Why 5% of Vauxhall's Meta ads drove 80% of their reach in the last 30 days

5.1% of Vauxhall's Meta ads drove 80% of their total reach in the last 30 days. Five ads. Out of about a hundred. The top five accumulated 25.07 million lifetime reach between them. The remaining 95 plus ads combined for roughly 6 million. That is the shape of the entire UK automotive brand's paid social output in May, condensed into a single ratio.
The Pareto, measured
Most marketers feel a top heavy distribution in their own account. Few actually measure it across a competitor's library. Running Vauxhall's Meta page through Adsmom's MCP this morning, the numbers came back cleaner than most internal audits I have seen. The top ad, a Grandland image, pulled 13.08 million reach across its three days live. The second, a different ad ID carrying the exact same primary text, pulled 6.50 million over the prior fortnight. The third, a video version of that line, did 3.58 million. Add a Corsa GSE image (1.13 million) and a Mokka rally video (0.77 million) and you have 80% of the brand's reach in the entire window.
The median ad in Vauxhall's last 30 days reached roughly 13,000 people. The top ad reached 13.08 million. A thousand to one gap inside one brand's account.

A thousand to one gap inside one brand's account.
The single 24 hour spike
The number that makes this concrete is what happened on May 30. The current top ad, a static Grandland image with the line "The Vauxhall Grandland combines GRAND design with premium tech, both inside and out. Book a drive to experience it for yourself!", reached 9,465,589 people in 24 hours. That single day represents 30% of the entire 30 day reach for the brand. From one ad ID. From one image. From a Friday.

Most teams would treat that as a glitch, scale it back, move on. Vauxhall did the opposite. They had already been recycling the exact same primary text. The first ad to carry it (a video) ran May 11 to 14 and pulled 447 thousand. The second (also video) ran May 14 to 17 and reached 3.58 million. They then switched the asset to an image. That image ran May 18 to 30 as ad 267459e3 and accumulated 6.50 million reach. The fourth ad ID, 2d413788, ran only May 29 to 31 and pulled 13.08 million, including the May 30 spike.
Four containers, same copy, 23.6 million combined reach in three weeks. Reach scaled roughly 29 times from the first ad to the fourth. The shift from video to image accounts for some of it. The cleanest reading is that the team trusted the copy enough to put it back into the auction four times in three weeks. That is rare. Most teams discard a creative the moment a new brief lands.

What this means for the brands sitting next to Vauxhall
Pareto in paid social is not a new finding. The 80/20 has been folklore for as long as digital media has existed. What is new is that the shape can now be measured for any brand, in any market, in a single prompt. Cross checking Vauxhall against any competing UK auto page, the question gets practical fast. If 5% of your account is driving 80% of the reach, what is the other 95% costing you? Is the long tail a deliberate test bed, or is it the wreckage of briefs that did not work?
There is also a methodological point. If you only checked Vauxhall's weekly report top line (4.85 million reach growth across the most recent week), you would not see either the 5% concentration or the 9.46 million single day spike. The shape only becomes visible once you pull every ad in the library and rank by reach, then chart the cumulative curve. That is the part our MCP makes practical. Pull a brand's full ad set, sort, compute the Pareto, and move on.
If next week's Vauxhall data shows a fifth ad ID with the same primary text scaling past 13 million, the recycling tactic is intentional. If not, May 30 will stand as a particularly clean illustration of how one creative carrying one line can carry an entire month.
Adsmom Explore pulls every Meta, TikTok and Google ad a brand has run, then lets you rank, filter and chart by reach. Pulling Vauxhall's full 30 day library for this post took a single prompt.
See Adsmom Explore