I think you are leaving out a really important aspect of the early internet- content discovery was really hard for users. If you were looking for something about Pets until google became dominant, you were just as likely, if not more likely, to type pets.com into your browser as you were to go to a search engine.
Getting the initial traffic to your site was really hard in those days, the domain was really important for that.
I could gopher topics by early 90's, and Infoseek, Yahoo, WebCrawler, etc. were a full text search of pages by mid 90's.
I distinctly recall searching usenet across multiple servers.
Even before this, when data only flowed through uucp (or Fido), search was albeit queued, readily available.
Your mileage may have varied; i just want to be clear that it was not that it did not exist, but new-comers would have a steeper learning curve. Today, it is expected, nay, demanded as a human right to be able to search the entire internet from a uniform and single klick search box.
(yes, old crotchety, "in my day..." :) )
So you weren't a normal, mainstream, new user to the internet. You knew how to use it, you knew where to go to search.
I was the same in the 90s (minus Gopher skills), I learned about Yahoo, Altavista, Infoseek, etc. and became a wizard to my friends and some relatives because I could find things on the internet, didn't need to know the website address beforehand or click somewhere in AOL.
It did exist, it just wasn't accessible and generally available, nor it was a tool people knew how to use.
So domain names up to the early 2000s were pretty important to capture the mainstream market, not the Gophers.
Actually that was very normal for a new Internet user of the time. In the early 90's there was no web as far as the public was concerned so Yahoo (1994)/Altavista (1995)/Infoseek (1994) did not exist yet. For users of the pre-1993/1994 Internet, to use it you had to learn a hodge-podge of protocols and software including UUCP, gopher, FTP, telnet, NNTP etc. since that was the Internet back then. It wasn't until online services like AOL started providing web gateways (1995/6?) when 'normal' people really started flooding onto the 'net and it took a few years longer before there were useful web frontends for the majority of services.
Imho the September was really kicked into overdrive with smartphone and mobile Internet adoption facilitating the social media hype.
Prior to that people would at least need to sit down and dedicate attention to whatever they were doing online at their desktop.
It required at least some technical base knowledge.
While nowadays pretty much everybody can, and will, pull out their smartphone even during very short breaks: It' all short attention span, relegated to devices were multi-tasking is way more of an hassle, which means people are also way less likely to do research on whatever they are discussing.
It's gotten to a point where in many places writing anything longer than Tweet length is considered something bad, as most peoples attention span just can't cope with long-form forum posts anymore.
I remember my school ran a contest where you had to find obscure information on the net. Things like when the inventor of the saxophone was born? You had to submit an URL so using the library was not an option. It was insanely hard, nowadays it's just one wikipedia click away.
The fact that you had even heard of gopher would have put you in a very small minority of internet users once dial-up internet access become relatively accessible (so, post-1995ish or so).
As a smart but not necessarily brilliant teenager getting on the web in mid-late 90s- Maybe Christmas 1996 I- well my family really- got a computer that could get on the internet, search engines were mediocre at best. I did have a family friend come over who was an old hand at the internet and knew all the tricks of the day- using operators like AND, OR, NEAR, NOT etc- and he was able to yield much higher quality results, but I was personally never able to replicate his abilities.
Google and Pagerank changed all that, but up until then, it was all very tricky and each search engine had its niche- Yahoo with its directory, Ask Jeeves had a user friendly interface, lycos and altavista had some niches of their own as well that I can't quite recall- or maybe they just each had brief shining moments in the sun. At one point I had desktop software that was a meta crawler that would enter a query into each one. The full text searches on keywords that may or may not really be relevant to what you wanted were really not all that good.
My favorite search engine at the time was Dogpile. It was a meta search engine that probably did much the same as your desktop software. Usually I could get relevant results out of it more quickly than using individual search engines.
