How to Export More Than 1000 Rows From Google Search Console When Your Store Has Thousands of Product Pages

    Getting full query and page data out of Search Console for a large product catalogue

    Published
    9 min read
    By Elara Keldovik
    Updated

    Most store owners who hit the 1,000 row limit in Search Console assume the problem is the export button. They click "Export", get a CSV with exactly 1,000 lines, and start looking for a paid tool that will "unlock" more. The button is fine.

    The limit sits in what the Performance report is willing to show in a single table, and for an online shop with 3,000 or 30,000 product URLs that means the file you download covers your best sellers and almost nothing of the long tail where most catalogue pages actually live.

    The good news is that every route around the cap is free at the scale of a small or mid sized store. The less good news is that each route answers a slightly different question, and two of them quietly drop data in ways that make totals disagree. This guide walks through the options in the order a store team usually meets them, with a worked example and a final checklist.

    Where the 1,000 row ceiling actually sits

    The Performance report shows one dimension at a time in its table: queries, pages, countries, devices, search appearance or dates. Each of those tables is capped at 1,000 rows, and the export copies what the table can display. So a store that ranks for 22,000 distinct queries over the last 3 months will see the top 1,000 by clicks and nothing else in the download.

    Three details matter for a catalogue site. First, the cap applies per table, not per property, so the same export gives you 1,000 queries and separately 1,000 pages. Second, the chart totals at the top of the report are not limited, which is why the sum of your exported rows is always lower than the headline clicks figure.

    Third, some queries are never shown at all because Google filters rare or potentially identifying searches for privacy. On a shop with many long product names, that anonymised share can be large, and no export method brings those query strings back.

    Data retention is the other boundary. The interface and the API both keep 16 months of history. If you want year over year comparisons for Black Friday, you need to start storing exports before the older months roll off, which is a point in favour of the automated routes below.

    Slicing by catalogue folder: the no code route

    The simplest way to get more than 1,000 rows out of the interface is to ask for many smaller tables. Each filtered view has its own 1,000 row cap, so ten filtered exports can hold up to 10,000 rows. A store has a natural way to slice: its URL structure.

    If your categories live under paths such as /shoes/, /bags/ and /accessories/, add a Page filter for each folder and export the Queries and Pages tables one folder at a time. The Page filter accepts "contains", "does not contain", "exact URL" and a custom regex option, so you can also split a large folder by brand slug or by the first letter of the product handle. A regex like ^https://shop\.example/shoes/[a-f] catches one sixth of the shoe catalogue in one view.

    Date slicing is the second lever. Seven exports of one month each will show different long tail queries in each month, because the top 1,000 changes with seasonality. Merging them gives a fuller picture than a single 6 month export.

    The weakness is labour and overlap. You will be pasting 20 to 40 CSV files together, deduplicating rows that appear in several slices, and you still cannot see query and page together in one row, because the interface table only holds one dimension at a time. This route works for a one off audit. It is a poor fit for anything you plan to repeat weekly.

    The API, Looker Studio and Sheets add ons compared

    Once filtering by hand gets tedious, the next step is a tool that talks to the Search Analytics API. The API returns up to 25,000 rows per request and lets you page further with a start row parameter. It also lets you combine dimensions, so one row can say "query X sent 14 clicks to product page Y from mobile in Germany". That pairing is what a merchandiser actually needs to decide which product description to rewrite.

    Route Cost Rows you can reach Query and page in one row Setup effort History kept
    Interface with folder filters Free 1,000 per filtered view No None 16 months
    Looker Studio connector Free Well beyond 1,000 per table Yes About 30 minutes 16 months
    Sheets add on using the API Free tier Tens of thousands per pull Yes About 15 minutes 16 months
    Own script on the API Free Up to the documented daily ceiling Yes 1 to 3 hours 16 months unless you store it
    Bulk export to BigQuery Free tier usually covers a small store Everything Google exports, daily Yes About 1 hour As long as you keep the tables

    Looker Studio has a native Search Console connector with two tables: Site Impression and URL Impression. Build a table chart with Query and Landing Page as dimensions, set the rows per page high, and export to CSV or Google Sheets from the chart menu. Sheets add ons such as Search Analytics for Sheets wrap the same API in a sidebar and can schedule monthly backups into a tab.

