ChatGPT answers, unedited

For exporting granular social ad data into a data warehouse for custom analysis, I’d shortlist these platforms:

My recommendation by scenario

1. Enterprise / high-volume / serious custom modeling → Fivetran

• Good choice if you have Snowflake, BigQuery, Redshift, Databricks, or a similar warehouse.
• It can replicate ad platform data into your warehouse and supports advertising connectors including Meta/Facebook Ads and other platforms. www.fivetran.com
• Strong fit if your analysts want to join ad data with CRM, product, revenue, and attribution models.
• Pair well with dbt for standardized campaign/ad-level models; Fivetran provides ad reporting transformation models across multiple ad platforms. fivetran.com

2. Marketing team owns analytics, engineering support is limited → Supermetrics

• Strong choice if the main users are growth/marketing analysts.
• It supports automated transfers of marketing data into warehouses such as BigQuery, Snowflake, Redshift, and others, with incremental loading and normalized schemas. supermetrics.com
• Good coverage for social platforms and marketing sources. docs.supermetrics.com

3. Data engineering team wants flexibility → Airbyte

• Better if you want to customize connectors, run infrastructure yourself, or avoid vendor lock-in.
• Expect more work around monitoring, schema changes, and connector maintenance.

Features I’d evaluate before choosing

For “granular” social ad analysis, verify the platform can extract:

• Campaign → ad set/ad → creative hierarchy
• Daily breakdowns by:
  • campaign
  • placement
  • audience
  • geography
  • device
  • creative
• Spend, impressions, reach, clicks, conversions
• Video metrics (3s views, completion rates, watch time)
• Attribution windows and conversion breakdowns
• Historical backfills
• API quota handling
• Schema-change management

A common architecture

A setup I see work well:

Meta Ads / TikTok Ads / LinkedIn Ads / Reddit Ads
→ Fivetran or Supermetrics
→ Snowflake / BigQuery / Databricks
→ dbt transformations
→ Looker / Tableau / Power BI / custom analysis

If you expect to build custom attribution models, incrementality analysis, creative fatigue models, or MMM, I’d lean toward Fivetran + dbt. If you mainly need campaign reporting and analyst self-service, Supermetrics is often faster to deploy. supermetrics.com

A couple of details would change the recommendation: which warehouse are you using (Snowflake, BigQuery, Databricks, etc.), and which ad platforms matter most (Meta, TikTok, LinkedIn, Reddit, X, Pinterest)?

For this use case, I’d shortlist Supermetrics and Funnel, with Supermetrics as my default recommendation.

My pick: Supermetrics

It’s particularly well suited if your priority is getting granular paid-social data into a warehouse and then doing your own analysis in SQL/dbt/etc.

Why:

• Broad ad-platform coverage: Facebook/Meta, TikTok, LinkedIn, Pinterest, Reddit, Snapchat, Google Ads, etc. are supported. supermetrics.com
• Warehouse-native delivery: supports BigQuery, Snowflake, Redshift, Azure Synapse, Databricks and other destinations. supermetrics.comfunnel.io
• Granular dimensions/metrics: you can select fields and build transfers around the level of detail you need rather than being limited to a canned reporting schema. docs.supermetrics.com
• Incremental loading and backfills: useful when you're maintaining a durable analytical dataset rather than just refreshing a dashboard. supermetrics.com
• It also handles the API/schema plumbing, so your data team doesn't have to maintain individual ad-platform integrations. supermetrics.comfunnel.io

Strong alternative: Funnel

I'd lean toward Funnel if you want a more marketing-data-centric transformation layer before the warehouse. It supports hundreds of connectors and lets you configure custom dimensions/metrics, while exporting to destinations including BigQuery, Snowflake and Redshift. funnel.io

How I'd choose

One caveat: don't choose based merely on the number of connectors. For your use case, I'd validate the exact Meta/TikTok/LinkedIn fields and breakdowns you need, historical backfill limits, API refresh latency, and how each tool handles platform attribution changes. Those details matter much more than connector count.

If you tell me your warehouse (Snowflake/BigQuery/Redshift/etc.) and which ad platforms you use, I can give you a tighter 2–3 platform comparison, including likely architecture and tradeoffs.

