Interview with Patrick McKenzie on Programming and SEO

by on June 21, 2011 | posted in SEO Theory

Patrick McKenzie is someone – that if you’re a frequent reader of this blog – you probably haven’t heard of. But you should have heard of him. And now, you will have heard of him. He’s best known for his incredible contributions over on Hacker News, the social news site for hackers. On Hacker News, users – very smart, mostly programmer-type users – use the voting “karma” system to upvote the comments they find most useful. Patrick’s comments have made him the  2nd most heralded user in terms of total votes on the site (his alias is Patio11) – and for a community with such intelligent users and high standards, this speaks volumes about his breadth and depth of knowledge.

I think he’s particularly relevant to the readers of this blog because he offers a unique perspective into the world of SEO. He isn’t really known in the overarching SEO community (which is perhaps his want), but on Hacker News, he is greatly respected for his SEO knowledge. You can find some of his solid SEO advice on his blog’s greatest hits section. He is also a great hacker, and has a good domain knowledge into that world – and has some cogent advice on whether SEOs should learn to program – which is a big talking point in this interview.

Patrick is a unique case study in that he lives in Japan and has been there since 2003, but fits the Hacker News mold in that he self-orchestrates his own projects, currently Bingo Card Creator and Appointment Reminder. You should check out his blog and follow him on Twitter here.

1. You are well known in the Hacker News world as an expert SEO there. Outside the hacker world, you aren’t known for your experience or expertise in the space, other than a one-off post you did for SEOBook recently. Do you have any interest in branding yourself as an SEO? As someone that is somewhat disattached from what may be considered “standard” SEO in terms of community, advice etc, is there anything you notice of particular interest about the community and the advice, people and ideology that comes out of it? Is there anything in particular about being a hacker (besides the obvious skills) and having that mindset that helps with SEO work? Or, perhaps, hurts it?

Patrick McKenzie: I am not terribly interested in branding myself as a SEO.  First, it is a branding which has definite negative connotations for many people whose opinions matter to me. Second, there is very little business incentive to do so.  I have enough consulting clients at the moment without venturing further into the SEO field, where

a) most people have little money
b) hucksters abound
c) I would be competing with established brands (including those of friends) on one side and an army of low-cost low-value providers on the other

Also, I don’t want to pigeonhole myself as an SEO for the same reason I don’t describe myself as a “programmer.”  I program, sure. It underlies most things that I do to make money.  However, when you combine programming with all the other soft skills (marketing, communication, etc), the value of it increases tremendously.  Some of my consulting clients have me program for them, but I can justify a rate much higher than an equivalently skilled Ruby/Rails programmer.  Ditto for SEO.

The SEO community has many facets to it.  I sincerely appreciate many members, including Aaron Wall, Rand Fishkin, and a few others whose blogs lit the path for me some years ago.  On the other side of the coin, many SEOs primarily seem to extract value (for example, by doing affiliate work for seedy industries) rather than creating it. I could honestly do without that sort of thing, which is why you don’t see me hang up my bingo spurs and just start slinging acai berry or whatever the scam du jour is.

The hacker skill- and mindset are INCREDIBLY useful for SEO. Seriously, if you can’t program, programmers who know SEO look like absolute wizards.  Here’s something which actually happened, slightly anonymized: “Find the top 50 US cities by population for which we don’t have an article written in the database.”  There are many SEOs for whom that is a solid week of boring work.  A programmer should be able to knock something together in an hour or two, at the absolute most. (I could probably get it done in ten minutes if I were racing, but in an actual workday, I’m guessing slacking would make than an hour.  Why rush, anyhow.)  Got keyword lists?  Scripts to the rescue.  Need a tweak made to your CMS?  Open up your editor of choice and just do it. Wish you could get free links from authoritative sites?  OSS.

The mindset that Google’s algorithms are not an oracle but are, instead, a collection of sophisticated algorithms *which can be understood* is also key to being an effective SEO, as opposed to just parroting the tactics you read about in a blog post in 2006.  The Internet is a complex system with emergent behavior. Complex, system, and emergent behavior are not scary words to engineers — or they shouldn’t be, at any rate.

I wish I could disagree with you on many of those points but I have to resentfully agree in most cases. Something I have thought about constantly while doing my job is “what value am I really creating for the world, or am I just exchanging who gains benefit from created value with others?”. Somewhat paradoxically to current public opinion, I believe the content farms of the world have done the most SEO value-creation rather than value-exchange in recent years by focusing on strong SEO. Of course they have deleted or overlapped real value in some cases but they have also filled the gaps of many previously unanswered queries.

