Warby Parker's Google ads are 93% text. Gymshark's are 42% video. What five DTC brands ran in one week

Most advice about competitor ads is about Meta. Google's Ads Transparency Center is the free source for Google, and one thing it shows quickly is what format a brand actually leans on. I counted five US DTC brands' ads that were shown between 18 and 24 October 2026.

What stands out

• Warby Parker is a search-text company in this data: more than 2,000 of its ads are text. Its new ads that week were almost all text too (22 of 23).
• Gymshark and Glossier lean on video, and so do their new ads: 6 of Gymshark's 7 new ads and 10 of Glossier's 13 were video.
• Volume is not strategy. Allbirds and HelloFresh show under 100 ads each, and none were new that week.

Caveats

One week, five brands, one country, and a snapshot from 24 October, so this is a method demo, not a trend. "Shown this week" counts ads shown at any point in the window, not ads launched in it. The counts include every ad shown for the brand's domain, which includes 4 to 8 ads per brand placed by other advertisers (HelloFresh 0); keeping only the brand's own account changes the totals by under 1% for each brand (Warby Parker 2,198, Glossier 274, Gymshark 93, Allbirds 89). I did not check whether a paused and restarted ad changes its first-shown date.

Do it yourself

The Transparency Center lets you search a domain, but counting formats by hand across hundreds of ads is slow. With the data exported as CSV (one row per ad), the count is a few lines:

import csv import sys from collections import Counter, defaultdict mix = defaultdict(Counter) for ad in csv.DictReader(open(sys.argv[1])): mix[ad["searchTerm"]][ad["format"]] += 1 for domain, formats in mix.items(): total = sum(formats.values()) print(f"{domain}: {total} ads, video share {formats['VIDEO'] / total:.1%}, {dict(formats)}") 

Save a dataset export as ads.csv (each row has searchTerm and format) and run python3 format_mix.py ads.csv. On my 24 October snapshot it prints:

allbirds.com: 94 ads, video share 7.4%, {'IMAGE': 39, 'TEXT': 48, 'VIDEO': 7} gymshark.com: 97 ads, video share 42.3%, {'VIDEO': 41, 'TEXT': 32, 'IMAGE': 24} warbyparker.com: 2206 ads, video share 3.2%, {'IMAGE': 88, 'VIDEO': 71, 'TEXT': 2047} hellofresh.com: 89 ads, video share 1.1%, {'IMAGE': 18, 'TEXT': 70, 'VIDEO': 1} glossier.com: 278 ads, video share 23.0%, {'VIDEO': 64, 'TEXT': 160, 'IMAGE': 54} 

What I built

I built a scraper for this (Google Ads Transparency Scraper). It costs $0.002 per ad, and has an "only ads shown in the last N days" input. Which brands would you want counted next? Tell me in the comments.

AI assistance: I used Claude to help draft and check the numbers; the data and the Actor are mine.