Yes, the earliest directories / search engines worked extremely well in large part because there minimal content. When there where 7 knitting websites they don’t need to worry about SEO. It was only after the users started getting hundreds of results for most searches that search engines needed to worry about filtering and prioritization.
People started to understand how PageRank worked, and managed to get George W. Bush's official biography page at whitehouse.gov to the top of the search for "miserable failure." As I recall, this worked until at least 2008, maybe a couple years past that.
There were manual indexes for sure, but my very first experience with the web was someone excitedly showing me how they could dial in, open this web browser thing, and then... nothing, because they didn't know any websites.
We typed in a few things, but it didn't go anywhere.
I remember he called the browser a "web crawler", which was probably the search engine someone told him to use to find sites. We had no idea.
Otherwise, I have a lot of memories exploring the web by typing in words followed by TLDs and writing down the interesting ones. It was another year or two before aggregators cropped up, copying and returning the first ten results of any manual index.
I don't remember how I discovered my first web site, but it might have been from a computer magazine at the time. Later, I would use Yahoo! as my search engine, eventually switching to Google.
I find it funny that you mention half a dozen ways to search the web and everyone seems to be focusing on just gopher because that is the only thing they can explain away. Google must have dug up Steve Jobs reality distortion field.
another issue I don't see talked about much in these comments is trust. Amazon had to develop a name that people could trust. I remember when people would agonize on whether to make a purchase because they were afraid they would be scammed.
I can totally second this as someone who began surfing web in 1997 as a university student.
Content discovery was next to non-existent. You had these directories like Y! or Lycos which more or less mirrored yellow pages. My group of friends would exchange interesting site links over e-mail, floppy disk or the good old way, write them down in a note book. Before you ask, internet surfing was only available in expensive public kiosks so browser book marks weren't of much use. Only when I got my own PC + internet in 2000 did I began using browser bookmarks.
And then Google became popular around 2002 just when I entered post-graduation and changed web surfing forever.
People were commonly using Lycos and Yahoo as search engines by 1997. Sure, some of the more exciting content wasn’t stuff you’d think to search for and thus would often get shared via word of mouth (like the “Bert is evil” site that parodied the Sesame Street character in compromising photos...after all, which sane persons would search “Bert +hitler” ?)
I think the limited time many people had on the internet (as you said, in some instances at Internet kiosks where you’d be limited to an hour at a time. Or on expensive dial up) probably contributed to people curating and sharing links offline between friends. But I do remember using search engines in that era specifically because AltaVista was widely regarded between myself and my peers to be the best search engine in terms of keyword syntax and the accuracy of the results. Remember this was the era before search engines popularised entering in natural language as a search query. Ask Jeeves (later renamed to just “Ask”) was actually the first to popularise that and even that was pre-2000 (it also largely sucked compared to keyword driven queries but I guess the tech wasn’t quiet there yet).
I think it was 1999 when my friends and I first discovered Google. Back then Google was keyword query based too. The natural language side of Google came much later (mid-2000s at a guess). But what sold Google was its minimalist home page and the accuracy of its results. The minimalist home page was novel because search engines were considered internet portals before Google. Yahoo! Would have online games, chat rooms, site recommendations and email all accessible from its landing page and it was a similar story for many of the other search engines as well. They considered themselves the homepage of the internet (and in many ways they were right). Whereas Google went the opposite way and said “let’s strip as much guff from our landing page as possible” and is modem uses really welcomed that (plus the accuracy of its results too). It’s ironic just how heavy Google’s landing page has become.
By 2000 the web had already felt like it had shifted from its adolescent years of anything goes and was starting to grow up. The stigma of meeting strangers online was fading and businesses were adopting the web as more than just a niche marketing tool - in part helped by Online payments becoming a thing with PayPal, WorldPay etc.
> Yahoo! Would have online games, chat rooms, site recommendations and email all accessible from its landing page and it was a similar story for many of the other search engines as well. They considered themselves the homepage of the internet (and in many ways they were right).
Still a winning business model, because this line could easily describe Facebook.
