Introduction: an innovative search engine, and an unfortunate timing

TinEye reverse image search engineI always liked TinEye Reverse Image Search; I admire their original idea and their implementation.
The way TinEye presents the results is invaluable to authors trying to enforce their copyright.

...and I really felt sorry for them when Google Images came up with a nearly identical solution - "search by image"- offered for free, which could put them out of business.

I also admired TinEye's first reaction at Google's coming in the niche (Big G was actually working at the project since years). TinEye had a better technology, but less resources, and Google Images eventually caught up.

The comparison crashes TinEye hands down:

Mona Lisa portraitGoogle Images offers more results, quicker, and for free (TinEye is free for non-commercial usage).
Also has another advantage: having archived the html pages the images are retrieved from, it can tag images using alt texts, anchor text and texts in the proximity of the image. This makes pertinence algorithms possible.

Add the unlimited resources in terms of budget to invest in hardware and research, and it is crystal clear TinEye has been very unfortunate not having enough time to grow enough to establish itself, be known, and well profitable...
Ubi Major Minor Cessat

My unsolicited business advice to TinEye

In order to survive, TinEye has few options:

  1. Be acquired by another search giant (Bing?)
  2. Offer something unique that Google Images doesn't

I have no knowledge of undergoing negotiations – I wouldn't be surprised if they existed! – regarding the first point, but in this occasion I will elaborate on the second option: innovate again.

An unsolved problem to address and make money

There is one thing that neither TinEye nor Google Image do, but would solve a problem and be worth good money:
Once you are presented hundreds of images, you have no clue on whether you are permitted to use them or not (even whether G itself could use them is questionable).

Giving the answer would make the difference

Of course TinEye is well aware of the need for attribution and has been very keen to point out how to find it out using its reverse image search engine, but the suggested approach goes along these lines:
"use TinEye to search for similar images all over the web, visit each referred web site, see if some of them is an image stocks site listing author and copyright notice, or locate the author site among them, contact him...."

...that's a time consuming process which could largely be automated by a machine for a good variety of images.

TinEye should be that machine.
Not an easy task, but not even matching images in real time is!

An interesting search for an interesting example

On its home page TinEye sports some pre-prepared "Interesting searches", used to demo its capabilities.
One of them until a few days ago was "Mona Lisa", which returns a SERP like this:

Mona Lisa image search on TinEye

While the result set is somewhat impressive, it misses something, as it doesn't answer a few simple questions:

  1. What is the subject portrayed?
  2. Who is the original author?
  3. What is the license applied to the picture?

If you want to use images in your blog/magazine/logo/etc., no matter who are, those are the answers you (should) want.

I'm sure TinEye engineers could conceive a why to provide, directly within the result page, answers such as:

  1. The subject is the world famous Mona Lisa portrait,
  2. by Italian artist Leonardo da Vinci
  3. The image is available under Public Domain license on Wikimedia Commons

I wish I could see these two little precious pieces of information right on top of the result page, without leaving to me to dig them on the web following an unknown number of links.

Will TinEye accept the advice?