The Google Ads setting that can make a campaign look successful while enquiries stay flat

A Google Ads campaign can look successful when the conversion action it counts is easier to complete than the enquiry your business needs. Check both the event definition and whether the campaign is using its goal for bidding before judging the result.

A page view, download, submitted enquiry and sales-qualified opportunity describe different customer actions. If they appear under one loose label, a rising total can conceal a weak enquiry pipeline. The dashboard may be counting exactly what it was configured to count.

The detail worth reviewing is the relationship between the action and the campaign goal. Google explains that primary conversion actions are generally used for bidding when the campaign uses the associated standard goal. Secondary actions are generally observed through All conversions. Custom goals are an important exception: actions included in them can be used for bidding regardless of their primary or secondary setting. Google Ads explains conversion goals and their bidding use.

That makes this an audit question before it becomes a settings change. Trace the reported action back to what the person actually did, then follow genuine enquiries through receipt, CRM creation and sales qualification. The checks below help you separate a counting problem from a campaign problem.

Why this is worth reviewing now

Google’s 13 October 2026 Demand Gen update announced a Data Strength Uplift metric and expansion of Data Manager. It also described enhanced conversions for Google Analytics and Display & Video 360 as coming soon. That wording is not confirmation that every feature is available in every account.

The business lesson is broader than any one feature: better measurement begins with deciding what should count. A new integration cannot decide whether an enquiry is commercially suitable for you.

Write a conversion dictionary

List each action appearing in your reports. Give it a plain-English definition, a trigger, an owner and a business purpose. The definition should be specific enough that your marketing manager, developer and sales lead would classify the same customer action the same way.

“Lead” is usually too vague. “A successfully submitted service enquiry that creates a record in our CRM” is more precise. “Qualified opportunity accepted by sales after checking service fit” is a different stage and should remain visibly different.

Keep useful browsing actions separate from customer commitments. Viewing a pricing page can be an interesting signal, but it is not equivalent to asking for a proposal. Otherwise, a report can improve simply because it counts more easy actions.

Trace one genuine journey from beginning to end

Choose a safe test process agreed with the people managing your systems. Mark test records clearly so they do not become real sales work. Follow the path from the landing page through form submission, acknowledgement, inbox delivery, CRM record and reporting.

At each stage, record what you expected and what happened. A visible thank-you message does not prove the sales team received the information. A CRM record does not prove the advertising report counted the event correctly. A recorded event does not prove that the customer was suitable.

• Check required fields and error messages.
• Confirm the record reaches the intended team.
• Check whether repeated clicks create duplicates.
• Inspect source information without placing personal data in campaign URLs.
• Make sure failed submissions are not presented as completed enquiries.

Use the same process for a phone enquiry or booking if it is a major path. A perfect web-form setup is not a complete measurement plan for a business that wins most enquiries over the phone.

Separate missing data from poor-quality data

A gap in reporting can come from several places: a broken tag, an unsupported integration, a consent choice, a blocked browser request or a journey that crosses systems. Those causes need different responses. Do not solve every gap by collecting more personal information.

Poor-quality data is a different problem. Duplicate submissions, spam and ambiguous status labels can create a misleading picture even when every technical connection works. Agree a consistent way to classify unsuitable contacts before judging campaigns by their lead totals.

Document what you cannot observe. A report with an honest limitation is more useful than an apparently complete dashboard built on assumptions nobody remembers.

Ask sales for a small amount of reliable feedback

A complicated CRM classification system often fails because it is too much work. Start with the minimum information needed to make a campaign decision. For example: suitable enquiry, unsuitable enquiry, duplicate, or still being assessed.

Add a short reason when it changes the marketing response. Wrong service, wrong geography and unrealistic timing are different problems. “Bad lead” leaves the campaign team guessing.

Keep the review window aligned with your buying cycle. A consultancy lead may need several conversations before qualification; a product purchase can be confirmed much sooner. Comparing yesterday’s new leads with last month’s completed opportunities produces an unfair conclusion.

An illustrative audit

Imagine a training provider reports 80 conversions in a period. Its team discovers that the total combines brochure downloads, enquiry submissions and repeated thank-you-page views. The number was not fabricated, but the label “leads” made it misleading.

The useful correction is to separate the actions and validate the underlying records. The team can then ask which enquiries are suitable for the course and whether downloads assist later decisions.

Do not simply delete inconvenient data to improve appearances. Preserve the history, annotate the definition change and explain why comparisons across the change date need care. Otherwise the next report may present a measurement correction as a sudden performance collapse.

Make a budget decision with the corrected evidence

After the audit, decide whether the immediate need is better measurement, a clearer offer, more suitable traffic or more response capacity. Increasing spend is only one possible action.

If measurement is unreliable, fix and validate the important event before using it to judge a larger investment. If enquiries are suitable but go unanswered, review the handover. If the account attracts the wrong work, examine campaign and landing-page scope. Each choice should follow the evidence, not the availability of a new feature.

Record your decision rule in advance: the outcome you want, the observation period, the evidence you will use and what would make you reconsider. Avoid importing another business’s conversion-rate target without understanding its market and definition.

Give the next specialist a useful brief

Bring the conversion dictionary, one traced journey and a list of unresolved discrepancies to a discussion about paid advertising. If the problem sits between the form and your sales system, include workflow and integration requirements in the brief.

That gives the work a clear purpose: helping the business make a sound decision with evidence it understands.