Google images have launched some improvements to their interface. This blog post looks at what these changes are, how they will help users, and what improvements should be done next.

Firstly let’s look at what user-need ‘google images’ fulfils. There seems to be 2 types of search tasks for google images:

Scenario type A – specific and basic

The most popular searches in google images seem to be either abstract keywords such as girls, tattoo, wallpaper. and the fastest rising at any one time are either the most popular celebrities or seasonal celebrations. It is reasonable to assume that users are indeed looking for images regardless of size, context etc.

source: google Insights for image search [ http://www.google.com/insights/search/# ],

Scenario type B – Ambiguous and deeper

I postulate that there is also a heavy user-base looking for the images in the long tail who are searching for images fitting niche requirements; for a presentation, for a very specific item, to put context around a certain concept etc.

The above two mental models are very different, and the interface and flow should cater for both sets; Scenario A require a straight-forward simple shallow task flow, and Scenario B require deeper interrogation of the results-set (assuming they haven’t already completed and advanced search).

————————————————————————————————————

This week there are 3 main changes to google images:

1 – The number of results per page has increased (up to 1000).

2 – The image text information has been moved to a hover pane.

3 –Clicking on the image takes the user to an overlay with the image on the overlay and then website behind it.
[see the full release article at The Official Google Blog]

The first 2 new enhancements focus on improving the usability of the results set and satisfy scenario A

1 – Increasing the number of results per page view is an obvious improvement. The old site only allowed one small screen of results and users had to paginate frequently which made it difficult to compare images. The next improvement could be to allow users to select how many they would like to see on each results page, and how the page can be presented: large icons and no text, small images and text etc. Users like this flexibility to suit their needs and a lot of leading image and shopping sites have this facility.

2 – The new hover pane is also a familiar feature as users like to preview more information before making the commitment to click through. The relevancy of additional textual information around the images will depend on the task in hand. In general, users in scenario A are interested in reels of images of their favourite celebrity or topic whereas Group B require context around the image which they now must hover over to access.

The main criticism on the forums is that the padding between images is too small and hurts the eyes when a user is searching through lots and lots images for any amount of time. Take a look at www.bing.com for a similar results presentation with large padding.

3 – Now, when you click on an image, the 90s-style frame holding the image above the website which contains the image has been replaced by the image overlaying it. Right away the user can see the image in some context. They can choose to explore the underlying website by closing the overlay, go back to the results, or take the image from the overlay (copyright allowing). This is a good improvement and makes the task flow more efficient for both scenarios A and B.

In summary the changes help all users (padding issues aside). However I would argue that the changes are basic improvements, not ground-breaking enhancements that I would expect from google.

————————————————————————————————————

What next?

So, what helps a user complete a task?

1. A great search facility
2. Relevant results list which is easy to scan
3. Tools to hone in on specific results such as filters, sorts, tags and related searches etc.
4. An easy flow from the results to the item

Google are constantly improving the first item (1) matching search and results via their algorithms and image recognition. The latest google images changes improve (2) the navigation and scanning within the results set, and (4) the flow to the image item has been improved. Next I would like to see enhancements for (3) by further extension of the filter function, and extending the ‘see similar’ and ‘related searches’ functions with semantic relationships. This will allow the user to hone-in on particular results thereby reducing tedious scanning through huge lists of eye-watering images.

And a more relevant and targeted set of results will definitely help their new targeted ad placement service based on the type of image the user wants. More income for google means more money to re-invest in improving the service. It’s a win-win situation!

There are formal channels for submitting feedback on google images – that’s where I’m off to next…. care to join me?