TikTok Ads to BigQuery: the routes, the real costs and the data traps

Is there a native TikTok Ads to BigQuery connector?

No. TikTok does not export to BigQuery, and Google's BigQuery Data Transfer Service does not list TikTok among its sources. When I read the service's documentation on docs.cloud.google.com in November 2026, its marketing platforms were Facebook Ads, HubSpot, Klaviyo and Mailchimp, plus Google's own advertising products. TikTok Ads was absent.

So every TikTok Ads to BigQuery pipeline uses one of four routes: a managed connector, an open-source pipeline such as Airbyte, a script you write against the TikTok API for Business, or reports exported from Ads Manager and uploaded. All of them read the same TikTok reporting API, so the data quirks later in this guide apply whichever you pick.

Which way to load TikTok Ads into BigQuery suits you?

Marketing teams without an engineer should use a managed connector. Teams already running Airbyte or a similar stack can add TikTok as one more source. Custom code makes sense only when you need fields or report shapes no connector offers, and manual uploads only for a one-off analysis.

What does TikTok Ads to BigQuery cost?

Expect the connector to be the main cost and BigQuery to be close to free for most brands. The prices below are what each vendor listed on its own pricing page when I checked in November 2026. They change often, so confirm before buying.

Windsor.ai charges $19 a month on annual billing ($23 month to month) for its Basic plan with 3 data sources and daily refresh, $99 a month annually for Standard with hourly refresh, and offers a forever free tier for a single source and account. BigQuery is included as a destination on all its plans. Coupler.io's Starter plan is $24 a month annually ($32 monthly) for 3 accounts refreshed daily, Active is $99 a month for 15 accounts, and Pro is $199 a month with hourly refresh. Fivetran bills by monthly active rows, with 500,000 MAR a month free and a $5 base charge per standard connection on paid plans. Airbyte's open-source edition costs nothing in licences but needs a server and someone to run it.

Google's BigQuery pricing gives every billing account the first 10 GiB of storage and the first 1 TiB of on-demand query processing free each month, with queries beyond that at $6.25 per TiB. Ad-level TikTok data for a single brand is measured in megabytes.

• Avoid the BigQuery sandbox for production: Google's documentation says sandbox tables expire after 60 days.
• Audience breakdowns by age, gender, country or platform multiply row counts, which matters on row-priced tools.
• If TikTok is one of three paid channels, a single multi-source plan usually beats a separate connector per platform.

A worked example: a DTC brand with two TikTok accounts

Consider a direct-to-consumer brand with two TikTok ad accounts, one billed in US dollars on Eastern time and one in pounds on London time, running around 200 ads at any moment. It wants daily ad-level spend, impressions, clicks, conversions and conversion value in BigQuery, joined to Shopify orders in Data Studio.

Two hundred ads across 365 days is about 73,000 rows a year. At 300 bytes per row that is roughly 22 MB, nowhere near the free 10 GiB. Reloading the last 7 days every night rewrites about 1,400 rows per run, which costs nothing meaningful in BigQuery. A Data Studio report reading a partitioned summary view scans a tiny fraction of the free terabyte.

The connector cost depends on how many sources travel with TikTok. With TikTok, Meta and Google Ads on one plan, Windsor.ai's Basic tier at $19 a month annually covers it; two TikTok accounts alone fit Coupler.io's Starter plan. The hidden work is in the model, not the bill: converting pounds to dollars at a stated rate and deciding whether a reporting day follows New York or London. These figures are my own arithmetic from the published prices, not a quote.

Why don't TikTok numbers in BigQuery match Ads Manager?

Usually for one of three reasons: late conversions, time zones or attribution settings. None of them is a connector bug, and all can be designed around.

Late conversions come first. Airbyte's TikTok Marketing documentation, updated October 21, 2026, notes that TikTok's reporting API has about 11 hours of data latency, so yesterday's figures are still changing when a nightly job runs. Its connector defaults to a 3-day attribution window for incremental syncs, configurable from 0 to 364 days, and recommends at least 3. A pipeline that only appends yesterday will undercount conversions permanently.

