コードなし・低コストのデータ取り込み：BigQuery DTSの新機能Zero-code, low-cost data ingestion: New BigQuery DTS capabilities

Google Cloud has outlined a set of expansions to its BigQuery Data Transfer Service (DTS), positioning the tool as a code-free, lower-cost way to move data into BigQuery. The update matters because data ingestion remains one of the most persistent and expensive problems in modern analytics, and the changes appear aimed at reducing the engineering effort that teams spend keeping pipelines running rather than analyzing the data itself.

The core argument behind the announcement is a familiar pain point. According to Google, many enterprises are caught in what it describes as a costly paradox, spending more than 100 hours a week building and repairing fragile, in-house extract, transform, and load (ETL) pipelines, or contending with unpredictable third-party tools. BigQuery DTS is presented as a way to eliminate that engineering burden by acting as a fully managed, zero-code data movement solution that automates ingestion into BigQuery. The stated goal is to let teams shift from pipeline maintenance to what Google calls strategic data science in minutes.

The central news is an expansion of the connector ecosystem. Google says it is rapidly broadening its integration landscape to eliminate data silos across databases, advertising, and marketing platforms. In practice, connectors are prebuilt integrations that handle the mechanics of authenticating to a source system, pulling data on a schedule, and landing it in BigQuery tables without a customer writing or maintaining custom code. Adding more of them widens the range of systems that can feed BigQuery directly, which is the kind of incremental but meaningful improvement that reduces the number of bespoke pipelines an organization has to own. The source material specifically highlights new advertising and marketing connectors, though the full list of additions was not detailed in the excerpt available.

To understand why this is significant, it helps to place BigQuery DTS in context. BigQuery is Google Cloud's serverless data warehouse, and DTS is the managed service designed to populate it from external and first-party sources. Historically, DTS has been closely associated with Google's own advertising and marketing products, such as Google Ads and Google Analytics data exports, making it a natural fit for teams that need marketing performance data alongside operational data. Expanding beyond that base toward more databases and third-party platforms is consistent with a broader push to make BigQuery a central destination for heterogeneous enterprise data.

The zero-code framing also connects to a wider industry trend. Managed ingestion and so-called ELT approaches, where raw data is loaded first and transformed inside the warehouse, have grown popular as an alternative to hand-built ETL. Competing and complementary tools in this space include third-party services such as Fivetran and Airbyte, as well as Google's own Dataflow and Datastream for streaming and change-data-capture scenarios. BigQuery DTS is likely to appeal most to teams that want scheduled, batch-style loading from supported sources without operating additional infrastructure, whereas more complex or real-time transformation needs may still call for those adjacent tools.

For readers evaluating the update, a few prerequisite concepts are worth keeping in mind. Data silos refer to information trapped in separate systems that cannot be easily combined for analysis, and reducing them is the explicit motivation Google cites for the new connectors. A fully managed service means the provider handles provisioning, scaling, and reliability, shifting operational responsibility away from the customer. And the emphasis on eliminating in-house ETL reflects a cost-and-reliability calculation as much as a technical one, since fragile pipelines tend to consume engineering time disproportionately.

It is worth noting that the available material is drawn from a Google Cloud blog post, so the claims about time savings and customer trust are the company's own characterizations rather than independent measurements. The figure of thousands of customers using the service daily and the estimate of more than 100 hours a week on pipeline work are presented as framing for the value proposition. As with any managed connector platform, the practical benefit for a given organization will depend on whether its specific source systems are supported and how well the scheduled ingestion model fits its latency and transformation requirements. For teams already invested in BigQuery, however, a growing catalog of no-code connectors is likely to lower the barrier to consolidating more data in one place.