Record the test date and any failed step.

Do not assume that a pasted URL works because it looks correct in your editing screen.

If you also need to know whether Shorts viewers watch your long videos, use the separate Shorts and long-video audience check. Website tags answer a different question from audience overlap.

Save the tags with the completed lead

Your website must retain enough source information to connect the completed action to the link register. Seeing a tag on the first page is not enough. Check that the final lead record retains it.

Keep owner lead records separate from public metric records. The following are proposed table fields, not Actor output fields:

Attribution means the rule used to assign a lead to a recorded source. Choose that rule before reporting.

A simple proposed rule is to credit the video whose registered tag the website retains on the completed lead record. State how your website handles later tagged visits and whether it retains a tag for a later completion. Do not describe that behaviour until you have checked it.

Keep directly captured values separate from later interpretation. For example, a sales note saying “I saw a video” is different evidence from a retained link label. Do not silently combine the two.

A tagged session is a website session recorded with your chosen source tag. A session is a group of website activity under your website measurement system's definition. It is not a count of distinct people. Save the definition used for your report.

Test a complete lead before you count results

Complete a marked test lead through the actual placed link. Check the full path from the video to the final record.

1. Confirm that the destination receives the expected tags.
2. Complete the defined lead action.
3. Confirm that the final completion event appears.
4. Confirm that the form or booking record exists.
5. Check that the record retains the expected link label.
6. Match that label to the correct video key.
7. Check how repeat submissions enter the report.
8. Remove the test from business totals or keep it clearly labelled.

A duplicate rule states when two records count as the same accepted lead. Write this rule for your business. Repeated submissions must not quietly become additional people or additional qualified leads.

Do not remove every repeat without review. A repeat record could be a duplicate submission or a separate valid action under your rule. Preserve the source records and record your decision.

This is a test procedure for the reader to perform. The supplied evidence does not show a completed test or a working website implementation.

Match records and keep failures visible

A join combines table records through matching keys. First match the captured link label to the register. Then use the exact video key to attach public context.

Set a reporting period for completed lead events. State its start, end, and time zone. Record the public metric collection time separately. A public view total collected today is not automatically a view count for your lead reporting period.

If you compare public counters, use the guide to view-count compatibility to check their meaning and measurement windows. Do not turn a cumulative public counter into a period count without supporting records.

Use these failure states:

Unattributed means no retained video tag supports a video assignment. Do not assign these leads to the most-viewed video. Do not spread them across videos in proportion to public views.

Zero has a different meaning. Report zero accepted tagged leads only when your checked records contain none under the stated rules. If tag capture failed or the records are incomplete, mark the affected result as unknown or incomplete.

Original fiction: compare lead volume and quality

Everything in this example is original fiction. The video labels, sessions, leads, quality states, and reporting choices are teaching examples. They are not customer results or outputs from a tested setup.

Assume one reporting period and one defined completed lead action. Marked tests and duplicate submissions have already been handled under the fictional written rules. The session and lead records use the same declared attribution scope.

The extra review states are also original fiction. They show why an unfinished review must remain unknown.

The tagged-session-to-lead rate uses accepted completed leads divided by tagged sessions:

• Video A: 5 leads ÷ 100 tagged sessions × 100 = 5%.
• Video B: 4 leads ÷ 20 tagged sessions × 100 = 20%.

These are tagged-session rates. They are not view-to-sale rates or percentages of YouTube viewers who became leads.

Video A has more tagged sessions and one more completed lead. Video B has more leads that pass the quality rule: three compared with two. Neither observation proves that Video B will perform better in a future test.

Keep the totals separate:

• Tagged completed leads: 5 + 4 = 9 leads.
• Unattributed completed leads: 3 leads.
• All accepted completed leads: 9 + 3 = 12 leads.
• Leads known to pass the quality rule: 2 + 3 = 5 leads.
• Leads awaiting quality review: 3 + 1 + 3 = 7 leads.

The three unattributed leads stay unattributed. Their count is known, but their video source is unknown. Their quality is also unknown in this example.

The next useful action is to finish the reviews and inspect why tags are missing. The current counts can guide a further content test. They do not establish sales, revenue, viewing direction, or video-caused demand.