Google Ads Transparency Center

It is the disclosure layer Google maintains for its own ad systems: a searchable record of the creatives actually served across Search, Play, Maps, Shopping and YouTube. Where the Google Ads API reports your accounts, this record covers every verified advertiser on the platform - rivals included - which is what makes it a competitive-intelligence instrument rather than a reporting console.

Each entry names the advertiser behind the creative and the payer when the two differ, shows the creative itself, and attaches the pair of figures almost nowhere else discloses: an amount-spent range and an impression-count range, counted per surface. Dates served come stamped at both ends, and advertiser-supplied targeting arrives as explicit included and excluded criteria.

Two clocks govern depth. Verified-advertiser commercial creatives remain in scope for one year from their last display - a rolling window, so the freshest year is always the deepest. Election creatives carry a seven-year window and a second disclosure tier: verification status, a regulatory identifier (the FEC ID for US filers), an advertiser type from Presidential Campaign through Super PAC, and ballot and promoter statements. Millions of creatives sit inside those windows; Google declines to publish an exact total.

Every row resolves to a single creative with its advertiser, window and disclosure ranges intact. Four shapes cover most workloads:

row_shape        : video creative on YouTube
advertiser_id    : AR04022286554273505281
creative_format  : VIDEO
surface          : YOUTUBE
dates_served     : activity_start .. activity_end
ranges           : impressions_lower_bound..impressions_upper_bound,
                   spend_data_lower_bound..spend_data_upper_bound
--------------------------------------------------------------------------
row_shape        : text ad on Search
creative_format  : TEXT
surface          : SEARCH
geo              : served_country codes
targeting        : included vs excluded audience and topic criteria
--------------------------------------------------------------------------
row_shape        : image creative, cross-surface
creative_format  : IMAGE
surface_options  : MAPS | SHOPPING | PLAY | SEARCH
days_served      : num_served_days
flags            : contains_synthetic_media true | false,
                   unverified true | false
--------------------------------------------------------------------------
row_shape        : election ad disclosure row
regulatory_id    : FEC ID for US filers
advertiser_type  : Presidential Campaign | Senate Campaign | House Campaign |
                   PAC | Super PAC | Political Party | FEC Filer
retention        : seven years past the ad's last impression

These are the archive's own field structures, not illustrations of rows we cannot show - identifier, date and range values populate per the advertisers and windows you name. Get a sample of this dataset and the rows land in exactly this shape, cut to whichever advertisers, regions and date ranges you specify.

Eighteen columns make up the confirmed core: three establish who paid (advertiser_id, advertiser_name, payer_name), four stamp the serving window (activity_start, activity_end, num_served_days, creative_format), and the rest carry the disclosures that separate this record from a screenshot gallery - impression and spend ranges, per-platform placement, region targeting, verification status and the synthetic-media flag. Definitions map to the archive's own field structure gathered during the August 2026 research pass; no upstream data dictionary was invented to fill gaps, which is precisely why the remainder sits behind an explicit request note.

Geography - verified-advertiser creatives arrive worldwide. Unverified-advertiser creatives appear only for Europe and Turkey, a regional rule worth remembering before comparing volumes across markets. Election creatives exist wherever Google supports political advertising at all.

Temporal - two rolling windows instead of one fixed archive: commercial creatives stay in scope one year from their last display, election creatives seven years from theirs. Creatives age out on schedule, so the record rewards routine capture over occasional raids - take a full snapshot once, or keep the warehouse current so new creatives diff cleanly into yesterday's rows.

Granularity - the individual creative is the atomic unit, rollable up to advertiser level. Impression counts break out per surface, so a single creative's Search-versus-YouTube split survives the rollup.

API, files, or your warehouse. Daily, weekly, or hourly.

Your cadence runs independently of the archive's own rhythm. Take one full snapshot of an advertiser set, or keep the feed current so newly served creatives diff cleanly against yesterday's. Deliveries arrive typed and normalized to the field dictionary above with the entity ladder intact - advertiser, creative, surface, window - so joins against your CRM, your brand roster or your rivals' hold up without parsing free text off ad cards.

• Competitive creative intelligence - follow rival pipelines across all five surfaces with spend ranges attached, catching new campaigns while they are still running rather than in next quarter's deck.
• Share-of-voice estimation - impression ranges plus per-platform breakdowns approximate relative visibility before syndicated panels report the same story a quarter late.
• Political ad accountability - election creatives trace to payer, regulatory ID and advertiser type over seven years, with ballot and promoter statements intact.
• Synthetic-media monitoring - the altered-or-synthetic flag turns AI-content tracking from vibes into a query.
• ML corpora at library scale - millions of creatives pre-labeled with advertiser, format, window and ranges, ready for creative-scoring and policy models.
• Partner vetting - verification status and historical volume check an agency's or prospect's claimed footprint.

Competitive intelligence and product teams get the headline use case: rival creative pipelines with money context attached. Data scientists and ML engineers get a corpus that would cost six figures of labeling to assemble by hand. Journalists, academics and students get citable political-ad provenance reaching back seven years. Developers and builders stand monitoring and dashboards on normalized rows instead of scraped cards. Market researchers and consultants benchmark whole categories rather than cherry-picked examples, and sales and growth teams read a prospect's real creative footprint before the first call.

Within advertising data, this record owns the Google-side everyone-else layer: every verified advertiser's creatives with spend and impression ranges, yours or anyone's. Its nearest neighbour in kind is the TikTok Creative Center - Top Ads record, a leaderboard of top-performing short-form-video creatives - useful for inspiration and reach comparison, but ranked-view only, with no spend ranges and no per-advertiser history.

The Google Ads API record is a complement, not a substitute: it reports your own account performance down to keyword level, while this record supplies the market around it - run both and share-of-voice becomes arithmetic. The Meta Marketing API, X (Twitter) Ads API and TikTok Business API & Ads Manager records complete the platform set for account-side reporting. Political money trails run through the FEC Campaign Finance Data and FCC Political Files records; broadcast creative lives in the Internet Archive TV News Archive. If your question is "what is every verified advertiser running on Google right now, and roughly what did it cost", this is the record.

Three things worth knowing upfront. First, the commercial window is rolling: creatives leave scope one year after their last display, so a snapshot taken today is strictly deeper than one taken next quarter - capture beats backfill. Second, the impression and spend figures arrive as ranges, Google's chosen disclosure granularity - precise enough to rank advertisers and size categories, not to audit an invoice. Third, unverified-advertiser coverage is regional (Europe and Turkey only), so cross-market comparisons should be drawn within the verified-advertiser universe. Name the advertisers, regions and windows you need, and the sample returns in exactly the schema shown above.