Time zones come second. TikTok's account setup help says an ad account's time zone, country or region, and currency cannot be edited after creation, and that time zone decides how results are dated in reports. Two accounts on different time zones produce days that do not line up, and neither lines up with a Shopify store set to a third zone.

Attribution comes third. TikTok's Attribution Manager help describes its default windows as 7 days after a click and 1 day after a view, and TikTok credits a conversion to its own ad. GA4 and Shopify use different rules, so TikTok conversions in BigQuery will rarely equal the orders those systems attribute to TikTok. Keep both and label them.

What can go missing in a TikTok Ads to BigQuery sync?

Deleted objects and some automated campaign types. Airbyte's documentation gives an optional setting to include deleted ads, ad groups and campaigns in report streams; leave it off and spend from a paused-then-deleted ad disappears from history while still counting in Ads Manager totals. It also notes that TikTok may omit the modify_time field for Smart+ ads, and records without it are excluded from the ads stream.

Throttling is the other gap. TikTok rate-limits by access token and by advertiser, and Airbyte retries specific throttling error codes with backoff. A home-built script without retries just skips the failed day. Placements are worth checking too: TikTok brought its TikTok Ad Network, formerly Pangle, to US advertisers on October 5, 2026, so confirm your connector exposes a placement field before Ad Network spend gets mixed into in-app results.

• Turn on deleted objects, or snapshot ad names daily into a dimension table.
• Compare ad count and total spend per account against Ads Manager after the first backfill.
• Add a check that alerts when the latest date in the table is more than two days old.
• Check that Smart+ campaign spend appears at the campaign level even if ad-level rows are thin.

How should TikTok Ads data be modelled in BigQuery?

Model it for reloads, currencies and joins. The fact table should accept overwrites of recent days, carry each account's time zone and currency, and keep the IDs needed to match Shopify orders or GA4 sessions through UTM values.

Partition the daily fact table by date and cluster it by advertiser and campaign. Store spend and conversion value in the account currency alongside a currency code, and convert in a view. Keep ad, ad group and campaign names in dimension tables keyed on TikTok IDs, since names get edited mid-flight. Put audience breakdowns in separate tables so they never inflate the main totals.

• Grain: one row per date, advertiser, campaign, ad group and ad.
• Load pattern: merge or delete-and-insert the last 7 days each run; widen to 28 if you lengthen attribution.
• Columns to keep: advertiser ID, time zone, currency, campaign objective, placement where available.
• A cross-channel view that unions TikTok, Meta and Google Ads with identical column names.
• A Data Studio data source on that view, using the owner's credentials.

When should you skip BigQuery for TikTok Ads?

Skip it when TikTok is a test channel with one account and modest spend, and the question is simply whether it works. A partner connector straight into Data Studio is quicker and cheaper; my TikTok Ads to Data Studio guide compares those options.

Move to BigQuery when TikTok becomes a real budget line: several ad accounts or currencies, a need to tie spend to Shopify or WooCommerce orders, history you want to keep beyond what a connector caches, or a blended report across more than five platforms. Data Studio blends stop at 5 sources with equality joins, while one unioned BigQuery table has no such ceiling.

How I set up TikTok Ads to BigQuery for clients

I begin with a free review of your ad accounts, time zones, currencies and the reports the team actually uses. Then I recommend the connector or pipeline that fits your volume, and quote a fixed price for the load, the BigQuery model, the cross-channel views and the Data Studio dashboard on top.

The pipeline runs in your Google Cloud project and your connector account, with the reload window, currency rules and day definition written down so nothing depends on me. Retail dashboards from projects I've led at Greenwolf Tech Labs, tracking MoM revenue growth, conversion rate and online versus in-store sales, rely on exactly this kind of clean, joined source underneath.