Forum Moderators: open
Q: Have any European search engine companies been able to establish themselves in one or more European countries to compete with the dominant global players (from Google to Lycos to MSN to Yahoo)?
- I know of Espotting in the paid-listing category, and Fast for general search. But has any Euro-centric general search company been able to attract significant traffic without being a "Google clone"? (Or some other derivative of the global engines?)
[Or is Fast the primary success story here in Europe?]
[webmasterworld.com...]
(I accidentally posted the Dutch view in your introductory thread)
So there are some local search engines left. Most are insignificant in terms of marketshare. A rare exception has been named by Vitaplease, dutch based Ilse.
The main problem for local players seems to be scalability and revenue opportunities. In order to get a competitive crawler engine going you need to develop a technical infrastructure and a competitive algo for ranking results. Only targeting one country doesn't give you any advantage whatsoever in that process.
To get this right is not easy. Even the big international players are wrestling with both aspects.
Google is probably the only player at this point getting it right at both ends.
And frankly, from a users perspective, who needs a local engine, when the international players give them the best local results anyway?
International players all offer localized interfaces. On most portals users don't even realize/care who's feeding the websearch.
Take the example you provided, Kvasir. Websearch at Kvasir is provided by Fast. You can go through all popular portal in Europe and will inevitably find an international player feeding results.
They run their own directory with their own editors and have been quite succesful in that aproach.
In Sweden and Norway similar directories have been built - either with the help of Looksmart or by local editors. The back fills are often from FAST, but the directories have been built locally.
For other countries, the Europe Overview [webmasterworld.com] is a good way to start researching.
And frankly, from a users perspective, who needs a local engine, when the international players give them the best local results anyway?
One of the advantages of for example ilse.nl was next day (free) inclusion in their index, whereas with Google it could take up to 2 months. So the user could have more relevant results.
Now with freshness from Google that advantage is lost a bit, however Google Fresh inclusion is inconsistent or unpredictable.
Ilse with their recent search engine update stopped this next day inclusion, even though they still claim it is supported.
[webmasterworld.com...]
You can go through all popular portal in Europe and will inevitably find an international player feeding results.
Not quite true. AFAIK, Switzerland's search.ch has its own results AND is in the top 3 of SE's in Switz.
It should be noted that local directories - as opposed to crawler based searchengines - do in fact exist in many countries.
A directory is a completely different animal though. You need local editors to do that. Most local directories are part of local portals. So a typical situation would be a local portal, with it's own directory only listing local sites, plus a websearch fed by an international player.
even thow DeepIndex is not yet a big fast neither a big google, and doesn't drive high trafic values in all european countries, it might be considered as player.
In the last days, with the bought of Fast and Altavista by overture, the market got opened. Got more openend because users do not forcely like that all engines are in the hands of 2 or 3 companies. I won't take the political implications in the near east, which might give also sense...
Anyway the market got openend, to other SE's, might be european ones, but also coming from other world area's.
Hope's this give another point of view.
Best regards
Gilbert
<<there is no such thing as a Euro-centric general search company, at least not on a significant scale. Fast has been the only worldwide player that was based in Europe. But they were worldwide, not specifically oriented to Europe. And the comapny has just sold their websearch divison to US based Overture.>>
There are, yes, some/many(?) local engines trying to do interesting things. But none I've found that have successfully transformed their technology or business model into the kind of proposition that made a local portal drop google or fast for a new engine.
There is a unique story in Exalead, inasmuch as they've faced down Google in a major local portal (aol.fr) and survived (so far). Despite AOL going to Google for its results globally, aol.fr stuck with Exalead. Exalead's explanation: customizability, the ability for the portal to generate and serve its own results, rather than someone else's generic ones, and finally the ability, as with other engines, to "geo-localize" results. If someone in France searches for health (sante) in France, they won't see a Canadian health organization (in Quebec) at the top of the list, but the site of an organization that is actually in France.
Then again, their customers are dominantly enterprise/corporate -- not portals across Europe. So any pent up demand for their product is not yet apparent. But the importance of search, long a commodity before Google, is certainly in a state of flux.
