The Idea Behind

Why we built aiPhotoSearch and why you should use it to find your next stock photo.

Why The World Needs A Better Stock Photo Search

I only bought stock photos here and there until last year but I had to buy a lot this past year. That is how I came across the inefficiences of the current stock photo search process:

  • Multiple marketplaces with different inventory and price points that you have to search individually to get the perfect photo at the best price
  • Lots of photos for your keywords that you need to comb through manually with no easy way to narrow down your results
aiPhotoSearch solves both of those problems. First, using API. Second, using AI.

How To Use aiPhotoSearch

You begin your photo search just like you always did - using keywords. As an example, lets say we are looking for a picture of a dog to include in our landing page to improve conversion (e.g. Can a Puppy Sell a CMS?). Lets start our search by using the keyword "cute dog".

For each search, we use APIs to grab results from multiple stock photo marketplaces and present the result them in one place. Each photo has a heart icon on the top left that you can click to like the photo. By liking a photo, you are telling the AI to look for more photos with similar features. By not liking a photo, you are telling the AI to stop showing photos with similar features.

In our example, I am going to like just one photo - the dog above a white banner. It is a good fit for our landing page use case. If there were other photos with banners, I would like them as well but there was only one photo with banner in this case. Note that the results below are from ShutterStock and Pixabay.

After liking all the photos that are similar to what we want, the next step is to click on Refine Results. This is the step that actually invokes the AI algorithm. The AI analyzes the photos you have liked with respect to the ones you did not and based on that ranks the remaining photos. The result is that you start seeing photos that are closer to what you want.

The picture below shows the refined results based on our previously liked photos. Note that I have liked the new photos with banners in them. This is to further refine the search results in the next step.

We can continue liking photos and Refine Results. Here is the result for our "cute dog" search after a couple of iterations of refining results.