If your priority is granular social-ad data in a warehouse for custom SQL/BI analysis, I’d shortlist Fivetran and Airbyte, with Fivetran as my default recommendation.

My pick: Fivetran

Fivetran is a strong fit if you want a managed, low-maintenance pipeline. Its Meta/Facebook Ads connector supports deep source replication and custom reports, including configurable fields, attribution windows, breakdowns, action breakdowns, and aggregation. www.fivetran.comgithub.com

The important bit for your use case is that it doesn't just dump a high-level campaign summary. Fivetran's connector supports granular structures such as ad-level insights and secondary tables for metrics like actions, action_values, and cost_per_action_type, preserving breakdown dimensions so you can do your own analysis downstream. beta.fivetran.com

It also supports common warehouse destinations including Snowflake, BigQuery, Databricks, Redshift, and S3. www.fivetran.com

When I'd choose Airbyte instead

Airbyte is compelling if your data team wants more control over extraction and schema or expects to build/customize connectors. Its Facebook Marketing connector lets you define custom Insights streams with a chosen account/adset/campaign/ad grain, fields, breakdowns, and action breakdowns. github.com

That's particularly attractive if "granular" means we want to control exactly what comes out of the Marketing API, rather than simply getting a vendor's normalized model.

Quick decision

Fivetran also has a Google Ads connector, so if you're trying to get Meta + Google + other paid channels into one warehouse, it becomes especially attractive. www.fivetran.com

One caveat: before buying, I'd test your actual required grain and dimensions against the connector. Meta's API has combinations of fields/breakdowns that aren't all simultaneously supported, so "supports Meta Ads" doesn't necessarily mean "can reproduce every report you can construct in Ads Manager." Airbyte explicitly exposes these custom-insight controls, while Fivetran exposes custom reports. www.fivetran.comgithub.com

If you tell me which social platforms (Meta, TikTok, LinkedIn, Pinterest, Snapchat, etc.) and which warehouse (Snowflake, BigQuery, Databricks, Redshift), I can narrow this to the best 2–3 options for your exact stack.

If your priority is granular social ad data → warehouse → custom SQL/analytics, my first choice would be Supermetrics.

Why I’d pick Supermetrics

• Broad paid-social coverage: Meta/Facebook Ads, TikTok Ads, LinkedIn Ads, Pinterest, Snapchat, Reddit, X, etc., plus Google Ads and other marketing sources. supermetrics.com
• Warehouse-native workflow: It supports BigQuery, Snowflake, Redshift, Azure Synapse, Databricks and others, with automated transfers and incremental loading. supermetrics.com
• Good granularity: For Meta specifically, its current connector exposes hundreds of metrics/dimensions and tables at campaign, ad set, ad, video, geo, age/gender, etc. levels. docs.supermetrics.com
• Useful for custom analysis: Once landed in your warehouse, you can join ad data to CRM, revenue, GA4, product, or attribution data and build your own metrics/models rather than being constrained by the ad platform UI. supermetrics.com
• Less engineering overhead: It handles API authentication, schema management, incremental loads, and error recovery rather than requiring your team to maintain individual ad-platform pipelines. supermetrics.com

The main caveat: I'd validate the exact fields you consider "granular" before buying. Ad APIs have platform-specific limitations and fields disappear/change over time; Meta, for example, has made recent changes to historical data, attribution windows, and breakdown availability. docs.supermetrics.com

My shortlist

For a team saying "we want raw-ish, granular ad data in Snowflake/BigQuery and then we'll do our own analysis", I'd start with Supermetrics, then compare it against Funnel on field coverage, historical backfill, API freshness, and cost at your expected row/volume level.

If you tell me your warehouse (Snowflake/BigQuery/Redshift/etc.) and the ad platforms you use (Meta, TikTok, Google, LinkedIn, etc.), I can narrow this down to the best 2–3 options and compare their granularity and likely tradeoffs.

For exporting granular social ad data into a data warehouse for custom analysis, I’d shortlist platforms based on how much control your data team wants versus how much managed transformation you need.