Joel Spolsky, who is the CEO of StackExchange (the company that makes StackOverflow) has a good take on this.  There are, broadly speaking, two types of questions: questions which have a correct expert answer, and everything else.  Content farms address a lot of demand adequately in “everything else”, and in that sense I agree, they are making the Internet better.  If you want to know how to pour water into a glass, eHow has got you covered.  (Less sarcastically, there are a world of common questions which the Internet used to fail at — basically, anything a middle aged housewife in Kansas would want to ask — and many of them can be usefully answered by a content farm writer.  Want to know how to make a paper snowflake?  Yeah, fold fold fold, cut cut cut, there you go.) The problem is with questions which have wrong answers. For $6 to $8 an article, you’ll get unique English text written (ideally) by someone who has some knowledge of what a semicolon does, but you will not get any research more complicated than “Google the question and rewrite the topmost answer”, which may or may not be accurate.  (My personal version of Hell: a web filled with eHow-quality rewriting of Yahoo Answers.)

2. SEOs are somewhat-frequently pushed to learn how to program from various sources. Given a site that is already built out and there are no on-page incentives to learn how to program, how would you gauge/describe the competitive advantage knowing how to program imbues? How much more productive do you believe yourself to be than an average SEO without programming ability? Are there certain job roles/situations where it makes sense to know programming and others where the advantage is negligible? On the converse, is there a sense in programming of the opposite as well, where programmers should know how to SEO, or is this largely ignored/not really cared about as much as it seems SEOs are getting pinged to learn how to program?

This depends on a given SEO’s role and what kind of organization they work in.  In many companies, the team who can actually make changes to the website, particularly those of a sitewide nature, will be skeptical of the SEO’s value.  Being able to “walk the walk” is an easy way to get buy-in from the engineering team: you don’t sound like a valueless airhead asking them to take time from things they think really matter to, e.g., redo pagination such that the site is actually crawlable.

Many programmers, particularly those of an entreprenurial bent, would be well served by learning SEO, which is a heck of a lot easier than learning to program.  At the very least, they should learn to appreciate how valuable it can be.  That said, you can be an excellent and well-rewarded programmer without ever getting near a website.  If you’re doing microcontroller code for car engines at Toyota, sure, you don’t even need to know how to *spell* SEO.

However, for SEOs, you cannot be an excellent SEO without either a) learning to program or b) hiring someone who can for you.  It makes no sense asking how more productive I am than an SEO who cannot program: that is like asking whether I can outrun a fish when on my bicycle. The fish is simply not capable of competing: I pick the area of competition and I say it’s a bike race.  (My main business, Bingo Card Creator, outpublishes every educational publisher in the world, combined, with regards to niche bingo cards.  This is precisely because I can trivially knock together a CMS that cranks them out and they cannot.  If Scholastic Publishing hired a non-programming SEO to attempt to wrest my bingo rankings from me, he would be crushed, despite their superiority in budget and resources.  They could probably buy enough links to grab [bingo cards] fairly easily, but that’s a miniscule portion of the business relative to the long tail of niche activities.)

Programming is also quite helpful for data analysis tasks, which are bread and butter for SEOs.  If you can’t at least script, you’ll be stuck with whatever Google Analytics gives you in their default configurations.  First, GA data is virtually useless after you scratch the surface of it for e.g. search keywords.  Second, think of your competitive position: if you’re an anomalously well-paid peon whose job is to push the right button in Google Analytics and maybe export to Excel when you are feeling adventurous, your job will very shortly be given to a less well-paid peon.  IBM et al are training armies of monkeys how to do “Enterprise Analytics” for $20 an hour as we speak.