That’s more a symptom of the lifestyle of any popular software project than it is an example of it being a winning business model.
Time and time again you see these big monolithic applications get displaced by newer more focused applications and people love them because these newer applications run faster / is easier to use / etc etc. Then as those applications gain popularity new features need to be built to continue growth. Whether it is through feature requests, to fight off competitors with other features, or just the businesses way of finding new ways to look individuals into a walled garden....soon this focused application becomes yet another behemoth that people start to moan about. Eventually something new, shiny and focused grabs the public’s attention and we flock to that like the fickle herd of pack animals that we are.
If memory serves, we started searching on AltaVista, moved to Northern Light and then to Google. Certainly directories played a role but really, AltaVista / NL was the bees knees.
This brought back a long lost memory of when before we had internet at home. My dad took me to an internet cafe and I had printed a whole list beforehand of sites I wanted to visit. I had just guessed a bunch of random domain names based on my interests and things I thought would have a cool website. Stuff like porshe.com, titanic.com, mountainbiking.com, spaceshuttle.com.
Speaking of yellow pages, I had an Internet-specific yellow pages, I think it came from an order from Outpost.com. It was a physical book, just like the original phone book.
I think I still have it, and if I can find it I'll reply with details.
It was fantastic for finding things, because you could explore it, and because search was garbage back then.
I remember that era pretty fondly and I don't ever recall just typing in random URLs, except maybe jokey ones like sex.com or fuck.com for kicks. We had Yahoo's search engine at the time and worked well enough. And to a lesser extent Lycos/Hotbot/Inktomi/DMOZ. The pre-Google web was pretty interesting and a bit more sophisticated than gets credit today. I also feel you were less likely to discover retailers on the web randomly. You'd more than likely hear about them first on the news or from friends and go from there. I don't think it was common to just discover some random e-tailer like we do today, put in our credit card, and expect the package in a couple days. You learned about Amazon from 60 minutes or your friend who subscribed to Wired magazine or the guy at the software store, not from a search engine.
I think the whole pets.com and travel.com is just the usual marketing logic at work and not really related to the limitation of search engines. Its just shady marketing tricks, a bit like how we see the .biz and other gimmicky tlds today.
It was really common to type in random words as domains expecting something. People would make mistakes so people would buy domains with common spelling mistakes.
I have a random memory of the website at hanson.com having a banner alerting visitors that the website of the popular late-90s boy band was in fact at hansonline.com, and that you would only find information about guitars made by the Hanson company should you further browse the site at which you currently found yourself.
Curiously, that domain now redirects to some concrete company, Hanson guitars existing instead at hanson-guitars.com.
Browsers also tried to autocomplete <word> as www.<word>.com (and .net and .org, iirc) and send you there if it resolved. At some point they switched to sending you to search engines instead, but i bet there are flags somewhere in Firefox' bowels that will still do that for you if you really want.
The first time I had the chance to use the internet I typed in "whitehouse", hit enter and whitehouse.com loaded. It turned out to be a porn site, the teacher saw and I got banned from using the internet - in 40 seconds.
White House was the first time I ever seen porn / naked women. Cue me spending the next 6 months sneaking to load the page on dialup all hours of the day. To be a kind again. Now we have TBs of data but tend to be uncontent sometimes.
Heh, was about to post this one myself. My buddy got in big trouble in the high school computer lab making this typo. I wonder what percentage of domains in the late 90's were simply "adult" sites trying to capture traffic via typos. It certainly felt like a lot.
In late high school or maybe during college, my wife was going to look for something at Dick's Sporting Goods by navigating to a domain that any ordinary person might expect. The result was not what she expected.
LOL, I used to fuck with people by getting them to type this into the browser. Unfortunately, these days it actually takes you to Dick's Sporting Goods.
To that end, it's worth reading up on the history of sex.com. The domain historically (and currently?) generates tonnes of revenue through advertising, and has been hijacked more than once.
