This article aims (humbly) to return to the history of the search engines that have existed in France since the creation of the Web (Lokace, Ecila, Deepindex, Exalead, Orange, Xaphir, Qwant …), a p I have been fortunate to live from its origins since I work on the Web and have been passionate about research tools since 1993. The idea will also be to try here to learn from the past and to understand why all these projects were failures and identify what has always lacked these tools to really break through, over the years …
Small preamble: in this article,
- I will only talk about engines that have developed their own search technology (algorithm / index), thus excluding a myriad of m tamoteurs, tools using existing technologies (mostly those of Google or Bing at present ) or creators of software overlay, as interesting as it is (Kartoo …). Even if I will come back to the end of this article …
- I will only talk about engines of French origin, so exit Lycos, Inktomi, Hotbot and other AltaVista (even if, for the latter, several French were part of the team, but the engine had above all a vocation US and was clearly created across the Atlantic, for Digital at the time).
- I will only talk about general search engines (the whole web), and not those (very numerous), limiting themselves to a single approach or a single topic.
- I will not speak of the directories (Nomade, the sailing guide, etc.), whose operation was very different from that of the engines themselves.
Another important point: note that I may not agree with me about what I write in this article, I understand it easily, but I set out here my personal opinion which I share with me- mrs …
The pre history: the 1990s
This having been established, we therefore begin with the beginnings of the Web and the 1990s. France has, at that time, few websites and websites, few pages (for the anecdote, my first book, Internet guide to connexion, published in 1994, included in the appendix the exhaustive list of French websites in a few pages (which were not written in small print). At that time, we first find Lokace (1994-2002), an engine that has never shown itself to be very relevant in its results, which certainly quickly halted its fall. Apparau00eet in the crowd e Ecila (1996-2001), a little more relevant, but who never found his audience, either.
In this first decade, these two engines never drilled for a simple reason: the others (mostly American: Altavista, Lycos, Webcrawler, Infoseek, Excite, etc.) were better, even on the French-speaking web! And directories were often preferred to search for French sources of information. That said, these ancestors were pioneers in the field and participated in the discovery of the Francophone Web in their own way.
Ecila engine home page. Source: Internet Archive
2000s: Deepindex and Dir.com
In 2002, the DeepIndex boat, a project set up by Gilbert Wayenborgh, the first real attempt at a French search engine. In its wake, Dir.com, created in 2003 by two Iliad / Free employees, tried its luck. I must say that this last engine has always left me with excellent memories. First of all, its designers were very nice and we even did some joint projects at the time. In addition, their engine was really of good quality, while they had very few means (to say almost no means), Dir.com having been torn up within Free, a little in full view of everyone. They were not then tired of being followed by their parent company, leaving the technology abandoned after a few years when it seemed to me to be really good.
Deepindex, for its part, will hold out a little longer, but this courageous project will also abandon the game a few years later, in view of the significant investments necessary to continue the adventure at a time when economic models for an engine of research were still very vague (the Espotting / Miva and Mirago advertising networks also experienced many problems at that time, hampering the engine in their cart).
The home page of the Dir.com engine. Source: Internet Archive
Time for regrets: Orange and Exalead
Among the actors who will become major thereafter, one can well on quote KE, search technology resulting from Echo and which will equip many engines of France Telecom then Orange, and in particular Voil and LeMoteur. This algorithm + index pair, developed by an enthusiastic, sympathetic and motivated team from Sophia Antipolis, was really of good quality and often ahead of its time (sometimes too much, by the way). With a weary team reduced over time and very little resources, the Orange Search team has often worked miracles while the tool was primarily seen by its leaders as a possibility for the business of ‘insert in-house (natural and advertising) links in search results, procedures prohibited by Google. And then, one day (2017), Orange cut the cord and passed its engine on Google technology. The KE engine remains a big regret for me, because it really had its place on the French market and it could certainly have become profitable if it had been helped by its hierarchy this time. Shareholders have decided otherwise
We can’t really ignore Exalead, created in 2000 by Fran ois Bourdoncle (author of the Cow 9 / Refine tool on Altavista l ‘poque in particular) and Patrice Bertin. A beautiful engine, a beautiful story. We believed hard as iron Until their acquisition by Dassault Syst mes in 2010, which will sound the death knell for the engine. It is now spinning in a corner of the Web (the cemetery of the search phants?). For Exalead too, I have a big regret because it was a real beautiful engine, which could really have a good future. But perhaps its designers have never known (or been able) to make a choice between their activity of development of intra-site engines and of web search engine proper. But here too, investments to stay on the road as an engine – necessarily global – could also be considered very (too) large, with hazardous profitability without proper advertising management.