Perhaps the better phrasing would’ve been “how much more productive are you than someone that can hire someone but doesn’t know how to do it themselves”. In the 2011 world of SEO, the majority of SEO tasks for small to medium size websites have been done for SEOs (by tool providers such as RavenSEOMoz and Ontolo), thus the identification of “fish on your bicycle” is not an appropriate analogy. There a few one-off ideas I get for tools that I wish I had the programming ability to implement. However, they are one-offs and I rarely feel inhibited by that disconnect – perhaps for a lack of imagination. For websites of medium to enterprise size, there is generally real capital involved that allows a programmer to work alongside an SEO. But there seems to be comparable benefit for one to one thing and the the other to know the other – it just seems that one can either be mediocre or slightly above average at both or be really good at one. I would not argue that programming is far more difficult to learn than SEO, so perhaps the road back to SEO from learning programming is much easier to do. However, there is a compensatory “marketing” mindset that also tends to be more inherent in SEOs and not in programmers, so even if programmers learn the basics of SEO, they may lack the marketing fundamentals that can identify viral marketing opportunities and the idiosyncrasies of using psychology to obtain a link even should they learn the basics of title tags and the like – based on those same characteristics that made them great programmers. Of course, this is not black and white and many SEOs are terrible marketers and many programmers great ones, but when speaking in generalities, this normally remains true.

With regards to SEO tools providers, suffice it to say that I believe the needs of creative, successful SEOs do not map one to one with provider offerings.  Those tools target the least savvy segment of the SEO market, because that is where the money is, and some of the value propositions are… dubious.  If your workflow exactly matches that of the designer for the tool, though, more power to both of you.

Here’s a frequent example from client work: link acquisition is probably the hardest thing to do in SEO.  Many of the tools providers presume a process that looks something like “Identify target page, send email to webmaster, request link on it.  Do this 10,000 times with little variation.”  They then semi-automate this process.  I have deep misgivings about it, partly because I have 100,000 requests from UK casinos on why they really deserve a link from my educational bingo pages.  (The answer appears to be “we were capable of coming up with a fictitious female persona.”)  There are better ways to get links: better quality links, with less drudgery, in a fashion which does not inconvenience hundreds of people whose only crime was ranking for one’s desired keywords.  Programming skill makes some of these much easier, but they aren’t one-size-fits-none solutions.

3. Programming is a term frequently used synonymous with scale. Your current enterprise, Bingo Card Creator, is a unique niche but it also seems to be a limited one (please clarify if this is quite off). Do you have plans to branch off to build a bigger business? How much maintenance does Bingo Card Creator require? Based on your previous responses, it seems that you believe SEO is either A) affiliate crap or B) full of entrenched businesses, so I’m guessing you’re of the thought that very few pure SEO opportunities exist that actually create value for you are on the horizon. Do you have a “next venture”? Do you see Bingo Card Creator as the same business it is now in 10 years?

Bingo Card Creator is my claim Internet fame and fortune (tongue planted firmly in cheek), and I love having it around to be my laboratory for testing new ideas.  However, the path forward for my business is not expanding my evil bingo empire.  In December I launched my second software product, Appointment Reminder (, which does automated appointment reminder phone calls, text messages, and emails for professional services businesses.  I think the upside to this is much, much higher than BCC’s is.  AR is a very young product and only has a few dozen customers at the moment, but I hope to continue growing it using many of the lessons I’ve taken from BCC.  It also uses a very important lesson that I have learned about business models: it is much easier to build a business on recurring revenue than to constantly have to get new customers for one-off purchases.  I have something on the order of 4,500 paying customers for BCC (and 200,000 users), and my revenue for today is *zero* until I can find a completely new person to sell it to.  By comparison, every person I sign up for a $80 a month subscription to AR is worth about $1,000 a year for the life of their business. As for future directions for BCC, I plan to continue experimenting with it, but I wouldn’t suspect it to be a vastly different business in 10 years time.  Then again, it is only 5 years old this month, so who knows.

With regards to pure SEO opportunities which create value: either you’re in publishing, you’re selling stuff to people, or (most promisingly, in mind mind) you’re in publishing in the pursuit of selling stuff to people.  Like I mentioned, I’m not an SEO by profession: I make and sell software. SEO is one of the tools in the box that helps me do that.  I certainly think that SEO can create value in the context of a larger business, and would not malign the field to say it was purely scam operators and black hat tactics. However, intellectual honesty requires me to point out that, if I were to list high-value affiliate niches, I would be listing a rogues’ gallery of unpleasant businesses to be in.  What kid says “I want to market payday loans and rebill scams packaged as nutritional supplements when I grow up”?

Also, as the Internet increasingly becomes important to legacy companies, I would expect competition to skyrocket.  Does that mean you shouldn’t start today?  No.  Does that mean the game is getting harder with each month one waits?  Yes.

