Attribution and customer journeys

Why a ChatGPT campaign can show purchases and zero goal conversions

Explain purchases with zero goal conversions in ChatGPT Ads. Inspect campaign settings, event detail and source connections before diagnosing tracking failure.

Start reading
Editorial illustration: Two terracotta parcels stand beside a green ceramic form with an empty star-shaped opening.
Editorial illustrationVisible purchases can belong to event detail without matching the selected campaign goal.
The working guide

What you can work through.

Attribution and customer journeys
  • A goal summary can differ from attributed event activity.
  • Inspect the attached event setting before editing the campaign.
  • Keep purchases, leads and other actions separately labelled.

The report contains attributed purchase value, yet its conversion total is zero. Before declaring a tracking failure, ask what the total counts. In ChatGPT advertising reporting, an event can appear in attributed activity without being one of the campaign’s configured goals.

That distinction matters for both diagnosis and client communication. “No goal conversions” is not necessarily the same statement as “No attributed purchases”. It also does not mean that the campaign optimized for purchases simply because purchases appear elsewhere in the report.

Inspect the campaign’s goal definition

OpenAI’s conversion tracking documentation distinguishes sending an event, defining a conversion event setting and attaching that setting to a campaign. These are separate parts of the measurement setup. A website can send purchase events without the relevant campaign treating that purchase setting as its goal.

Retrieve or inspect the campaign configuration through the supported account workflow. Record the attached setting identifiers and the intended business action. Check the setting itself rather than relying on a friendly name that someone may have copied from another campaign.

Do not change the configuration during the initial investigation. First preserve the observed setup and report. An immediate edit can erase the context needed to explain why the earlier period showed zero goal outcomes.

Read summary and event detail separately

In the dedicated conversion endpoint, the outer conversions field counts campaign goals. The nested attributed_event_count can also include non-goal activity. An empty conversion_event_setting_breakdowns array means that no goal settings matched that event detail. Reading these fields together provides a concrete check before treating the apparent mismatch as a tracking fault. Their definitions are in OpenAI reporting.

In a hypothetical campaign optimized or measured for a different action, two attributed purchases might appear in event detail while the goal summary remains zero. That can be a coherent report. The purchase activity belongs in a clearly labelled event row, while the zero remains a statement about the configured goal count.

Do not sum every event row and replace the goal total with that number. One customer journey can generate several different actions, and the events may represent different business stages. Adding a registration, a lead and a purchase does not create three new customers.

Working reference

Purchases and goal counts can differ

  1. Event detail

    Two attributed purchases can appear as event activity.

  2. Goal summary

    Zero means no outcomes matched the campaign goal in scope.

  3. Check

    Inspect goal, source and attachment date before assuming tracking failure.

Hypothetical campaign with a different configured goal.

Check whether the apparent mismatch is intentional

The campaign may deliberately use a goal other than purchase. For example, the team may evaluate a lead action while inspecting attributed purchase activity as additional context. Verify that choice with the campaign brief rather than assuming the largest-value event should always be the goal.

If the intended goal was purchase but the configuration differs, classify the issue as a setup discrepancy. Record when the setting was attached and what historical interpretation is supported. Do not assume that correcting the setup will rewrite every earlier reporting period.

The goal-attachment date is a separate question from whether events were received. A correct current configuration does not prove that the same goal applied throughout the period you are reviewing.

Follow the event’s source

When several websites or integrations send events, confirm that the setting points to the intended source. A purchase event with the right name from the wrong source will not answer the campaign’s measurement question. Names alone are insufficient evidence of a correct connection.

Use the source identifier, event name and configuration record together. If the report still appears inconsistent, preserve a minimal example with the campaign, setting, source and relevant report fields. Avoid including customer details that are unnecessary for diagnosing the configuration.

Event-source auditing covers that connection more closely. This article concerns the distinction between goal summary and event detail, not the full installation of browser or server tracking.

Explain the result in business language

A useful report note could say: “The campaign recorded no outcomes matching its configured lead goal in this period. The attributed-event detail includes two purchases, which are shown separately.” That statement gives the client the actual distinction without claiming that either field is broken.

If the configuration was unintended, state the discovered mismatch, the correction path and the limit on historical interpretation. Do not promise recovered conversions or improved optimization until the relevant behavior has been verified. The fix is a setup action, while its effect remains something to observe.

Keep the labels in the metric dictionary. Analysts should not alternate between goal conversions and all attributed events under the same dashboard heading. A definition change can otherwise masquerade as a performance increase.

For implementation, use OpenAI conversion tracking and the reporting documentation. The distinction is useful even when the report is consumed through another interface: inspect what each number counts before deciding that a ChatGPT campaign produced nothing or that a tracking system failed.

Sources and scope

Distinguish attributed event activity from the campaign-goal summary when a report appears internally inconsistent.

Working methods and examples are editorial suggestions. Check current platform requirements and available features before implementation.

Your next chapter

See what campaign reporting covers.

Explore reporting in AthillyAds and how it fits alongside your own measurement of enquiries, purchases and other business outcomes.