The author's primary complaint seems to be that Medium does not update the pubdate when articles are updated, consequently Google and other search engines give the article diminishing relevance over time. This strikes me as more of a problem with search engines than with Medium.
In my experience, modern search engines give far too much relevance to date, especially for evergreen content. I am purveyor of evergreen articles, and it is sincerely quite obnoxious to put a lot of research and effort into an original article, only to have some flip shop do a lazy rewrite a few months later, citing my article as its primary source, yet ranking higher in search because it is newer. It's a perverse incentive that encourages low-effort regurgitators, and discourages original work. Grumble, grumble.
I don't fully disagree with you, but I think it depends on the domain.
You're right that it's annoying the way Google will show me results of a blog post recently published by someone who just "discovered" something that's been known for decades, and what Google shows me is a poor summary of more in-depth original work done elsewhere.
But at the same time, I hate the way when I search for solutions to an issue involving computers, I'll Google shows me nothing but results that are 10 years old, and no longer work because the operating system has changed.
Respect for high quality old content, mixed with a more prominent UI for the time filter (it used to be soon Google, but less so now) is the right trade-off imho.
I've noticed there are blogs that will do fake updates so the date in the Google search results will be from last week or something recent, even though the content itself is the same it has been for a couple years.
This is definitely a problem for people who write true evergreen content, or do frequent updates without updating the pubdate, but across the internet I would imagine there is much more old content that is out of date than old content that is still just as relevant, so it seems like a smart strategy.
I was thinking about this too. Search engines should regularly crawl a page and notice that the content has been updated right? So it seems like they should ignore pubdate as it could be gamed, and just measure the real "last updated" date based on content changes.
> So it seems like they should ignore pubdate as it could be gamed, and just measure the real "last updated" date based on content changes.
It's hard to imagine a mechanistic measure of the 'real' last-updated date that couldn't also be gamed. (If I'm being really cynical, it's hard to imagine anything that couldn't be gamed; so I guess I mean 'easily gamed', for some value of 'easily'.)
That’s definitely true. I often find myself clicking first on a more recent search result and then backing out and clicking on older content if the initial choice didn’t live up to expectations. Since I probably still had decent time on page, etc for the first link, I’m guessing I’m making Google think that content is more relevant than it is.
Medium has become the go-to publishing platform for more and more entrepreneurs and businesses. However, I'd argue that most people would be much better off publishing content on their own domain. This is especially true for "evergreen" content, since Medium does not allow you to update the date of your posts—meaning they will perform worse in search rankings after a while.
I started my own business, Nomad Gate, on Medium about 3.5 years ago, but have realized that I'm much better off hosting everything myself. After making the switch, both my traffic and revenue more than doubled.
I wrote an article/PSA describing my experience with Medium (and moving off it), how they are making it harder and harder to actually leave their platform (no more custom domains, no more updating canonical links, etc), and how to safely do content syndication to Medium after moving your content hub elsewhere.
I hope it will help some of you make better decisions about your content strategy, and I'm happy to answer any questions you might have for me!
For anyone wanting to build their own blog and start publishing content, what do you think is the best way to gain traffic? Just share it everywhere and hope it grows or perhaps use a dual Medium+self-hosted thing? (for a time being)
I'm in the process of doing that, and oh man the amount of minute details that I have had to solve just to get things running. Granted, I did it the hard way building everything from scratch with Gatsby but I guess it was sort of an experiment too, so the time wasn't necessarily wasted. Recently I started fiddling with the OG tags, JSON-LD etc, and that was a quite tedious task. Hope it's all good now.
It obviously depend on your niche, but as long as it's not super competitive I've had pretty good success appearing in the top 10 with articles posted on my own (fairly new) domain.
Perhaps the best advice I can give is to don't focus too much on SE algorithms (just make sure you don't mess up any tagging etc), but rather focus on creating very good, well researched content. If you do that, people will share, and Google will notice. Personally, I have had success with writing long form articles. Many of the articles on Nomad Gate is 4-5000 words or more. At that point you will get a lot of long tail traffic as well.