My recommendation by scenario

1. You have a data engineering team and want maximum analytical flexibility

• Choose Fivetran, OWOX, or Weld.
• Land the rawest available data in your warehouse.
• Build your own models for:
  • campaign → ad set → ad hierarchy
  • creative performance
  • audience/placement breakdowns
  • attribution joins with CRM and revenue data

2. You are a marketing analytics team without much engineering support

• Choose Adverity or a similar marketing-focused platform.
• You’ll get more normalization across Meta, TikTok, LinkedIn, Google, etc., without building as much transformation logic. www.adverity.com

3. You need very granular Meta/TikTok/LinkedIn data
Prioritize platforms that expose:

• ad-level objects (not only campaign summaries)
• creative IDs and metadata
• placement/device/demographic breakdowns
• conversion action arrays
• historical backfills
• raw API fields when possible

For example, some connectors explicitly support granular Meta structures such as campaigns, ad sets, ads, creatives, placements, demographics, and conversion metrics. datawarehouse.io

Architecture I’d generally recommend

Social APIs
   |
   v
ELT connector (Fivetran / OWOX / Weld / Adverity)
   |
   v
Cloud warehouse
(Snowflake / BigQuery / Databricks / Redshift)
   |
   v
dbt models
   |
   v
BI + custom analysis + attribution models

If this is for a serious analytics stack, I’d lean toward OWOX or Weld for warehouse-native teams, or Adverity for a marketing operations organization that needs standardized reporting across many channels. weld.appwww.owox.com

A key question that changes the recommendation: are you primarily pulling Meta/TikTok/LinkedIn ads, or do you need all paid media + CRM/revenue joins for attribution?

For granular social ad data → your data warehouse → custom analysis, I’d shortlist these three:

My recommendation

I'd start with Fivetran if your priority is genuinely granular, warehouse-native analysis. It is designed around replication into your warehouse rather than primarily around marketing reporting, which makes it a better architectural fit when your team wants to build its own attribution, incrementality, cohort, LTV, or creative-level models. www.fivetran.com

Choose Supermetrics instead if marketers need more control over the available ad-platform dimensions/metrics and you want a marketing-data specialist rather than a general-purpose ELT platform. Its documented Facebook Ads coverage is particularly extensive. supermetrics.comdocs.supermetrics.com

One important thing to evaluate in a POC: don't just compare connector counts. Test the exact fields you care about—ad/ad-set/campaign grain, creative IDs, placement, geography, age/gender, hourly data, conversion/action types, attribution windows, and historical backfill—and verify how each vendor handles API changes and Meta/TikTok reporting limitations.

If you tell me your warehouse (Snowflake/BigQuery/Redshift/etc.) and the social platforms (Meta, TikTok, LinkedIn, Pinterest, Reddit, etc.), I can narrow this to the best 1–2 options and compare granularity, historical backfill, pricing model, and raw-vs-modeled data.

If your priority is granular paid-social data → warehouse → custom SQL/BI analysis, my first recommendation is Funnel.

Why Funnel

• Broad social coverage: Facebook/Instagram, TikTok, LinkedIn, Pinterest, Snapchat, Reddit, etc., with hundreds of marketing connectors. funnel.io
• Warehouse-native exports: supports destinations including BigQuery, Snowflake, Redshift, Databricks, SQL databases, and cloud storage. funnel.io
• Granularity: you can select the dimensions and metrics you want exported rather than being limited to dashboard-level aggregates. funnel.iohelp.funnel.iohelp.funnel.iosupermetrics.com
• Creative-level data: Funnel can export creative assets and metadata—including Facebook Ads at the ad/creative level—to S3, GCS, or Azure Blob for downstream analysis. help.funnel.io
• Good fit for an analytics team: it gives you a managed ingestion/normalization layer without forcing your engineers to maintain individual ad-platform APIs.

One important caveat: “granular” doesn't necessarily mean raw impression/click event data. Ad platforms generally expose reporting data that's already aggregated by dimensions such as date, campaign, ad set, and ad. Funnel explicitly notes this limitation. help.funnel.io

Runner-up: Supermetrics

Supermetrics is probably the strongest alternative. It advertises 170+ marketing sources, incremental warehouse loading, and direct integrations with BigQuery, Snowflake, Redshift, and others. supermetrics.com It also provides explicit field-level documentation and warehouse schemas for Facebook Ads. supermetrics.com

My pick: Funnel if you're building a durable marketing-data layer for analysts/data scientists. Supermetrics if you want something particularly mature around field selection, reporting, and warehouse transfers.

If you tell me your warehouse (Snowflake/BigQuery/Databricks/etc.) and the ad platforms you use, I can narrow this down to the best 2–3 options and compare API coverage, historical backfill, refresh latency, schema flexibility, and approximate cost.

If your priority is granular social/ad data in your warehouse for your own SQL/modeling, I’d shortlist these three:

My pick: Improvado

For custom analysis rather than dashboarding, I'd lean Improvado. Its positioning is much closer to "marketing data engineering pipeline": extract granular data, transform/normalize it, and land it in your warehouse. It currently lists 1,000+ data sources and explicitly supports Snowflake, BigQuery, Redshift and Databricks. improvado.io

I'd evaluate it against Funnel if you care more about self-service configuration and controlling exactly which fields get exported.

One important diligence point: don't just compare the number of connectors. Ask each vendor for a sample raw export/schema for your exact ad platforms (Meta, TikTok, LinkedIn, Pinterest, Reddit, etc.) and check whether they preserve ad → ad set → campaign → creative, breakdowns, attribution windows, conversion types, and historical revisions. That's where "granular" can mean very different things between vendors.

For granular social/paid-ad data into a warehouse for custom SQL analysis, my first choice would be Supermetrics.

Why:

• Broad ad-platform coverage: Facebook/Meta Ads, Instagram, LinkedIn Ads, TikTok Ads, Pinterest Ads, Reddit Ads, Snapchat, X, etc. are supported. supermetrics.com
• Warehouse-native delivery: It supports BigQuery, Snowflake, Redshift, Databricks and other warehouse destinations, with automated transfers and incremental loading. supermetrics.com
• Good granularity: You can select dimensions/metrics and work from field-level source data rather than being limited to prebuilt dashboards. Their documentation provides field lists and warehouse-table schemas. docs.supermetrics.com
• Low engineering overhead: It handles API connections, schema management, incremental loads and error recovery, which is attractive if your data team doesn't want to maintain individual ad-platform APIs. supermetrics.com

When I'd choose something else

Improvado — I'd evaluate this if you're an enterprise/multi-channel organization and want heavier cross-channel normalization and a larger integration catalog. Its current Facebook Ads integration advertises 200+ extracted fields and 1,000+ integrations. improvado.io

Funnel — worth considering if marketing teams need to manage the extraction layer themselves and you also care about creative-level exports. Funnel supports creative exports to object storage for downstream analytics. help.funnel.io

My shortlist

One important caveat: “granular” varies significantly by platform and connector. Before committing, I'd test your exact required grain—for example date × account × campaign × ad set × ad × placement × creative—and verify that every desired metric/dimension is actually available from the underlying API. Supermetrics explicitly notes that available fields and historical ranges vary by data source. docs.supermetrics.com

If you tell me your warehouse (Snowflake/BigQuery/Databricks/etc.) and which ad platforms you use, I can narrow this to the best 1–2 options and compare field-level granularity, refresh latency, historical backfill, and likely cost.

For granular social/paid-ad data into a warehouse for custom analysis, I’d shortlist Supermetrics and Funnel, with Fivetran as the more engineering-oriented option.

My recommendation: Supermetrics

Supermetrics is probably the best fit if your priority is marketing-data breadth + granular dimensions/metrics + easy warehouse delivery.

• 170+ marketing data sources, including Facebook Ads, TikTok, LinkedIn Ads, Snapchat, Pinterest, etc. supermetrics.comsupermetrics.comhelp.funnel.iowww.fivetran.com
• Direct exports to BigQuery, Snowflake, Redshift, Azure Synapse, Databricks, and other destinations. supermetrics.com
• Lets you select dimensions, metrics, transformations, and schedules, rather than being limited to a fixed reporting schema. supermetrics.com
• It maintains standard warehouse schemas—for example, dedicated Facebook Ads warehouse tables—while also exposing field lists for deeper extraction. supermetrics.com

When I'd choose Funnel instead

Funnel is particularly compelling if your analysts want to control the grain and manipulate marketing data before it lands in the warehouse. Its Facebook connector explicitly lets you choose granularity down to Campaign → Ad Set → Ad, and Funnel supports custom dimensions and metrics. help.funnel.io