Results for these two beautiful technologies (Orange and Exalead): RIP, quite simply. But they constituted my opinion with Dir.com the 3 most beautiful search technologies made in France which could have had a role to play in France in particular and on the Web in general. In the 3 cases, it was the shareholders, the buyers or the N + 1 who did not follow. A lack of long-term vision?
The Exalead engine is still online. Source: Exalead
Xaphir and Qwant’s rats
In recent years, we have also seen the appearance of Xaphir, an engine designed Epinal in 2013 by the company Xilopix, and the victim of Cuill syndrome (from the name of this engine, notably launched by Louis Monier, ex-Altavista, ebay , Google, etc. and who missed the boat in 2008 by exiting too quickly under pressure from shareholders). Catastrophic research results at launch have sealed the fate of Xaphir’s technology, which was acquired in 2017 by Qwant. Wanting to get out too quickly of a badly finished and not enough engine, they thus exposed themselves to a few drinks, weary But did they really have the choice?
Research results on Xaphir for the request abound at the time. Source: Ultimate SEO
As for Qwant, I have spoken about it several times on Ultimate SEO in the past, so what to say more? Born in 2013, after an absolutely catastrophic communication from the start, the engine multiplied strategic errors to finally arrive at a horrified annus in 2019, concluded by the forced departure of its CEO Eric L andri at goal 2020. Result : this engine represents today a big stone in the shoe of the government, so much so that the question is not (more) really for our political leaders to find an alternative Google, but rather to find an alternative Qwant And There, things are clearly much more difficult And installing the engine on the posts of French officials will not change anything overall (they will return to Google the next day). In addition, if you have to force people to choose such or such an engine rather than their personal choice to try to obtain any hypothetical success, it is that there is really a problem In short, Qwant is today in a state of cruel death (and its SEO contest Qwanturank has done nothing to improve things) with as a result, 7 years after its launch, traffic shares not exceeding 1% in France. We can regret it, because there are certainly internal of this company very valuable people, who must sometimes wonder what they came to do in this gal re
Today, are we?
Finally, if we go back on the various French search engines that have emerged for almost thirty years, we see that they have all been failures. Why ? Several reasons for this, in my opinion. Bulk:
- Sometimes they were not good and irrelevant in their results. This was clearly the case for some, let us not hide it. But these remain exceptions in the end, because others were very good, or in any case, they did the job rather well compared to the Web as it existed at their time.
- They were killed by shareholders or executives who did not believe it (A syndrome will never work!). And it was a shame because some would obviously have deserved a better future.
- They were designed too early, a time when economic models for search engines were far from obvious (this is one of the reasons for Google’s success – but not the only one -: he arrived after everyone, a time much more conducive to the economic development of such a tool).
- The investments necessary to evolve with the size and complexity of the Web were right of the goodwill, but obvious from the start.
- They have been victims of strategic errors and / or sometimes catastrophic communication.
We could certainly find other reasons, but my opinion, the main ones are there. And then, at the beginning of the 2000s, the meteoric success of Google did not leave much room for many. 20 years later, it’s too late
Indeed, the current world search landscape has become quite simple: at the global level, there are only two search technologies available: those of Google and Bing. Plus some local crumbs for Yandex (Russia), Seznam (Czech public), Baidu (China) or Naver (South Korea). Bar point.
What is missing today about the French search engine?