I wonder if it's because people are likely to try `sex.com` in a web browser for laughs (and get surprised when it turns out to be a real site), but aren't likely to do so with `books.com` or `travel.com`.
This is somewhat second hand but I worked with someone who claimed to know the person who owned sex.com (he was trying to buy it from him at the time, and I'm trying to be deliberately un-assertive) in the mid 90s (also fair warning, this was a long time ago and I probably misremember stuff) and said he paid a ton of money in bribes to the search engines/portals back then - which was what most people assumed, but those companies insisted wasn't true, so it fit a narrative people liked.
I don't think sex.com was an expected direct hit back then - and if I'm remembering correctly, even in the later 90s, you would have had to type http://www.sex.com to get it to even load right - there wasn't a ton of convenience and most browsers relied on a heavy set of built-in bookmarks to get people into portals and search engines.
That said, I believe it was totally true that they likely paid six or seven figures for a year or longer deal to guarantee top three for a ton of porn search terms. I assume that area was allowed to be seedy in order to mitigate risk of even allowing that content to be included in the first place, and companies didn't feel bad getting money from porn sites who had a lot to throw around.
Should add that as I understood it the business model was that sex.com resold front page space to affiliates that were relevant to the search terms that led to their paid listing being listed higher for a particular keyword. That’s how they could afford to bribe/partner with portals/search engines pre-Google.
For context, squatters would often land you on some bad sites and URI syntax was unforgiving, so it just isn’t something I remember most doing. There was a general desire for simple branding on the off chance you can get it typed in or printed on a business card or in an ad, but it was not the common entry point.
Also of note: UX expectations back then didn’t trust typing into a box and seeing any result first - it really was the solid ranking of Google results that made that UX something browser vendors adopted or more cynically, that Google pushed first with partnerships then with ownership of Chrome.
And the implicit threat of your bank “outing” you via the transaction if you try to cancel. I may be misremembering but I felt like there was a good chance what I was viewing legally may end up suddenly illegal with the way online legislation was going in the late 90s, and I’m talking just like laughably soft core stuff by comparison to today, and that permeated the online porn business and kept it seedy for a long time.
No idea what you’re talking about. There was the run up to the communications decency act and obscenity prosecutions but these were far more risky for the peddlers, not the consumers (and some were prosecuted). There was the old hilarious CP80 initiative by the Mormons (with a nice tie-in to the SCO v Linux debacle) but that never had any serious legs to stand on. Meanwhile discreet billing for naughtiness predates the internet and was especially notable in the days of 900 numbers - assuring discretion was pretty well stated. Considering that child porn was essentially legal and you could walk into adult stores in broad daylight on Times Square not too long ago I think you are misremembering things or probably too young for context.
What changed is the cultural mainstream acceptance.. that something like PornHub can be a mainstream company. The market itself was alive and well pretty much the day after the movie camera left Menlo Park.
Many sites didn’t do this right even large ones and often browsers if my memory serves right automatically put www and .com on any single word typed by default. That said yes, that should have worked, I just remember it not working enough that it wasn’t a normal or expected UX pattern until Google and maybe even Chrome.
I remember being on Tymnet and just trying addresses via a program to see if I found anything. Most interesting thing I found was a Fed Reserve address once (did not do anything with that).
there used to be (sadly not anymore) a very active forum on fuckyou.com and it was mostly populated by people that had just randomly typed it in one day.
I used AltaVista and Hotbot all the time, as well as other search engines and curated pages, and there was no problem finding things. I believe most people used search engines and aggregation pages since entering the address in the browser bar would simply yield an error if you entered it wrong. Connecting the address bar to a search engine is fairly recent and was disputed a lot.
Google did nothing particularly innovative or new, they just had the cleanest interface, their page was fast, and provided good results. That's what made them successful.
Google had the best results by far. I remember around 2000 most “regular people” had pretty much given up on the web. It was google and maybe Napster that got people interested again...