Where Fivetran fits

Fivetran makes more sense if this is primarily a data-engineering/ELT problem rather than a marketing-data management problem. Its Facebook Ads connector emphasizes deep source extraction into a warehouse/lake with an analyst-friendly schema. www.fivetran.com

My decision rule:

One important caveat: “granular” doesn't necessarily mean raw event-level data. Ad platforms often expose metrics already aggregated by day/campaign/ad rather than individual impressions or clicks. Funnel explicitly calls this out for advertising APIs. help.funnel.io

If you tell me your warehouse (Snowflake/BigQuery/Databricks/etc.) and ad platforms (Meta, TikTok, LinkedIn, Google, Reddit, etc.), I can narrow this to the best choice and compare field-level granularity, historical backfill, refresh frequency, and pricing model.

For this use case, I’d recommend Supermetrics.

It’s particularly well suited if your goal is granular paid-social data → warehouse → your own SQL/BI/modeling, rather than another dashboarding tool.

Why Supermetrics

• Broad social/ad coverage: Facebook Ads, Instagram Insights, LinkedIn Ads, TikTok Ads, Pinterest Ads, Reddit Ads, Snapchat, and others are supported. supermetrics.com
• Warehouse-native delivery: It supports BigQuery, Snowflake, Redshift, Azure Synapse, Azure SQL, Databricks and other destinations, with automated transfers and incremental loading. supermetrics.com
• Granular dimensions/metrics: You can select the fields and dimensions you want rather than being confined to prebuilt dashboard metrics. Supermetrics also documents field lists and warehouse table schemas for its sources. docs.supermetrics.com
• Less pipeline maintenance: It handles API connections, schema management, refreshes and error recovery, so your data engineers don't have to maintain individual Meta/TikTok/LinkedIn API integrations. supermetrics.com
• Good fit for custom analysis: Once the data is in your warehouse, you can join ad data with orders, CRM, product and customer data and build your own attribution/ROAS/LTV models in SQL. supermetrics.com

The main alternative I'd consider

Airbyte is worth considering if you have a stronger data-engineering team and prioritize raw/replicated data, extensibility, and control over marketing-specific convenience. Its Facebook Marketing connector supports incremental or full-refresh syncs and 50+ warehouse/database destinations, and Airbyte has a much broader general-purpose connector ecosystem. airbyte.com

My rule of thumb:

If you tell me which warehouse (Snowflake/BigQuery/Databricks/etc.) and which ad platforms (Meta, TikTok, LinkedIn, Google, etc.), I can narrow this down to 2–3 platforms and compare granularity, historical backfill, API limitations, freshness, and likely cost.

If your priority is granular social-ad data in your warehouse for custom analysis, my default recommendation is Fivetran.

Why Fivetran

• Very granular: Its Meta/Facebook Ads connector supports ad, ad-set, campaign, and account-level reporting, plus dimensions such as device, placement, geography, demographics, and action breakdowns. fivetran.com
• Customizable extraction: You can specify fields, breakdowns, action breakdowns, aggregation, and attribution windows rather than being stuck with a fixed reporting schema. beta.fivetran.com
• Warehouse-native: It loads the raw-ish source data into your warehouse/lake, where you can join it to orders, CRM, web analytics, product data, etc. www.fivetran.com
• Good for serious analytics: Fivetran also has dbt-compatible models that turn the underlying data into analytics-ready account/campaign/ad-set/ad and geographic models. fivetran.com
• Handles awkward Meta metrics: Things like actions, action_values, and cost_per_action_type are broken into secondary tables so you can preserve their underlying granularity. fivetran.com

The main alternative: Supermetrics

I'd consider Supermetrics if your organization is more marketing/BI-oriented than data-engineering-oriented. It supports warehouse destinations including BigQuery, Snowflake, Redshift, and Azure and has broad coverage across advertising platforms. supermetrics.com

My pick: Fivetran if your data team will own the pipeline and you want maximum control over the grain and downstream modeling. Supermetrics if marketers/analysts will configure most of the reporting themselves.