This morning I listened (again) to a really good IndieHackers podcast with Laura Roeder where she answers this question (thanks Courtland!). Her suggestion was to buy traffic from Facebook, IG, etc. at the beginning. Once you're seeded with some regular visitors, continue to work on getting more organic traffic coming in.
But like everything, 90% of success is showing up, so first you actually have to tell people the blog exists and post stuff!
I wouldn't take advice from me since I post very infrequently and no one really reads the posts that I do write. However, I'll throw out one other option which is posting on both Medium and your own domain. Medium allows you to set the canonical URL so you shouldn't take an SEO hit.
Not if you update the content. When reading up on interfaces that change rapidly, I'll often throw $CURRENT_YEAR into the search bar. If your post has been updated with current content, I'd like it to be included in my results.
It depends on the content and the type of updates you make.
As long as I do a significant update to an article, and make sure it's generally up to date and still entirely relevant, then I'll update the date to match.
When people look at the dates (in SERP) they generally want to know if the article is still relevant or if they are likely to find outdated information there. At least that's how I think.
Automatically updating dates without actually revising the content on the other hand... That's not something I'd want to endorse.
As far as I'm concerned, posting any content of value on any platform or service that you don't pay for and don't have a good control over is a bad idea. It doesn't matter if it's Medium or Facebook or Instagram or Quora or anything else. If it's, say, on one of the Facebook properties that puts up barriers for people not registered with them (and thus the content is not easily accessible on the open web), it's a terrible idea.
Those who value their content should (when financially feasible) ideally use paid services or paid hosting where the provider's interests and the publisher's interest are aligned.
> Those who value their content should (when financially feasible) ideally use paid services or paid hosting where the provider's interests and the publisher's interest are aligned.
Or, when financially constrained but technologically inclined, use a custom domain on a service like Tumblr or Github, to allow migrating later down the line.
The author touches on this:
"If I had used a custom domain for my Medium publication, it wouldn't have been much of an issue. In that case, I could have just set up 301 redirects to the new articles and deleted the old ones from Medium. All the SEO juice would have been preserved, and Google would have shown the updated article in the search results."
This is basically B2B media, which has been around forever. We built a very profitable company about writing content that people in niche industries need to do their jobs, and then selling ads around it.
There are many factors that go into what industries we cover and who within that industry we write for, but the biggest one is probably that they all make buying decisions for stuff that costs a lot of money: https://www.industrydive.com/industries/
If your site is http://efficiencyiseverything.com I think it could be a hit with a graphic redesign and maybe converting some posts to YouTube, Instagram, Facebook, etc. It’s a neat idea but slow and a little hard on the eyes.
This exactly. Companies exist to make money, and that's of course a good thing. But the way that some tech companies choose to go about that (growth at all costs, maximizing attention by clickbait, and ultimately selling whitespace for advertisments) may not be the best long-term way to do things.
It's like the real-world option between choosing to publish your own little newspaper versus writing a small column in someone's else newspaper, where you have no control over.
> As far as I'm concerned, posting any content of value on any platform or service that you don't pay for and don't have a good control over is a bad idea. It doesn't matter if it's Medium or Facebook or Instagram or Quora or anything else.
Playing devils advocate for a moment: what about HN?
I get this place generally behaves differently but the points you discuss doesn't explicitly exclude HN.
There are many places that I didn’t mention. And I did forget HN, to be honest. But you’re right — what I said applies to any long form and/or thoughtful/useful content, even on HN.
I've seen similar discussions about Stack Overflow as well. At one time people really welcomed comments being posted on there but some people have since become a little jaded due to how strictly moderated that site is. While it keeps the place in order, it does also mean anyone who might disagree with a moderators decision - even when they have genuinely thoughtful content to post - is left out from the community.
One day people might say something equivalent about HN. So your points are valid about HN if just at a theoretical level.