I don't know, but perhaps vitaplease's Scandinavian example is similar (Jubii.dk). If you beat LEGO, you gotta be doing something right... ;). But it's so hard to get a handle on all the various iterations of engines and portals and directories in Europe. Seems that few analysts have even wrapped their heads around the whole thing.
<<Only targeting one country doesn't give you any advantage whatsoever in that process.>>
This seems to be the paradox -- the more you succeed at "geo-localiziation," the more you narrow your market. You CAN conceivably help a country portal keep its audience by providing better, more relevant geographical results (when/if appropriate) to its users. Thus they turn their first instead of Google.
But who wants to help only one portal in one country -- unless you can convince many of them to buy your technology together. No one's done this yet. Maybe they will.
But for now, as I will probably write, it's a Google, Google, Google world, and Europe just lives in it... ;)
<<To get this right is not easy. Even the big international players are wrestling with both aspects.>>
That seems to bear out too. No one considers their formula finished (or says they don't). They keep working at what it means to provide a country/place/language with more relevant results, paid or otherwise. Again, flux. More flux than in, say, 2000.
<<Google is probably the only player at this point getting it right at both ends.>>
It is hard not to be impressed with the leaps and bounds they've made in the last two years.
<<A directory is a completely different animal though.>>
It is. It's certainly part of some engines' search equation, but it's just not the same as search.
> it's a Google, Google, Google world, and Europe just lives in it... ;)
Users in Europe don't care for Google being a US company. It has been that way from the beginning of the web as mass medium, think Amazon, Yahoo, Lycos, Altavista Ebay - all huge brands are international on the web.
For web publishers, webpromoters and ecommerce the total Google dominance is obviously a problem. This problem anyhow, being dependant on one channel, is in no way related to US vs local or European issues.
Cough cough, not sure vitaplease gave that example. Not that he couldn't though ;)
>there is no such thing as a Euro-centric general search company, at least not on a significant scale
Agreed 100%. Euroseek.com tried once to cover all of Europe. They didn't make it and sold the technology and domain name for the price of a candy bar afaik. With all the different cultures, languages, habbits and ways of dealing with the net, nothing comes close to providing a euro-centric search engine or portal for that matter. Lycos is trying and making results in some countries. Mind you though that they often build on top of already excisting sites/portals and use local people to run the sites.
Local people know how to give local stuff to local people. That's hard to beat.
Exactly, only that does not pertain to websearch. It's definitely true for directories, and one of the reasons why local portals very very often do websearch combined with directories is the latter are perfect for local results, besides offering an proven revenue model: yellow pages.
As for websearch: that's a technical thing. It's based completely on programming. Specific local knowledge is hardly a factor in programming a search engine. Languages are. But not countries.
A search engine able to handle more than one language can easily handle all languages, or at least all based on similar alphabets.
Basically delivering local results is nothing but a filter operation, or rather running multiple filter operations. It certainly is easier and more cost effective to spider all the web and run those filters for many languages.
Basically delivering local results is nothing but a filter operation, or rather running multiple filter operations. It certainly is easier and more cost effective to spider all the web and
run those filters for many languages.
I am not sure about this heini. It would be better for a localised search engine to start with a good local search index rather than attempting to distill a local country search from a more global index. Though the technology platform used for the global search could easily handle a simple country search. It may be better to generate the country specific search indexes and merge them for the global index.
Though I think that any European wide search engine would be run across the linguistic fragmentation in the same way the Roman Empire did. Latin used to be the common language in Europe (the lingua franca :) ) but eventually it was supplanted by the local languages which it never quite managed to replace. With the web, the common language is currently English. However as more countries get better connected, the diversification of languages will increase.
From a marketing point of view, a Europe wide search engine is a nightmare as it would be fighting marketing battles in every European country. Perhaps Google and the bigger players have the financial and marketing resources to do this but they could also be throwing away a lot of money. It may suit them to have a single macro search engine with country specific filters on the results and marketing the single brand name in the same way that Google does.
