Trends come and go, this is true for the seachengine-sector as well as for other sectors. Regularly we hear of a new Google-killer (does anyone still remember Cuil or WeFind?) and we hear about the semantic web at least a few times a yea and on top of that, we want the results to become more relevant all the time. One of the few developments that managed to stand its ground in the last few years is the integration of vertical searchengines into the regular SERPs: Google calls it Universal-Search, Bing is doing something similar in the United States and we can assume that Yahoo will follow suit with something, too.

What is interesting is the different approaches that Google and Bing take. Microsoft decided to purchase interesting companies and to push technologies as well data-sources into the SERPs as exclusive partners. In the United States, we can see this nicely with Farecast and after the takeover of Ciao, we can expect something similar to happen here in Europe in the shopping-sector. Google takes a different approach and collates data from different sources, tries to strukture them and use them to fuel their vertical search. The shopping- and local-search is a nice example for this.

For Google's method to work well they need as much data in structured form as they can get, so that they can show more than a simple textual search. Part of this data-base can be crawled from the web, where they probably use customized crawlers for large domains, but it would be far more comfortable to have webmasters make the data available for them. Luckily, there are a number of current developments that try to do just that: Microformats and RDFa. Both of these technologies have in common that they can mark certain data in already existing HTML so that a searchenginecrawler can automatically pick them up. For user-ratings, this means that the name could be noted as text, while the rating itself could be on a scale. This is a nice idea in principle and Google is forcing the signaling by showing those pages that already integrate these tags a little more prominently in their SERPs (Rich Snippets).

Sadly, there is a catch: once Google has collected enough data to build its own product, they start being a competitor. This can be nicely seen in progression of the shopping-vertical: at first (back in the day it was still called Froogle) shop-owners were asked to upload their data both structured and in a format that was convenient for Google. After that, the integration into the Google-SERPs started: shopping got increased visibility and additional shops entered their data. Because Google wanted to have those informations as complete as possible, they tweaked the algorithm so that shops ranked better when they filled out lots of fields. Last year saw the coup-d'etat: the integration of links that went straight to the shops had to make way for Google's own price-comparison. Those shops that originally provided all those product-informations to help Google set up their product are not guaranteed to show up anymore.

What's interesting is this: “How do I deal with Google's wish for more structured data?”. Seeing how we can expect Google to move into many different markets in the future (real estate, cars, jobs, classified ads, ratings and so on), there are many large German websites that should be concerned by this. I am still undecided: on the one hand you are not going to be able to stop Google from setting up their own product and you will probably see an increase in traffic during a transitional phase. On the other hand, you are helping out a future competitor, that has proven in the past that they make very few mistakes and are quick to learn from the ones they make. What is your take on this?