    Google documents a daily cap on how many rows the API exposes per search type for each property, in the tens of thousands. Most stores never touch it, but a marketplace style site with 200,000 URLs will, and that is where the next option starts to make sense.

    Worked example: a 14,000 SKU store and its long tail

    Take an illustrative home goods shop with 14,000 product pages, 600 category pages and about 40,000 organic clicks a month. The default Pages export shows 1,000 URLs. In this example those 1,000 URLs account for 29,000 clicks, or 72.5 percent of the total. The remaining 11,000 clicks are spread over pages the export never lists.

    Now the team splits by folder. Kitchen, bathroom, lighting, textiles and garden each get their own filtered export, and lighting is split again by brand because it holds 4,200 products. That gives 9 exports and 7,350 unique pages after deduplication, covering 37,400 clicks, which is 93.5 percent of the headline figure. The missing 6.5 percent sits on pages with one or two clicks each.

    Next, the team pulls query plus page pairs for the same 3 months through the Looker Studio connector: 61,000 rows. Filtering for product pages with more than 500 impressions and a click through rate under 1 percent returns 412 URLs.

    At an illustrative average order value of 68 euros and a 2 percent conversion rate, lifting those pages to a 2 percent click through rate would add roughly 2,900 clicks and 58 orders a month, about 3,900 euros in revenue. None of those 412 pages appeared in the original 1,000 row file.

    Choosing a route: six yes or no questions

    Work through these in order and stop at the first answer that sends you somewhere.

    1. Is this a single audit you will not repeat for at least 6 months? If yes, use folder and date filters in the interface and accept the manual merge. If no, go to question 2.
    2. Do you need query and landing page in the same row? If no, filtered exports may still be enough; go to question 5. If yes, go to question 3.
    3. Is anyone on the team comfortable with a Google Sheet that refreshes from an add on? If yes, start with a Sheets add on and schedule a monthly pull. If no, go to question 4.
    4. Do you already build reports in Looker Studio? If yes, add the Search Console connector there. If no, a Sheets add on is still the lighter choice, so revisit question 3 with help from whoever owns analytics.
    5. Does the property have more than about 50,000 URLs with impressions, or do you need more than 16 months of history? If yes, go to question 6. If no, the API through a Sheets add on or a short script will cover you.
    6. Can someone create a Google Cloud project and attach billing, even if the free tier ends up covering the cost? If yes, turn on bulk export to BigQuery now, because it only collects data from the day you enable it. If no, run a scheduled API script that writes to Sheets or a database and keeps the history for you.

    The order matters. Teams often jump to BigQuery for a 3,000 product shop where a Sheets pull would answer every question in an afternoon.

    Checklist before you trust the export

    Whatever route you choose, run these checks before the data goes into a merchandising or content refresh decision.

    • Compare total clicks in your export with the chart total for the same dates and search type. A gap of 5 to 15 percent on a store is normal because of anonymised queries; a gap of 40 percent means a filter or date range is wrong.
    • Confirm the search type. Web, image, video and news are separate, and shopping heavy stores often get meaningful image traffic that a web only export ignores.
    • Check whether the property is a domain property or a URL prefix property. A URL prefix for https://www will miss a separate shop subdomain.
    • Deduplicate on the full URL plus query, after stripping tracking parameters if your platform adds them to canonical product links.
    • Note the data freshness. The most recent 2 to 3 days are usually incomplete, so cut them from weekly comparisons.
    • For BigQuery, use the URL level table when you analyse product pages and the site level table when you compare totals, since the two aggregate impressions differently.
    • Store a dated copy. History older than 16 months is gone from every Google interface.

    With those checks done, the long tail stops being a guess. What nobody has a clean answer for yet is how much of a store's growth hides inside the anonymised slice that no export returns: if 30 percent of your impressions come from queries Google never names, how confidently can you say which product pages need new copy and which simply have no demand?

    Related Articles