How We Built a Medical Tourism Startup?
July 7, 2020Our first steps why we choose this idea to help people reach medical institutions globally from anywhere, and how we got started in the first steps.
How We Built Image - SEO Fairy
Problem space. On the medicaltourism.review platform we have more than 400 representatives who upload clinic photos every day - both to their listings and to the treatment packages they advertise on our marketplace. Over time this has grown into tens of thousands of images, 99% of which had no description at all.
Before AI, we considered generating automated, context - based captions, but quality never reached the bar that justified the effort. That changed when AI became good enough to produce marketing - friendly, SEO - ready descriptions.
We’ve always wanted to help representatives maximize their marketing and SEO performance. Manual captions, however, would have required extensive training materials and a lot of tedious work - yielding results that rarely captured each image’s full potential.
Solution journey. When OpenAI GPT arrived, we tested it on increasingly complex tools, then looked for “low - hanging fruit” we could implement quickly. Image descriptions were a perfect candidate - but only if we could keep effort low for both our SEO/content teams and the representatives.
After experimenting with many prompt iterations, we landed on a workflow that fed GPT both the image and its contextual listing data. The output had to be SEO - friendly, descriptive, readable, and - most importantly - sound good to us.
Testing sprint. Once the quality met our standard, we rolled the tool out platform - wide and generated sample captions for review.
Following several rounds of tweaks and spot checks, we reached a point where the vast majority of descriptions were accurate and useful. A few images were inevitably hard to describe, but overall the results exceeded expectations. Confident in the quality, we ran the tool across the entire image library in one batch, then audited random samples for hallucinations or errors. We found almost none.
Results & impact. The payoff was immediate: representatives who had invested time uploading images now enjoyed fully optimized captions with zero extra effort, yet retained the option to edit them whenever they liked. In practice, few felt the need - the AI captions already reflected SEO best practice, and most marketers lacked the expertise to improve them at scale.
Patients benefited too. Better captions made clinic and package images easier to discover via Google, which prioritizes visual content over plain text. Listings with optimized images saw higher content - rank scores, improved search positioning, and better click - through rates - lifting overall site SEO in the process.
Leadership lessons. I always push for solutions that deliver high quality with minimal ongoing maintenance. Creative thinking can meet rigorous standards, delight users, and avoid needless bureaucracy. Image - SEO Fairy is one more example of how clear goals guide a team toward elegant, labor - light answers. Set the expectation - high quality, low manual effort - and the right ideas surface.
Find out more about AI-SEO Fairy on HumanFace Tech >>>