To keep campaigns new and relevant, Poshmark leans on AI to enhance product images and refresh video assets. The fashion resale platform has also employed AI to help it translate TV ads into different languages to scale campaigns across countries.

Charlene Phan, director of growth at Poshmark, shared how AI has helped the company become more agile without spending more on talent or resources at Advertising Week New York 2023.

1. Enhancing product images

“In the fashion industry, you have a ton of different products and you want to test different backgrounds,” Phan said.

For example, a brand may want to sell a dress that was initially introduced in the summer with summer imagery in the product photo.

“But now it’s winter and you want to relist this item,” said Phan. “AI can [change the model’s] sandals into boots quickly and seamlessly.”

This way, brands can sell off leftover inventory without having to shoot new campaign assets, saving money and time.

2. Translating TV ads into another language

After launching a connected TV campaign in the US, Poshmark wanted to scale the campaign to other countries, said Phan.

“[We were interested in] launching in Canada. But in Canada, they speak French as well as English. So we needed to come up with additional creative for us to test.”

With AI, Poshmark was able to generate a voice-over version of the English commercial in a French Canadian dialect, which the company could then test in-market.

“The process itself is amazing to see,” said Phan. “The agency came back with a list of versions it could produce. If you want it in Spanish, do you want it in Colombian Spanish or in another dialect?”

3. Iterating and optimizing campaigns

AI can help marketing teams scale creative campaigns more efficiently without having to invest in additional talent or resources.

“You can come up with a handful of creatives and then AI will help and generate a few dozen variations of the same ad copy for you to run and test with,” said Phan.

This is especially helpful when it comes to video content, which Phan said should be refreshed every two to five days.

“Our team was always [trying to figure out] how [to] iterate on [video]. How do we [generate] new copy? How do we change out the music and then increase the output of [this type of] content that TikTok and Facebook are hungry for?”

A word of caution: AI can be incredibly helpful to marketing teams to jump-start the creative process, but it still requires human oversight to be successful.

“You need a human copilot with AI, you can’t let it run wild,” said Phan. “I think that’s one of the key learnings. If you’re just starting out with AI, I would look at your entire creative process and then figure out where you are lacking resources or areas that AI can pinpoint and really help you speed up development.”

This was originally featured in the Retail Daily newsletter. For more retail insights, statistics, and trends, subscribe here.