Case #027: The pivot that wasn't a country

Case #027: The pivot that wasn't a country

We were running a VPN offer in Germany. $2,600 spent, $2,200 back, a slow 15% loss that wouldn't quite die or quite win. The obvious move was to test a new geo. We did something stranger.

The campaign was German-geo but English-language creatives and an English lander. It half-worked — Germans with strong English converted, everyone else bounced. The data hid this because it was all bundled as one 'DE' campaign.

Instead of changing the country, we changed the language inside the same country. We translated three creatives and the lander into German, ran them against the English originals, same geo, same budget split.

The German-language set, first five days:

— cost per sale dropped from $31 to $19
— conversion rate on the lander climbed from 1.9% to 4.1%
— $1,300 spent, $2,050 back

The English set in the same window stayed underwater at a 12% loss. Same country, same offer — the only variable was the language, and it was the whole game.

We killed the English creatives and ran German-only:

— $9,400 spent over the next two weeks
— $14,100 revenue
— 50% ROI

The lesson: a geo pivot isn't always about moving countries — sometimes the country was right and you were just speaking to it in a foreign tongue.