First, confirm conversions are not happening, not merely unrecorded
Start outside TikTok. In analytics, filter sessions by the campaign parameters (UTMURL tags identifying traffic source.) on TikTok destinations and look for orders, sign-ups or other goals; in the shop or CRM, look for orders and leads whose source is TikTok. If there are some, the conversions exist and the problem is that TikTok cannot credit them: the event may not arrive at all, or the click identifier may be lost between tap and page. The second and fourth rows of the table apply, and the TikTok pixel not firing page covers the event side in depth; this page only takes the first check.
If analytics shows TikTok sessions that bounce and no goals anywhere, the visitors are really leaving. Then read the gap between destination clicks and landing page views in Ads Manager, which the TikTok traffic setup describes as a direct reading of how many taps the page lost while loading. A wide gap points at the page in the in-app browser. A narrow gap with no action on the page points at the optimisation goal or the promise. Keep the answer to that first question in front of you as you read the table.
Diagnosis table
Diagnosis
| Symptom | Check | Likely cause | Fix |
|---|---|---|---|
| Plenty of cheap clicks, visitors leave at once | Optimisation goal of the ad group; objective of the campaign | Ad group optimising for clicks or page views | Build an ad group on the conversion event |
| Shop shows TikTok orders, Ads Manager shows none | Conversion event in Events Manager for those days | Conversion events not arriving | Repair the event and confirm it in test events |
| Destination clicks far above landing page views; high bounce on mobile | The page opened from a test ad inside TikTok on a phone | Landing page slow or broken in the in-app browser | Lighten the page; remove pop-ups and walls; test in-app |
| Events arrive but few are attributed to ads | Click identifier on the landing URL after redirects | Click identifier stripped by redirects or parameter cleaning | Preserve query parameters through every hop |
| Page loads well, visitors leave without scrolling | The video's claim beside the page's first screen | Video promising something the page does not offer | Land on the exact thing the video showed |
Fix, verify and prevent each cause
An ad group told to find tappers
The ad group's optimisation goal decides who delivery looks for. A click goal finds people who tap, and the feed has many people who tap out of curiosity and leave. A landing page view goal is closer, but still stops at arrival. If the business records purchases or leads, an ad group optimising for anything less is buying the wrong people efficiently.
Fix: create an ad group that optimises for the conversion event, stepping to a more frequent event such as adding to cart while the final event is rare, as the campaign structure guide explains for choosing the goal.
Verify: after learning completes, the conversion rate (CVRShare of visits or clicks that convert.) from tap to action rises even if the price of a tap goes up.
Prevent: match the goal to the outcome when the campaign is planned, and keep traffic campaigns for building audiences that conversion campaigns use later.
Sales that happen out of TikTok's sight
When the shop has orders that Ads Manager does not, the event is missing, mistimed or attached to a pixel the ad
account cannot use. This page takes only the first check: place a test order and look for the event in the test
events view; for a shop, the event to look for is CompletePayment, carrying its value.
Fix: if it is absent or empty, follow the TikTok pixel not firing page and the pixel and Events API tutorial, then add the server route so blocked browsers stop hiding sales.
Verify: test orders arrive once with value, and reported conversions begin to track shop orders.
Prevent: include a test conversion in every site release.
An event that does arrive can still be missing from the figure you are reading. Ads Manager credits a conversion to an ad only when it falls inside the attribution window set for the ad group, and reports can lag behind the shop, so an order placed days after the tap, or only in the last day or two, may not show yet. The shop counts every order whatever its source. Before calling the campaign dead, check the attribution setting, leave a few days after the date you are judging, and compare like with like; the platform's current documentation, such as the TikTok Ads Help Center, lists the window options and how quickly results appear.
A page that fails inside TikTok
TikTok opens links in its in-app browser on a phone, where heavy scripts, large images, pop-ups and consent walls cost visits that a desktop test never shows, and some payment buttons that depend on the device's own browser may not work.
