What is an initial load?

When you connect a Seller, Vendor, or Amazon Ads account, DataDoe requests historical Amazon reports so the app has a usable history. That first backfill is the initial load. After it finishes, DataDoe keeps data current with the usual daily and intraday fetch periods.

Progress in the app is the overall initial-load percent (all historical report windows together).

Why does DataDoe split the initial load into two phases?

Amazon limits how many reports we can request at once. If we requested every historical window immediately, recent orders and snapshot reports would compete with a large backfill.

Phase 1 requests a small, useful set first (recent orders plus reports that only need one window). Phase 2 waits until that set is finished, then requests the remaining history. The app becomes useful sooner, and Amazon quota is spent on the high-value reports first.

What starts immediately?

Phase 1 Amazon reports start as soon as the first enqueue runs. It includes:

• the first 4 weekly windows of sp_api_report_orders (about the last 28 days)
• ads_api_report_brand_metrics
• ads_api_report_reach_frequency_dsp
• sp_api_coupon_performance
• sp_api_myi_unsuppressed_inventory_data
• sp_api_promotion_performance
• sp_api_report_brand_analytics_market_basket_by_month
• sp_api_report_brand_analytics_repeat_purchase
• sp_api_report_fba_fulfillment_customer_returns
• sp_api_report_fba_fulfillment_longterm_storage_fee_charges_data
• sp_api_report_fba_inventory_planning_data
• sp_api_report_fba_overage_fee_charges_data
• sp_api_report_fba_recommended_removal
• sp_api_report_fba_shipment_replacement
• sp_api_report_fba_storage_fee_charges_data
• sp_api_report_flat_file_returns_data_by_return_date
• sp_api_report_listings_data
• sp_api_report_myi_all_inventory_data
• sp_api_report_referral_fee_preview
• sp_api_report_reserved_inventory
• sp_api_report_restock_inventory
• sp_api_report_seller_feedback_data
• sp_api_report_v2_settlement_report_data_flat_file_v2
• sp_api_report_vendor_forecasting_retail

These sources are not gated on phase 1 and start on their own schedules:

• finances transactions
• catalog
• FBA inbound shipments
• Data Kiosk

Recurring (ongoing) report schedules start their clock on the first enqueue pass, the same as before. They skip that first run and then follow their usual cadence.

What waits until phase 2?

Everything else in the initial report backfill waits until every phase 1 report for that connection has finished (processed, failed, cancelled, or dropped). That includes:

• remaining order weeks (about 2 years minus the first 4 weeks)
• archived orders
• multi-window Ads backfill
• vendor reports with many periods
• FBA ledger, removals, AFN shipments, sales-and-traffic, and similar multi-window reports

Phase 2 rows are counted in overall progress from the start, so the percent does not jump to 100% after phase 1 and then drop when the rest of history is added.

How do I know when the initial load is done?

Use the overall progress shown in the app. DataDoe also treats the connection as still importing until there is no pending initial-load work, including reports that are waiting in phase 2.