In order for a local search engine to compete successfully it would have to do a better job than Google or the other bigger players at a local search index and give better results. But if it is not marketed effectively, it would not be a waste of time and money. Rather than having one single search engine for Europe, a constellation of effective local search engines would probably be the way to do it. Approaching the problem without a properly thought out localised marketing and search index plan would guarantee failure like Euroseek. Google probably has the muscle to market well in each country but they would be competing with every local search engine and directory in each of these countries. I'd guess that pushing Google as the great macro/generic search engine would be the way Google will play it since it has the traffic level to be the significant player in every market. Though given the cyclic nature of the business, Google will have a lot of other big players aiming at its core market and what happened to Altavista will give them sleepless nights.
Regards...jmcc
Since I'm not a programmer, I certainly would be interested in any experienced opinions.
To clarify: It's certainly a natural, or at least normal way for any search engine to start out from a local POV. Working one language. But looking at german or
french SEs, it's also clear they start by indexing languages, not countries. French, as spoken in France, Canada Switzerland etc.
This is a very important point heini and it is probably the fundamental reason why Exalead is able to fight off Google. (For the moment. :)) In terms of countries, Europe is clearly defined but in terms of languages, the demarcations become very fuzzy. At worst, the main problem for Google with the US market is that one of the biggest European markets linguistically speaking, is almost identical in that it is difficult to tell an Irish or UK website from a US website based on language - all of them use some variant of English. It is the same problem that faces Google with a French language website and a Canadian (Quebecoise?) website. Breaking it down on a smaller scale for Europe, you have language splits inside countries, Belgium and Switzerland being good examples. Approaching this from a linguistic viewpoint, the name of CNO website may make sense in the language spoken in a country but it may not actually be related to that country. The French, German and US, UK, Swiss, Belgian, Irish SEs all run into this problem of linguistics.
The way I see it, is that many of the local SEs that try to cover a particular country make the mistake of starting with a relatively poor (on the .com/net/org side) search index. While deciding what is and is not an Irish website based on whether .ie is easy (.ie is a managed registry so some connection to Ireland as well as entitlement has to be proven to get a domain) deciding what is an Irish CNO website is difficult.
To establish the engine on a market, or at least find an audience, quickly country becomes a demarcation.
Agreed. This is mainly from a marketing viewpoint as it is easiest to group the users in a country as being the target market for the advertising and audience.
But, where in this process of programming a spider and a ranking algo do you see the advantage of being a local engine?
Like the old joke about politics (all politics is local) all good local search results are local. Where Google and the big players are excellent is in delivering non-local or generic results. Where they fail is in delivering clear local results. This is a niche that a well marketed and designed local SE can occupy well. The big players can't do local - they only do global. Looking at the evolution of the web, the value of being able to provide local results will become far more important. The targeted sponsored links are the start of this but if you are searching for something and the targeted sponsor link could give you exactly what you want rather than something that may or may not be in the same town or city. If you are using WAP or some other enabled mobile system, and paying for the connection by the second, fast and accurate results will be required. I think that the main flaw of existing local SEs is that while they may have a good local (.de/ie/uk/es) search index, their CNO indexes are poor. The solution for SEs like Google is that they index everything and then try to distill a country or local search with filters. This trying every possibility is known in cryptography as a Brute Force Attack. It is not an efficient way of doing things. With a local SE, starting out with a clear vision of what it wants to achieve (a superior local search index) and going about achieving it will give the bigger players headaches. However from a marketing point of view, these small local players will have a very tough time against the big players. Though if they get the quality right and the small SE results are better, they could succeed in a niche market.
Furthermore, where do you see the advantage a local player might have when tackling more countries, with more than language?
For a crawler based search engine it surely doesn't make a difference if it spiders and ranks the www or lets say 10 countries with perhaps 15 languages?