4. You’ve been in Japan since 2003. Although based there, it seems like the majority, if not all, of your business is still generated through customers in the United States. Would you describe yourself as fluent in Japanese today? Have you ever thought about creating some kind of technology/software company over there, and if so (and even if not), could you help impart some knowledge about where Japan is technology wise, and how business is conducted differently on the web? Japan is known for their particular business practices (or perhaps we the opposite to them), so I find it interesting to see how that ports over to the web with how do they do business and/or complete transactions online. Relating it back to SEO and our previous topic of very little room for value adds in that regard still remaining on the web, there is still a perceived infancy in many other continents as far as search is concerned – and it’s something my company, and I’m sure others, have thought about as well, if not taken advantage of. Do you perceive this in Japan? Do you believe this to be accurate in non-Americanized countries?

I have an engineer’s lack of love for wishy-washy adjectives, and “fluent” is one of them.  I would prefer to describe it in terms of things I can do and things I cannot.  I can write a business plan in Japanese and get it approved by the CEO of a multinational.  I can be interviewed in Japanese about my business, and when I had a lapse in vocabulary I successfully made it sound more like a planned joke than forgetting a 7th grade vocab word. (

I go to Mass in Japanese and understand it.  On the other hand, I could not write a novel in Japanese, and some fairly routine tasks such as “Describe how to make eggplant parmigiana” would give me more trouble than you would expect.  (“Blimey, what is contextually appropriate way to say ‘slice into thin strips’ again?  And baste? No, not coat in, *baste*.”)  Then again, I could probably not write a novel in English.

I do occasional consulting for Japanese clients.  There are, perhaps, half a dozen people in the world who can do what I do and also speak enough Japanese to get the go-ahead from the decisionmaker at that sort of organization.  That said, I spend more of my time developing products for overseas than for here.  There are a variety of reasons for that.  One major one is that “fluent” is not good enough for success in business — almost every native speaker of English is fluent by dint of being born, but very few of them can write copy well enough to build a following or sell things to people.  If I were hypothetically to create Japan-facing products, I would be hitting that barrier to entry headlong.

Also, the Japanese Internet culture is quite different from the American one, in ways which have material differences for businesses. For example, it is very difficult to get folks to buy software from a website using a credit card, especially for business software, which is where lots of the money is in software.  Japan is a few years behind the curve in adopting SaaS — most software, including packaged software like MS Office, is sold to business by firms with long-standing relations with the businesses.  For most clients, I would be on the outside looking in.  (Conversely, the most indisputably talented SEO in the world could not sell to some of my clients here, because I have the “in” and they do not.  This is not that rare of a state of affairs in America, by the way, but it is virtually universal in selling to businesses here.)  Another difference, especially for consumer applications, is the primacy of the mobile Internet over desktop use.  That changes everything from the publishing environment to business models to how one would get links.  (Ever tried creating a link from a cell phone?  I don’t recommend it.  So if the folks you expect to link to you spend 90% of their Internet time on a cell phone, you’re going to need to find someone else to make links.  On the plus side, your competition may be as link-poor as you are.)

Do I think a company could do big things in Japan or a similar less-mature Internet market?  Certainly.  Would I want to be the guy to do it?  Nope.  Would I suggest it for an arbitrary Western business?  Candidly speaking, for the amount of work it would take to do a decent job of taking one existing product and getting it to success in Japan, you could probably launch multiple products in the US, and that is assuming you have somebody on the team who is savvy to business realities here, has a strong rolodex (metaphorically speaking), and speaks and writes Japanese as well or better than I do. That’s not impossible, but it is a tall order.  US multinationals routinely flame out on 8 and 9 figure investments in Japan.  One would need an appetite for risk and hardship to try it on a 5/6 figure budget, which is more than adequate these days for success in the US.

A quick pitch for my home town of Ogaki: we have a bunch of twenty-somethings here who figured out what the Japanese market wanted from the iPhone, productized it, and have been busy printing money for the last few years.  They’re largely engineering/music geeks whose main marketing strategy was, like many iOS developers, centered around “Put it on the App Store and pray”, and the Apple gods smiled on them. Could that have been done from San Fran by an American team?  Yeah, I don’t think that is an impossible order.  But, again, in terms of reproducible strategies, I wouldn’t rush towards doing it myself.

Thanks to Patrick for the incredible interview! Make sure you check out his blog, and follow him on Twitter here.

Previous post:

Next post: