Ecommerce ops · Production

GetYourGuide integrates Generative AI into activity provider onboarding to accelerate content creation and improve quality

The problem

Activity providers on GetYourGuide's marketplace spent up to an hour manually completing a 16-step product creation wizard, and the content they produced often contained missing or contradictory information that confused travelers, hurt conversion rates, and increased care department contacts.

First attempt

A first full-scale AI experiment was closed without launch because the UI failed to explain the AI tool, causing activity providers to drop off; providers in the treatment group also spent longer on non-AI pages out of frustration, and the planned A/B measurement approach proved unreliable.

Workflow diagram · grounded in source

Provider pastes existing content

trigger

“activity providers to paste their existing content (such as from their own website) into a designated input box”

AI generates descriptions and fills fields

ai_action

“We then do some AI magic to generate both the longer free-text content sections, like the full description, as well as fill out many of the structured fields, like what kind of transport is used and what locations should be tagged”

Provider reviews AI-completed steps

human_review

“The AI auto-completes 8 key steps, allowing activity providers to review content in minutes, reducing time spent on manual data entry”

Activity onboarding completed

output

“activity providers were more likely to complete onboarding for their activities”

Content quality drives activity performance

feedback_loop

“the higher quality content meant we saw a solid increase in activity performance for activities onboarded through AI”

Reported outcome

After UX and LLM refinements, all success metrics increased: activity providers were more likely to complete onboarding, content quality improved yielding a solid increase in activity performance, drop-off from the AI input page fell by 5 percentage points, and the feature was launched to 100% of the user base.

Reported metrics

Manual activity creation timeup to an hour

AI auto-completed wizard steps8 key steps

AI feature adoption rate60%

drop-off reduction from AI input page5 percentage points

Show all 9 reported metrics

manual activity creation timeup to an hour

AI auto-completed wizard steps8 key steps

AI feature adoption rate60%

drop-off reduction from AI input page5 percentage points

fastest observed onboarding completion time14 minutes

feature rollout coverage100%

overall success metrics resultincrease in all our success metrics

activity performancesolid increase in activity performance

supplier satisfaction with AI toolvery very convenient to use and very helpful

Reported stack

LLMHotjar

Source

https://www.getyourguide.careers/posts/revolutionizing-the-experience-marketplace-the-dual-impact-of-generative-ai-on-suppliers-and-consumers

Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

After UX and LLM refinements, all success metrics increased: activity providers were more likely to complete onboarding, content quality improved yielding a solid increase in activity performance, drop-off from the AI…

What tools did this team use?

LLM, Hotjar.

What results were reported?

Manual activity creation time: up to an hour; AI auto-completed wizard steps: 8 key steps; AI feature adoption rate: 60%; drop-off reduction from AI input page: 5 percentage points (source-reported, not independently verified).

What failed first in this deployment?

How is this ecommerce ops AI workflow structured?

Provider pastes existing content → AI generates descriptions and fills fields → Provider reviews AI-completed steps → Activity onboarding completed → Content quality drives activity performance.