Fix: open the page from a test ad inside TikTok, remove anything that covers the content on arrival, cut page weight, and check that the main action, including payment, completes in-app, as the TikTok traffic setup recommends.
Verify: the gap between destination clicks and landing page views narrows and analytics shows longer TikTok sessions.
Prevent: make an in-app test on a phone part of approving any landing page for TikTok.
A click identifier that never arrives
TikTok adds a click identifier to the destination address when someone taps an ad, a query parameter named
ttclid, and the pixel stores it so later events can be matched to the tap. A redirect that drops query parameters, a link shortener, or a site that
cleans parameters on arrival removes it, and conversions that do happen are not credited. The
pixel and Events API tutorial shows how to check for it.
Fix: follow the tapped link through every hop and make each one pass query parameters intact; exempt the identifier from any cleaning rule.
Verify: the landing address in a test tap still carries ttclid, and
attributed conversions rise while events stay level.
Prevent: add parameter survival to the checks for any redirect, domain or link-tool change.
A video that promises one thing and lands on another
A viewer taps because of what the video showed: one product, one result, one answer. If the page opens on a home page, a category grid or a different offer, the promise breaks on arrival and the visitor leaves. The content formats and hooks guide treats the hook as a contract; the landing page has to honour it too.
Fix: land each video on the exact thing it showed, with the same words as its spoken claim and on-screen text in the first screen.
Verify: analytics, read by the creative in the campaign parameters, shows deeper sessions from the matched pages.
Prevent: brief each video with its destination named, so the promise and the page are planned together.
Prevention: a landing page check in the in-app browser
Most TikTok conversion failures are visible in a minute on a phone. Before launch, tap a test ad inside TikTok and
watch: how long the page takes to show content, whether anything covers it, whether the first screen repeats the
video's promise, and whether the action, including payment or form submission, completes without leaving the app.
Check that the landing address still carries the click identifier (ttclid) and the campaign parameters. Place a
test conversion and confirm the event, such as CompletePayment, in the test events view. Confirm the ad group optimises for the outcome rather than
the tap. After launch, read the click-to-landing gap and the tap rate (CTRShare of impressions that result in a click.) beside conversions per
creative weekly, and repeat the in-app test whenever the page template changes. The
landing page audit adds a scored pass over message match, speed and form friction for pages
that carry most of the spend.
A supplement brand with busy traffic and no orders
A supplement brand runs a traffic campaign optimising for clicks and records 9,400 clicks in two weeks at £0.21 each, with analytics showing 4,100 sessions from TikTok and 11 orders. The shop's own report shows 14 TikTok orders; Ads Manager shows none, because the campaign has no conversion event selected. Opened from a test ad inside TikTok, the landing page takes several seconds to show anything and then displays a full-screen discount pop-up over a category page, while the best-performing video demonstrates one product. The team moves the pixel's CompletePayment event onto the order-complete push, lands each video on its product page with the pop-up removed, and builds a conversion campaign optimising for add to cart. In the following fortnight landing page views rise to 78% of clicks, analytics records 63 orders from TikTok, and Ads Manager attributes 51.
Common mistakes
- Judging TikTok on Ads Manager alone when the shop shows sales. A missing event, a lost
ttclidor a short attribution window can make working traffic look dead; compare with orders by source first. - Testing the landing page on a desktop browser. Heavy scripts, pop-ups and payment buttons fail in the in-app browser on a phone, so test the page there.
- Optimising for clicks with a working pixel. Once the event arrives, build the ad group on it, stepping to a more frequent event while purchases are rare.
- Sending every video to the home page. Land each video on the exact thing it showed, or the promise breaks on arrival.
- Adding a link shortener that drops the click identifier. Follow the tapped link through every hop and confirm
ttclidsurvives. - Reading the latest days as final. Recent orders may fall outside the attribution window or not have reached the report yet; give them a few days.