However, we understand the motivation that many people have to fight American hymnia in this area. Who processes and disseminates the results of research on the Web today infuses information, even opinions. The political risk is normal. It would therefore not be illogical for the French government, or quite simply the French citizens, to call with their eyes a motor made in France. But today and for that, we need to start from a few postulates that seem obvious to me:
- Creating in 2020 (and a fortiori later) a search engine (or at least the crawl + index part) from scratch, in France or elsewhere, is a total utopia. A h r sie. A fantasy. The Web has become too standard, too complex, and the human and technical investments to make it happen are absolutely colossal. Ms. Amazon (with A9) broke her teeth. Microsoft, along with Bing, has struggled to nibble Google traffic for years. Even Facebook and Apple, often tempted, have never risked it.
- If a start-up gets involved in this market, it would leave anyway with a handicap of more than 20 years behind its competitors. It may have been an advantage for Google in 1998 to leave after the others by living their errors at a more favorable time, but in 2020, the situation changed considerably. 20 years behind in terms of data: disability is now too standard.
- And if, by miracle, this start-up succeeds, it will immediately pass to the status of BBBBG (Born to Be Bought By Google) or BBBBB (Born to Be Bought By Bing). Do not dream: who says no a few hundred million (or more) dollars? …
Let us be clear: the creation of a new search technology at our time is impossible, unimaginable, irrealizable, at least for the part which consists in maintaining an up-to-date world index. Point. We can regret it for days and months, however it is so. An engine, in order to develop, must build on existing technology and, on this basis, propose, build something else. And, more clearly, if the goal is to fight against Google, there is only one solution: use its competitor, Bing!
And so, if we want to break the bad spell that systematically sends the trash the search engines from the cortex made in France for decades, it will be necessary to work with the Great Satan Microsoft to fight the great Lucifer Google. Or do nothing. Or see a short-term experiment die. Or maybe earn some short-term money through a buyout. That said, the choice and the axiom of departure remain quite drastic and limited.
Note that this is the approach of tools like Ecosia, Lilo, DuckDuckGo and others, which rely on existing technologies, and which ultimately draw not so badly as they are. further!
The truth will be in the user interface …
Another important question: if we use another technology (Bing in this case), what can we offer more, or rather different, than Google (and Bing, for that matter)? It is there, in my opinion, that lies the nerve of the War. However, I am convinced that there are still a lot of things to do with the user interface. For almost 30 years, a search engine has been a search form (ou00f9 we type keywords) and a SERP, or engine results page, with a certain number of natural links (blue links ) and advertising.
So well, we can not shake the habits of Internet users overnight (it would be a guaranteed failure), but we could do everything to give them the choice: either a display of the results Google (classic: form + 10 blue links + pub), or another way (or other ways), more original (s), even a little crazy (s), more or less reasonable (s) to display the results. Cuill, in particular, had shown the way at the time and, if he had provided relevant results, he would certainly be a major player in search today. Because his SERPs had a dog and offered another fairly innovative way of reading while ultimately remaining fairly academic. Cuill was just right. It was just launched too early, but its basic approach was good (in my opinion) in terms of the user interface Xaphir had also launched some interesting ideas in its time. But he too came out too quickly and didn’t have time to develop them.
Research result on Cuil l ‘poque (2008). Source: Abundance
It is therefore quite possible, in my opinion, to offer a good search engine, with a UX overlay, based on Bing’s search technology (with an agreement guaranteeing the protection of personal data, although not bad), although that is not a really differentiating criterion) but proposing enough innovations in the user interface aspects to be pleasant to use, beyond the essential relevance.
I therefore continue to believe hard, that an engine, or in any case, a research tool designed in France (even if it is based on foreign technology, which the internet user does not really care about – note that the latter already uses a computer and an American operating system, even a telephone of Asian origin – in short, plumbing is not his thing as long as the water is drinkable) to gain shares of walk and even, let’s be crazy, become profitable! And for me, the story is based on what we can offer the Internet user as innovations in the user interface.
The approach will also be essential: what the internet user wants above all is honesty, frankness, loyalty and transparency. If he has the impression that he is being lied to, that things are being hidden from him, he will not adhere to the project. A point never to forget Some may have fluffed their wings in the past!
It seems that in France, we don’t have any oil, but we do have ideas. So, let’s take the pole (index, algorithms) it is (Bing) and bring innovative ideas (interface, UX)! Let us put our minds to work and build a beautiful engine that speaks the language of Moli re! did you say? So, your synapses!