It’s not just valid theoretically. There is a cognitive bias on HN in certain hot button topics including psychiatry, addiction, and Facebook. Dissenting opinions there run the risk of getting you trolled and flagged under dubious pretenses.
Overall the moderation system here works better than any other I’ve seen on the internet, but like any organization of humans there are some dysfunctional areas where power prevails over logic.
Spelling corrections, factual accuracy corrections that kinda thing needs the ability to be edited.
Using HN to publish "lasting value" content isn't the same as publishing a book where likely you'd work with an editor to reach the final error free imprint. Even an e-book can be updated to allow corrections, and future print editions can be published with corrections or an errata.
After a certain time limit you can't even add a new reply to your own "lasting value" post that could contain these types of changes.
There are federated platforms that let you do this - or at least a lot closer to that kind of reality than a forum / HN style site does.
You do make an interesting distinction though. My counterargument would be "Ask HN" style topics where the entire content is hosted on HN - however I don't disagree with the point you're making either.
Stallman reccomends that if someone needs social media for publicity or business need, then you set up your own domains and only post "title, who, date, link" and get the useds off predatory platforms like Facebook or Instagram or Quora.
If you do that, you retain the meager control of your profile, and you don't ignore social media. 'Better' of both worlds.
Except that a least Facebook tends to show this kind of posts to no one. Instagram is worse in that your posts can't have external links at all. I guess all other social platforms will end somewhere on that continuum in their quest to keep you and your content on their site.
There is another twist. Facebook's terms of service say:
> [..] you grant us a non-exclusive, transferable, sub-licensable, and worldwide license to host, use, distribute, modify, run, copy, publicly perform or display, translate, and create derivative works of your content [..] when you share, post, or upload content on or in connection with our Products [sic] [..]
Emphasis mine.
So if you post a link, do you grant them a license to the link itself or to the content the link points to, because it is content posted "in connection" with their product?
Out of interest: Do you have a link to that Stallman quote. I'm interested to read more about it and about the context in which he said it.
Thank you for responding on my behalf, ksangeelee. That is indeed the source from Stallman that I was thinking of.
Ideally, we need to stay off these platforms. But I believe these are a good middle-ground to retain a presence when we have a need to be there. But we need to corral these useds so they end up getting away from the para-sites :)
Wikipedia. I wrote a series of items which were plain hard work, mostly owing to attitudes of other wikipedia editors. I saved some back, posted on my own site and made a good source of ad revenue over the past 10 years or so.
> Those who value their content should (when financially feasible) ideally use paid services or paid hosting where the provider's interests and the publisher's interest are aligned.
Which works until those paid services are bought out by someone with more money and radically change direction, disappear, or both.
Hosting your own content is the only solution. Anything else puts you at the mercy of someone else.
The problem is I get more viewers on Instagram than with hosted stuff. So ideally I would love to not use Instagram, but I guess I feel my hand is forced :/
What does paying have to do with it?
A paid service can have a bad control, and a free service can have good control. It's the quality of the system that matters.
> But to say that it's not possible for them is clearly a lie.
> Medium, if you ever read this: I don't appreciate being lied to.
> To top it off, they simply ignored the two (very polite) follow-up emails I sent them.
My guess is that Medium has an admin tool for customer-service agents to use, and that tool was likely updated and no longer allows for canonical link updating. As companies mature, the freedom of CS staff tends to diminish, as mistakes are made or privileges are abused.
The email reads "there is no way for you (or us) to set it to something else". I'm guessing "us" refers to the customer-service team, and not Medium as an organization.
In other words, I don't believe the author was lied to.
As for not getting replies, I'm shocked to learn Medium even has an open support email! I'd have to see the author's followups to know if they warranted a reply, but given that the author adamantly believes that deception is at play, it's possible that the emails aren't as polite as the author believes.
Well then this is quite simple for us to clear up then - Medium has mis-trained their customer-service team to think that they're not speaking for the company when they're speaking for the company. That can be fixed.
Yes, this is what I thought probably happened, as well. The "us" referred to the support agencies, and they likely were no longer able to change canonical URLs (surprising they were allowed to do so to begin with).