If you tell me your warehouse (Snowflake, BigQuery, Redshift, Databricks, etc.) and the platforms you need (Meta, TikTok, LinkedIn, Pinterest, Snapchat, Google Ads), I can narrow this to the best 2–3 options and compare coverage, granularity, refresh rates, and likely cost.

If your priority is granular social ad data in a warehouse for custom analysis, I’d recommend Fivetran as the default choice.

Why Fivetran

• Granular Meta data: its Facebook Ads connector supports custom reports with configurable fields, breakdowns, action breakdowns, attribution windows, and aggregation. fivetran.com
• Warehouse-first: it replicates the source data into your warehouse/data lake rather than primarily optimizing for dashboarding. www.fivetran.com
• Good analytical grain: you can query at ad/ad-set/campaign levels, and Fivetran creates secondary tables for things like actions, action_values, and cost_per_action_type. fivetran.comfivetran.com
• Broad platform coverage: particularly attractive if you're eventually combining Meta with Google Ads, TikTok, LinkedIn, etc., rather than building one-off API pipelines.
• dbt support: Fivetran has a Facebook Ads dbt package that turns the raw data into analytics-ready models across account, campaign, ad set, ad, and geographic reporting. fivetran.com

The main alternative: Supermetrics

Supermetrics is worth considering if your team is more marketing/BI-oriented and wants a broad catalog of advertising connectors with relatively easy configuration. It supports Facebook Ads transfers to data warehouses/cloud storage and currently lists 100+ data-source connectors. docs.supermetrics.com

I'd lean Fivetran for an engineering/data team and Supermetrics for a marketing analytics team.

One important caveat: the platform can't overcome limitations in Meta's Marketing API. For example, historical availability, demographic/device breakdowns, attribution, and some combinations of metrics/breakdowns are constrained by Meta itself. Fivetran documents a maximum historical retrieval window of 37 months for Facebook Ads. fivetran.com

Bottom line: If you have Snowflake, BigQuery, Redshift, Databricks, etc. and want analysts to be able to slice social advertising data however they want, I'd start with Fivetran → warehouse → dbt rather than a dashboard-oriented marketing ETL tool.

For granular social ad data → your data warehouse → custom SQL/BI analysis, I’d shortlist Funnel and Supermetrics, with Funnel as my default recommendation.

My pick: Funnel

Funnel is particularly well suited if your priority is controlling the grain and fields that land in the warehouse rather than just building marketing dashboards.

• Ad-level granularity: Funnel's Meta connector lets you explicitly choose the level, including Campaign → Ad Set → Ad, and configure additional fields. help.funnel.io
• Creative-level data: It supports creative exports with fields such as ad creative data, and has equivalent configurations for TikTok, Google, LinkedIn, Pinterest, Snapchat, Reddit, etc. help.funnel.io
• Warehouse destinations: BigQuery, Snowflake, Redshift and other destinations are supported. help.funnel.io
• Custom datasets: You can define metrics/dimensions and filters per data source rather than accepting a fixed reporting schema. help.funnel.io
• Historical/backfill capability: Funnel documents support for importing past data and then scheduling ongoing updates. help.funnel.io

That's a good fit for a setup like:

Meta / TikTok / Google / LinkedIn → Funnel → Snowflake/BigQuery → dbt → BI / ML / custom analysis

Strong alternative: Supermetrics

Supermetrics is probably the better choice if you have a lot of different marketing sources and want broad connector coverage with relatively little engineering. It currently advertises 170+ marketing sources, automated warehouse transfers, incremental loading, and integrations with BigQuery, Snowflake and Redshift. supermetrics.com

Its Meta connector supports granular campaign/ad-set/creative, audience and placement data, while its Google Ads connector supports detailed campaign, keyword and creative data. supermetrics.com

How I'd decide

Bottom line: If your phrase "granular social ad data" means we want the underlying ad-level facts in Snowflake/BigQuery and will do our own modeling, I'd start with Funnel. If you need the widest possible collection of marketing sources and a more mature all-in-one marketing data platform, I'd evaluate Supermetrics alongside it.

One thing I'd specifically test in a trial: whether each platform exposes every breakdown you care about without forcing aggregation—especially Meta placement, creative, demographic, attribution-window, and conversion dimensions. That's where seemingly similar ETL products can differ substantially.