You are right. It is hard to believe. There is a lot of content out there. But good writing is rare. And the more poor content there is, the more your gold content rises to the top.
Yep, it would definitely be hard to know in advance that an article will last 2-3 years. And it takes some expertise to find the right topic and write a good article that humans will find useful. That article needs.
I'm in the top 5, 6 years later
The way I discovered this myself is I saw that we get a lot of (accidental) traffic for "how to fix my windows microphone". I looked and we (Bungalow Software) are in the top 5, for this article I wrote 6+ years ago.
But there was a problem with this article. Can you guess what it is?
The problem is that this article isn't the right kind of traffic for us. We make speech therapy software. The Mic page was just a FAQ for our users. But that traffic is incredibly untargeted. If we're lucky .001% of those folks are interested in speech therapy software. So I did a little experiment. I spent a few hours doing keyword research and wrote an article on a therapy technique. We are #1 for searches on that topic and it's converting at about 1% to free trial signups. (A bit too early to tell how many paid signups that will generate. (It's generating as many free signups/month as would cost us $200 via adwords). (And past conversion calculations show that we at least breakeven on Adwords traffic.
But my article is soo much better than anything out there that I'm confident it will stay highly ranked for a long time.
So it'll be interesting to see how it does.
I recently updated that article and spent wayyyy too much time on the graphics, etc. But I enjoyed it. Total time invested on that page is probably 15 to 20 hours. BUT.... if it has a life of 5 years then that's $200 x 12 x 5=$12K.
You almost need to engineer/design the articles. A bit of up front effort could really increase the Benefit/Cost ratio. The mic article took me maybe 2 or 3 hours to write. Much less than the new article.
I apologize for (deliberately) being vague on the ROI on the new page I created. I don't want to provide a competitor with too much information. (We have a VC funded competitor to the tune of $4M or so).
The Tipping Effect
If you have the #1 spot for a topic, you get a huge "tipping effect". The longer your article is around, the higher google tends to rank it. So, if you can keep that #1 rank for 6 months or so, i'd suspect you can easily keep it a lot longer (especially if you do a small tweak to it ever now and then).
See what our own Patrick McKenzie wrote about evergreen content
1. Content must be very specifically target to your market niche, and complement your marketing.
2. The content must be the best (as rated by humans) on the topic. It should be the canonical article on the topic.
3. Topic should be evergreen. Don't write about the latest headline. Write something timeless. Like, How to fix your microphone or How to measure the quality of your software work enviroment