Sadly, it's not archived anywhere I could find, but it did mention that you could send them a CSV file of up to 50 URLs to update the canonical links for.
That doesn't contradict anything the agent said. I believe you that they could previously set canonical links. But it's highly likely that upper management changed that policy (I'm surprised they allowed it to begin with).
IMO, it's not too surprising that they tried to be a bit more writer friendly initially (to attract more quality writers and larger publications). I guess it no longer fit with their new business objectives, which is why they quietly removed it this fall.
I think most people on HN have set up their own domain and web services before or have the ability to do so without having to do much learning. So I don't think it's a stretch to say that on the scale of difficult things to do, hosting your own blog is very low.
But Medium is just so easy. It's trivially easy. It's uncluttered, clean, easy to just write an article and publish it or share for editing. The fonts are easy to read anywhere, it's fast, easy to find through Google because of @handles and works seamlessly on all devices. Oh and it's "free."
I've tried to figure out how to make self hosting this easy but honestly it escapes me. Maybe someone smarter than me can figure that out.
It's not uncluttered. Do you really want someone arriving at your article to be greeted with a big modal banner offering them to "make things official", as well as the un-hidable top and bottom banners?
...complaints about Medium's obnoxious banner and general drop in UX quality over the years are among the constants in the comments when they're posted here...
I don't understand why anybody would ever post anything to Medium, as there is a monthly article limit. If you want people to read your work, why would you post to a site where how much the audience reads _other people_ affects their access to yours? It makes no sense.
Medium became popular years ago because it allows authors to have an inbuilt audience (via their recommended feed) and a writing interface that's aesthetically pleasing.
In the years since, both advantages have weakened a bit as competitors now have similar features.
The monthly article limit only applies to authors who opt into the Member program to get revenue, which Medium has been pushing since previous monetizations attempts failed.
On a side note. Google should allow users to ignore page's age as a ranking criteria. I am not comfortable with the fact that some articles are lower in search results simply because they were written long ago.
Generally anything that is not time sensitive, either because it is about a generic or static topic, or because the author makes regular updates if it is not static. EG: a post on WWII airplanes would be evergreen, as the data is pretty much not going to change at this point. Similarly a regularly updated post on stealth airplane technology could also be evergreen because it evolves with the technology and is presumably always current and relevant.
A post on raising VC funding that was written in 2007 and not updated is not really evergreeen because while it was accurate for the time, it is probably not accurate and relevant in the current funding market.
I could quip that "Here's what Trump tweeted today" is the evergreen content, because you always know what Trump tweets are going to be like. But that wouldn't be a HN-worthy comment, so I won't.
It seems like they moved the article to tkrunning.com, which is a Medium custom domain created before Medium custom domains were disallowed. From there, tkrunning.com is behind CloudFlare, so I would guess they used CloudFlare's Page Rules, which lets you (among other things) setup 301 redirects at the CloudFlare reverse-proxy level.
I don't believe I'm violating their ToS, but you never know how they would interpret it. That's why I'm not sharing it publicly. But feel free to email me on mediumtrick at nomad dot email.
I don't think Medium is able to plug the hole that easily, but they might tell me to stop doing it or plug the hole just for me. I don't believe I'm violating their ToS, but you never know how they would interpret it.
No, it's not that, and I feel silly for thinking it could be that, because it would require iframe busting on a sandboxed iframe.
I think it has to do with being able to use Medium on a custom domain. I think perhaps Medium validated that the DNS pointed to Medium once and stopped checking.
Let's not be naive. There's __always__ a trade-off, __always__ some compromise. Why does the author sound so surprised?
We might not agree with all of Medium's decisions but we certainly have to understand they have the right to make them. How many platforms have to jerk the wheel left and then possibly jerk it right before we realize we're hitchhikers. We don't own the vehicle. They do. We knew that when we got onboard.
I'm not defending Medium. Not at all. That said, 2009 called and wants the cries of "OMG! Look at what Platform X did ___ to me and my content..." back.
