Skip to content
Signal network online · 14 chains Request Demo
SEO与数字营销

Google Ads to JSON: Fields to Check Before You Reconcile

PBy PFN DaiPFN Dai

For finance teams exporting Google Ads to JSON for CRM reconcile: where JSON actually comes from, which IDs, grain, dates and conversion rules to lock, and a narrow check before any bid change.

Brand promotion: this article introduces Guangsuan's Google Ads support for export and reconcile checks.

Direct answer: There is no JSON button in the Google Ads interface. If a tool needs JSON, query the REST form of the Google Ads API with GAQL, or stage data in a warehouse and emit JSON there. Before you reconcile with the CRM, record range with account time zone, keep stable IDs and row grain, fix click-time versus conversion-time, separate primary from secondary actions, and convert cost micros once. Check one action in one campaign group for a closed period before changing bids. This applies when the file will guide spend or lead decisions.

No JSON button in the interface

The download control above a statistics table offers Excel CSV, TSV, PDF, XLSX, XML and Google Sheets. JSON is not listed. XML sits closest in the menu, but it serves systems that parse XML, not a pipeline expecting JSON. Converting a CSV or XML download to JSON on your machine is a local programming step you own. Keep that script versioned and test it on a small slice you have checked by hand.

For the field list behind this check, see Guangsuan's guide to exporting Google Ads data to JSON.

Use the REST API when a tool asks for JSON

The Google Ads API supports two interfaces, gRPC and REST. Most official client libraries use gRPC with protocol buffers, while the REST interface uses JSON representations of the same resources. You write the request in GAQL, and a GoogleAdsService query returns rows carrying only the attributes, metrics or segments you selected, for example campaign ID and name with clicks, conversions and cost micros.

Access needs a developer token and an authenticated client. Google documents four access levels: Test Account Access, Explorer, Basic and Standard, with different daily operation limits and test-account limits. Keep tokens and credentials in a secret store, never in a shared report.

Keep IDs and row grain joinable

Keep customer ID in the file. Keep campaign ID beside campaign name, because names are edited while numbers stay stable. Keep every split that defines a row, such as date, device, and conversion action where used.

A campaign total and a campaign-by-day-by-device file are different grains. Export the grain you will actually analyse. Keep the raw download unchanged and do analysis in a copy, and name the file with account, report, range, basis and pull time so the same slice can be replaced on retry instead of doubled.

Write down range, time zone and conversion basis

Exports are dated in the account time zone, which may not match your calendar or your CRM timestamps. A request logged late on the last evening can land in different months on each side with no error on either side. Write the zone on the file itself. Use the segment control to split by day, week or month rather than joining several separate ranges later.

Conversion-time reporting places the event where it happened. Click-time reporting places it where the click happened. A September click that becomes an October signup therefore sits in different months under the two rules. Pick one rule and keep it on both sides of a period comparison.

Before you build a conversion report, open the conversion settings and write down basis, which actions are primary, counting method and attribution model. Only primary actions feed the Conversions column used for bidding. Secondary steps stay separate and should be labelled that way in the export. The export only shows what Google was able to record. Sales-confirmed numbers live in your own system.

Handle currency and cost units once

Cost in API rows arrives as metrics.cost_micros in the account currency. Divide by one million once and note where that step happened. Do not add raw totals across different currencies. Keep forecast value separate from booked revenue.

As a hypothetical teaching illustration with invented figures, not a client outcome: a file logs 52 ad-tracked contacts in a month. Sales accepts 17 as qualified under a written definition, 12 dated inside that month and 5 dated in the next month, with 3 reaching order. Each total describes a different stage.

Run one narrow pass before changing bids

Select one primary action, one campaign group and one closed period. Align basis, range and zone on both extracts. Separate ad-sourced records from organic and referral entries. Check counting behaviour and repeated uploads before calling a remainder unexplained. Leave any remaining gap marked as unresolved rather than assigning unknown traffic to Ads.

When few contacts ever qualify under the written definition, review search-term intent and landing-page fit before moving bids.

When Guangsuan can help

Guangsuan's Google Ads management covers account structure, search terms, landing pages, conversion checks and reporting, with no fixed enquiry promise. Scope and fees sit in a contract. Details are on the Guangsuan Google Ads management service page.

FAQ

Click time or conversion time for month end?

Use the rule that answers the question and keep it identical on both sides, with window and zone filed with the extract.

Does valid JSON mean ready to reconcile?

No. Parsing says nothing about completeness, grain, units, date rule or which actions were summed.

Can a missing row mean a hidden zero row?

Yes. Very quiet rows can drop from a filtered view, so confirm filters and grain before saying the term never matched.

Allocate capital with intent.

PFN Dai fuses transformer-based on-chain analytics with proprietary DeFi signal models — engineered for funds that don't guess yield.

Request a Platform DemoView Audited Performance