The biggest advantage that a local player has is that they are local. They would, or should, have a good idea of what is in their country. Google and the bigger players do not have this edge. However with the multiple languages, it would be better for the local player to work on creating a good search index in each local language before beginning the spidering process. The whole crawler based seach engine approach always makes me think of the Shakespeare analogy: An infinite number of monkeys on an infinite number of typewriters all hammering away trying to come up with the complete works of Shakespeare. This mindless crawling will probably produce some good results and the application of the search/ranking algorithm to all this data will produce good results if the input data, the websites and webpages, is good. But if it is poor then no brilliant algorithm is going to turn rubbish into search engine gold. In effect they would run into the Garbage In Garbage Out problem.
In some respects, a local SE is fundamentally different to a large SE. While a large SE will crawl all webpages and depend on the search results/ranking algorithm to provide relevance, the local SE is starting from a very specific point and has clear boundary conditions. Somehow I think that many local SEs, especially the ones that did not succeed, took the same approach as the big SEs.
Regards...jmcc
- - -
a) Conventional wisdom gravitates to 'Search is search, no matter where a particular spider is applied or what language its results appear in.' (Witness distribution of Google's often stupendous, fast-growing victories in traffic rankings across Europe, mirroring its US/global success earlier on.)
b) No major search engine appears to be posing a pointed, well coordinated challenge to Google.
- - -
Q: How do major European portal/search sites compete with Google?
A: What major European portal/search sites are competing with Google?
The most obvious exception is Fast (a search results provider, sans destination site), which has a) managed to fend off Google -- not in traffic, but at least in numbers of country-specific portals that use Fast to power results. And maybe some Overture/Fast/Altavista animal will pose an interesting challenge here/globally down the road.
But it seems Google is just waking the search world up and the signs of anyone yet getting their act together to try to compete are barely beginning to register.
So, in a nutshell, you are saying a local engine could do better with produscing local results, because they would be starting from a local, small, but high quality set of results. Where a global player would apply multiple filters to a global index, the local engine would scale the small quality core up.
Producing truly local results is a problem for global engines. Google is the only engine which has half solved this problem. They use two filters: local domain, and DNS info. Google uses them in either or mode, i.e. one of the two factors must be given. As we, especially jmccormac, have often discussed, both of those have major flaws. On the side of the ccTLDs not much can be done. Basically everyone can register them, even though regulations may not say so. But more important is the problem of assigning locality to generic TLDs. The share of gTLDs throughout Europe is high. So the problem needs to be tackled from the whois/dns info side. If I understand correctly where jmccormack is coming from then that would be to trace domains to exact physical location of the servers.
Anything else?
The biggest advantage that a local player has is that they ... have a good idea of what is in their countryHow is that achieved? You say a local engine needs to have a good set of data before the spidering actually begins - where's that set of data supposed to come from? All I can think of is a good local directory.
In any case, all of this relates to the difference between local engines and global engines. Nothing to do with the distinction of European vs US, or European vs global. I still see this distinction not having any validity in the field of websearch, neither from the users nor from the search engines end. Local vs global: sure. Language specific vs all languages: okay. But Euro vs global? no way.
The real question here is what exactly is the advantage of local results over global results? I see only one answer: ecommerce. For informational and entertainment needs results don't have to be local at all. They need to be in the users language. That's all. Evene if the query is for information clearly related to locality, the query itself will in 99% of all cases reflect that locality. Nobody would search for the opening hours of a library in Ourlittletown by entering opening hours library in a search engine. Ok, correction, many do, but they wouldn't get the correct answer in a local engine either. So, ecommerce drives the quest for local. The search engine as yellow pages.
Q: How do major European portal/search sites compete with Google?
Looking at AOL.fr, who stuck to their partnership with Exalead for local results, it has to be noted that AOL.fr has not gained marketshare from that move. They didn't fall though. From what I heard their users are well satisfied with the local search, but it has not been a key to significantly gain marketshare.
Perhaps a way to really gain a competitive edge over general global SEs could be applications based on crawler based indexes, which process those results and add value for the users. Exact localisation, integration with maps, direct communication applications come to mind.
Until then I fail to see any local engine being really competitive for global engines.