I'm not surprised and as I say in the article "A heads-up would have been nice, but still, that's their prerogative."
I do understand why they would want to move in the direction they are, but I also think authors in particular should be aware of the trends and make choices that will benefit them, rather than Medium, in the long run.
What's a bit sneaky here is that they have been selling authors on the fact that they own the content, and the fact that you can move it elsewhere later (including updating canonical links). It's a bit like a SaaS offering an export function, but after you put a lot of your data there they quietly remove it. For bloggers, you can't claim to give them proper data portability without a way of updating the canonical links.
I don't think I ever see any Medium content linked to anywhere else outside of HN. Am I just missing some other sources? Is it a case of Medium writers link to other Medium writers?
Unfortunately, I see a ton of mostly low-quality Medium.com links on my Pocket app's "Discover" feed. You can't block domains on that app, you can only flag links as spam, not of interest, etc. Flagging them doesn't seem to reduce the frequency of such posts appearing on the feed however.
It's important to acknowledge that centralized platforms like Medium have benefits, and a truly decentralized writing platform will only work if those problems are worked out to a level where writers have a good experience.
- The biggest thing they give a lot of writers is the promise of automated distribution through their topic and user graphs.
- They can ensure higher comment quality and real identity through review systems that cost money but have economies of scale.
- It seems like domain-wide features is still something Google uses so if they keep the average content quality high, there's an SEO advantage to using these platforms.
- They can invest in a follow graph protocol that's non-anonymous (unlike RSS) so you know exactly who your audience is as a writer.
- They can otherwise make more complex product investments that will take the open-source community a little longer to coordinate on.
Many of the points I mention above are possible in decentralized platforms, they just haven't yet come together in a cost-effective (zero monetary cost and high ease-of-use) way for writers.
Granted, publishing on Medium removes a lot of the author's own "branding"; but doesn't Medium also provide a certain level of visibility that would be quite hard to obtain if one were to publish individually?
We're having this debate internally. I'm trying to motivate teams currently blogging in Medium to come back to the new CMS we're building for the whole company. Briefly, side A says, Medium is a well-known platform with good discoverability and a great writing experience, and having a popular high-ranked site outside of our core helps SEO. The other argument is mostly, we want our web properties to be more than a brochure, we want the voice of our company to speak "from it" rather than some other random place (even if one presently popular), we should avoid moving between fashionable 3rd party platforms, etc.
I am currently having this debate as well. One angle to consider: Medium presents a path for your writers to grow their own personal brand - instead of writing on medium as a company account they can write as their personal account.
The popup mentioned in the article is a huge detractor from Medium, so I'm still waffling.
Can Medium function as a funnel to your CMS? I assume there's nothing preventing a company or individual publishing on more than one place. Post to Medium with links back to your CMS. Blurbs, teasers, full articles - whatever suits your content and goals.
you are correct, that's another facet of the debate (I tried to hit the high points). The argument against that is usually "as long as I'm there why don't I just write the damn article".
I mean, you don't write the article from scratch on each platform. The author should, ideally, be writing it locally and then placing it in all the places it's needed.
I can get 3000+ views pretty easily on a blog I haven't put much work in. (More like 30,000 or 50,000 if the post connects with people)
Somebody pushed me to put a post on Medium where it got 20 views.
I hear from people who are excited that they post to Medium and they get 60 views -- they think that is a lot but I think they are ignorant about what is possible.
So far as I can tell, Medium has tried to create a brand for blog posts that are a little bit better an average but really the posts there are a little bit worse than average. It seems like a ghetto for clickbait headlines that would appeal to HN readers but when you read the articles they are at best thinly sliced salami which doesn't satisfy.
Typically, less than 5% of an article's traffic has been coming from Medium for me (that's true for unpopular articles as well as more popular ones with more than a quarter million views).
So don't give the "visibility argument" too much weight.
> The main issue evergreen content is facing on Medium is that you can't update the date of an article.
