Race for Google (with abbreviations)

Journalist Miguel Helft considers the main stages of this race, noting that there is a lot of venture capital in the search market - not only to “beat” Google, but also to get your piece of advertising pie. Nevertheless, Helft concludes that despite the attempts of many companies, it is not clear whether they can catch up with Google. Google is not a monopolist like Microsoft, but nevertheless, at the moment there is no company that could overthrow Google.
We all have almost the same feeling about the dominant position of Google - but why? To understand, we segment Google's competitors and their approaches. In this article, we will use Emre's classification to understand why we all intuitively feel that Google will remain the king of search.

Categorization of competing technologies

There are two classic approaches to the development of a new company: the path of improvement and the path of cheapening. In the case of the search market, the cheaper way is impossible. Thus, there is only one way for competing companies - the way of improvement. In this article we will look at three opportunities for improvement:

• Improved technology - better (relevant) results.
• Improved interface - the results are displayed better.
• Vertical / Semantic Search - improvements relate to both technology and interface.

Technology improvement

PageRank is a phenomenon born by Google. Competitors have two ways to “get around” the leader: 1) natural language / AI and 2) the power of people.
In the AI ​​camp, we note two companies - Hakia and Powerset , which are making efforts to improve search results by understanding what you want to find. Although the benefits of understanding natural queries are undeniable, their impact on improving search results is far from being so straightforward.
It is not clear at the expense of what these companies can win. Firstly, the ability to enter regular queries is already implemented by Google and is not a competitive advantage. Competitive advantage can be significantly improved search results, but it is still difficult to imagine. A bit better is quite possible, but you can’t name a significant difference.
Search with the help of people is interesting because it gives one indisputable advantage that is inaccessible to the computer - common sense. del.icio.us is not often presented as a search engine, but it is perhaps the best example of searching with the help of people. The main thing that we get with the help of del.icio.us, is filtered high-quality results. The disadvantage is that we get results that are already popular among people. Indeed, it is quite predictable, but in some cases it can be a significant limitation.
Another problem del.icio.us is manufacturability. You are unlikely to get good results by searching for something, for example, on medical topics. ChaCha, which Read / Write Web wrote in the December review, uses a different approach. This company literally uses people to find the right information. But there are at least two drawbacks - the high cost of the search query and possible problems when scaling.
Although it is indisputable that services similar to del.icio.us are an order of magnitude better than static site directories, the search algorithm is hard to beat with the power of people. The speed and completeness of Google search is out of reach for people.

Improved interface

The motto of Google is simplicity. This is the key idea of ​​all products of the company and it strictly adheres to it. That is why so many competitors are trying to make search results more beautiful.
Snap and Microsoft Image Search show images to improve the interface. Images are really interesting and useful. I personally really like the Microsoft Image Search interface, which is a significant improvement in the display of search results. Unfortunately, both technologies should improve not only the relevance of the results, but also their number. In addition to what is good for image search, not so good for text search - it is difficult to determine the quality of the result by a small image.
Now let's look at the clustering of search results, which is essentially a vertical search, but with all the results. Clusty makes a huge impression by adding value to the search results without changing our perception of them. The presence, in addition to the search results, of the cluster tabs, which refines the results, essentially allows you to get an answer in no more than two steps.
The second search engine, SearchMesh , is owned by Google and the company's “playing field”. It uses Ajax and one of the modes allows you to display only the headers of the search results pages. I find this method surprisingly helpful. You can also “slice” the results vertically by searching blogs, news, or pictures. Both sites (SearchMash and Clusty) are very convenient and can enter the mainstream.
In the last subcategory of search interface improvements, Quintura and Kartoo are located , offering innovative search visualization technologies. Quintura offers “clouds” to provide clusters of search results. In general, the idea is similar to Clusty, but the graphical representation is different. Kartoo also uses a graphical representation for clustering results. The problem is the complex presentation of search results, which is unlikely to allow them to enter the mainstream. Although I have an idea of ​​graphical visualization, I cannot call any of the search engines intuitive or simple. In this incarnation, the visualized search will not become widespread.
Many of these interface improvements are correct and interesting, but none of them are able to beat Google. The main reason is the ease of playback. Of course, from a legal point of view, exact clones will not be created, but if a significant market share is gained by any technology, Google can easily make the necessary improvement. In addition, experiments with SearchMesh can already confuse the maps of some vertical search engines.

Vertical search

Much has been written about the benefits of vertical search, which leaves no attempt to move Google from their homes. These technologies improve search results, the interface, and also focus on a specific vertical. Many players have entered the search market for blogs, ads, products, jobs, etc.
Consider a simple example. If you want to find blogs in web 2.0, will you use Google or Google Blog Search? Even Google recognizes the benefits of vertical search for some topics, such as blogs or music. Thus, the vertical search is already with us and is not going to leave the scene. The key issue is the “piece of cake”, which can bite off the vertical search.

findings

In general, despite significant activity in this sector, Google is likely to remain “king of the mountain” in the foreseeable future. Different approaches will be more or less successful in biting the search cake, but it will take time, perfect implementation, a significant amount of marketing dollars and, of course, a new technology to create a big tooth. This is far from a simple combination of success.
Despite the fact that the first three tasks, although complex, are doable, the improvement of technology still requires an understanding. On the one hand, it should give better results faster, and on the other hand, another player may appear on the market - a personalized search . With this technology, results are not ranked using PageRank, but according to your interests. It is likely that the combination of vertical and search search will be able to give a fundamental improvement in search results even in comparison with Google. Nevertheless, we have already noted above - Google is not going to rest on its laurels.