I wonder what it’ll be to fix the Internet this time around?
Altvista's edge over the competition at the time was that it indexed the most content. But google came in with pagerank and had superior results. There was a 3 year span or so where people just attributed google's success to minimalism. in terms of minimalism, hotbot was a disaster, but it was the algorithm that truly set google apart.
minimalism has its good points, but falsely attributing it to google's success probably focused too much attention to it in the internet of the 2010s.
Exactly, the single feature that made me use google over yahoo/altavista was the clean interface, I was using dial up back then and it loaded very quickly the others were covered in ads.
Early search engines were pretty dumb and heavily weighted on keywords, which a generic domain name helped with as well.
Again, less of an issue now with more intelligent search engines. (Although strong brands still tend to a get a lot of traffic via searches on their name...)
Oh yeah, I remember discovering some great content from pages like that back in the day. I also vaguely remember the idea of content rings being a thing.
I thought it was much better back then, you could find niche sites that were of interest.
Google seems to have banned all the interesting wild content that made the internet so fascinating in its early days. Blogs seem to have disappeared completely for instance and I have to specify the site to search to find anything e.g. site:Reddit.com
However it does have an amazing ability to find comments on stackoverflow relevant to my needs based on a few keywords.
In some ways content was more discoverable back then. There was less of it, and it was objectively harder to get it online thus making the quality higher for the stuff that was there.
Plus Web Directories, and Link lists where IMO better than keyword searches. They allowed you to find and follow topics faster.
The explosion of content and google has made discovery harder unless you are one of the Top sites or platforms. All it takes it being on one of them, not being a good quality content.
Want to sell a widget, you better sell it through amazon because chances are Google is going to index those results as the Top Results not your webstore. Even manufacturers that want to sell direct have to also have Amazon stores because the Amazon listing is the one of the only ways to get a top search spot
The early internet, for me, was a 1-800 number BBS called Starfire. It had a splash page to the “internet” section that I’d assume was curated by the operator. This was probably 94 and 95.
This, also search engines used to give a lot of weight to workds in domains so if you searched for "book", the search engine was likely to give you books.com as first result.
On the other hand if you were in yahoo you were online. Type "books" into yahoo though and you'd be more likely to find books.com than amazon.com as the top page, and clearly "books.com" would sell books, who knows what "amazon.com" sells, something about Brazilian rivers?
Absolutely. If you know what you want to buy, it takes seconds to find the absolute cheapest place on the internet to buy it. Then it's up to you to decide whether you trust the site or not. Brand name helps with perception, but definitely not discovery.
That used to be true with Amazon. Not anymore, not for me. Now I have to decide if I trust and endless array of 3rd party sellers. It is often just easier to buy from a manufacturer’s website since I know I’m getting the real thing and it’s not expired or otherwise screwed up. Maybe I pay a few more bucks, but I’m ok with that for a lot of things.
(Exception: books and used)
It's always interesting to me how people on HN almost universally have such poor experiences with Amazon. I've had over 200(!!!) orders from Amazon last year alone and didn't have any problems with any order. In fact when my guaranteed next day delivery was late by a day they just gave me a month of prime for free.
And every time I decide to buy directly from the manufacturer I get punished - bought a Lenovo laptop around September, laptop turned up with a broken screen. Took over 3 weeks to get Lenovo UK to replace it and it was an exercise in absolute frustration, I could write a small essay about it. I was so upset at myself for not buying it from Amazon - I know if I did and had the same problem I would literally have a replacement posted to me the following day without having to fill out a dozen forms and spend hours on the phone with customer support.
I also prefer to order direct, if possible. If I think I might need to return it, Amazon.com is the way to go (or Bol.com for local stuff). It's always a pain to return to the manufacturer, they're incentivized not to take back the return while a retailer is incentivized to take it back.
Getting the initial traffic to your site was really hard in those days, the domain was really important for that.