Nor should you. The published date is immutable. You may, however, set a new _updated date_ with your last update. We need to separate these ideas, because there has been rampant abuse of rank preference for fresh dates. Tech (and I assume non-tech) gossip sites are terrible for this.
They seem to update all articles to be mere hours or days ago in perpetuity, despite no meaningful changes. This has to stop. It makes the dates meaningless.
This is definitely true. What I'm referring to in the article is cases when you've done a proper update of the article to make sure it's up to date, added new information, etc.
I actually hope Google will start comparing versions of articles before/after date updates, and penalize those that update the date without updating the content.
And while I appreciate good updated content, I still insist that the _publish_ date should not change. In your case, the publish date ought to be many years ago. The _updated_ date, however, may be re-written with those refreshes. Content should include both (when applicable) and indexes should reflect this.
If you think Medium is bad, take a look at the horror that is blogspot. Loads a ton of slow JavaScript, breaks scrolling, breaks the back button, flashes a different article before showing the one you wanted…
I really do not understand how people put up with this, especially companies that should care about their brand.
Luckily, it's no longer 2004, so no one in their right mind would start out blogging there ;)
I personally prefer Jekyll or Gatsby on GH Pages or, even better, Netlify. But if you're less technically inclined, WordPress.com is fine as long as you get a custom domain.
I only set up this blog for this specific article, so haven't given following much thought, but yes it does have RSS (https://bts.nomadgate.com/rss.xml). And people do still use that.
My main blog (https://nomadgate.com) offers more options: RSS, web push, email newsletter, community forum, etc.
The thing that's nice about medium is the social network around it, the claps, the comments, the hooks into Twitter, the reccomendations for new content to read etc
I post to my own blog, but really miss those hooks into the community. Maybe ActivityPub will be a way to solve this
Although content is king, there is a lot to be said for pushing the envelope when it comes to design. With CSS Grid and other delights of HTML5 I think that anyone wanting to write true 'evergreen' content needs to consider document structure more, so article, aside, figure, nav and all of these other tags get used.
I am currently working on a project with high resolution images and a 'deep zoom' OpenSeaDragon viewer for these images. Why would I want to limit myself to a platform or a CMS that does not allow this advanced presentation?
I just tried a Chrome Lighthouse audit on a typical Medium post:
And the time to interactive given was 19 seconds for mobile. That is a bit silly.
Evergreen content should not be dependent on another platform, it should rank well if the document structure is really good, plus if the content is actually there and people read it. I see no evidence of well crafted HTML5 documents trouncing clickbait blog content on Google but I don't think it has been tried, everything is a 'sea of divs and class tags' and not putting document structure first. However, if I was wanting to write something that still was relevant in 2028 then I would have to consider how the web would be crafted then and it won't be a 'sea of divs'.
it depends also on whether comments matter, we all know from HN that the comments are here and not there when it comes to useful articles, even if they are on a prestigious site. For an evergreen article do you necessarily need discussion and comments? If people are just learning how to take something apart and fix it then they are just needing to takeaway that knowledge, not go writing about it. If writing about some aspect of Brexit then that would be something you might want comments about.
Really you need others to be writing about your content, with link to the original article, not having them rip it off verbatim. But people that copy aren't necessarily going to rank as well if they are not using screen reader HTML5 and, if you do have something like a deep zoom tool to tell the story then they cannot rip off the content for that very easily. Most writers though are quite okay with stock Wordpress or other CMS and some lightly modified them from somewhere. Actually that is too much effort and hence the success of Medium et al.
I just threw the site together in a few minutes before publishing the post (experimenting with Gatbsy), so literally just added a max-width and changed the font styles.
In my experience, modern search engines give far too much relevance to date, especially for evergreen content. I am purveyor of evergreen articles, and it is sincerely quite obnoxious to put a lot of research and effort into an original article, only to have some flip shop do a lazy rewrite a few months later, citing my article as its primary source, yet ranking higher in search because it is newer. It's a perverse incentive that encourages low-effort regurgitators, and discourages original work. Grumble, grumble.