Technical journalism

I’m not exactly impartial to the following story, having been a technical journalist myself for long swathes of my professional career. I was employee #1 at Hexus, one of the longest standing computing news and reviews sites on the web today, working there between late 2000 and mid 2006. I’m also still part of the team at Beyond3D. Part of the discussion forums since late 2003, I took over as Editor-in-Chief of Beyond3D straight after leaving Hexus and I’ve been either Editor or Deputy Editor there ever since.

After trying and failing to get Beyond3D to generate an income to keep me working on it full-time, I made money as a programmer for a while, before deciding that, yes, graphics was actually the thing to get me out of bed in the morning, leading me to get a job working on PowerVR.

I never quite left Beyond3D when joining PowerVR, though. Imagination, PowerVR’s mothership, have always been understanding and supportive there, letting me work on it to keep it alive and running on the Internet and help Alex, the current Editor-in-Chief, where I could, given the bounds of my freshly signed contract and NDA.

That’s let me keep one eye on technical journalism as I see it, at least as far as understanding how it works for coverage of modern computer technology and graphics in particular.

Then there’s The Tech Report, a site I’ll come back to here and there in this essay. Even back when I was writing for Hexus, where The Tech Report were my competition, I was always a fan of Scott and Geoff’s work. I’ve followed them ever since I read Scott’s article on the 1.7 GHz “Northwood” Pentium 4, reading the site for free almost religiously ever since, checking back almost every day to consume all of their new content.

In the beginning I was a true PC enthusiast, spending every spare penny on my computer to improve it, absorb its new technology and keep myself on the cutting edge of what was going on, in order to inform my analysis for Hexus. And no matter whether I was working on my own pieces or consuming others, there was a really great community of websites out there, often doing it on a shoestring budget like we did at Hexus. I’d meet other writers at various events and realise that what I was doing was exactly what they were doing; consuming technology advancements personally, in order to do a job professionally.

I happened to meet The Tech Report folks at some event or other, probably some editors day for a GPU launch, hosted by NVIDIA or ATI. I struck up a friendship with Scott, in no small part because I’ve always loved talking to him about technology, because he’s always had an obviously great view of it from his writing on TR and he clearly loved doing that job for a living.

We all did it because we loved it. The Tech Report still exists today, in part at least, because Scott and his staff love technology and it bleeds out into the copy. They’re clear connoisseurs of tech, involved as much in enjoying it in their spare time as they can, although maybe not as much these days as we did back in the day, as everyone’s older and the time sinks of wives, husbands and kids have found their natural fit in the order of things.

If I look around the web today for the old sites that found their feet at much the same time as Hexus did, many of them are still there. Not only that, some of them are still run by the same founding teams! Anandtech is a bit different of course, and I’ll come back to them, but sites like Guru3D, Rage3D, Tweaktown, HardOCP,, Bit-Tech and PC Perspective have all been around for well over a decade now.

Looking at their content, things have obviously evolved, but the personalities of the writers at those places are still evident in the copy they produce. I can overwhelmingly read a piece on The Tech Report and Hexus and know exactly who at TR wrote it, or know if Tarinder was the author at Hexus, in among the other, newer editors there that I’m not really familiar with. I recruited Tarinder over a decade ago. I keep referencing the timespan because a lot of my memories of the early days feel like they happened recently.

One of the great and surprising things about these enthusiast and niche publications is that none of the writers ever had any formal training in journalism, technical analysis, or even degrees in written English. I can’t think of a single formally trained colleague from those early days, in any of the areas you’d think you’d need to succeed as a technical writer today. All we had was enthusiasm and access to PC parts, and the ability to cobble together some HTML. We figured out journalism and technical writing, whatever that means, ourselves. Codes of ethics sprang up out of common sense. Testing and analysis methods were developed, copied and iterated on. The ways and means for analysis that were developed in the very late 90s and early 00s are still very much in use today, in a lot of ways.

My relationship with these sites is a personal one as a result. I feel responsible in a tiny way for the way modern PC hardware analysis happens, because some of the methods for hardware analysis back in the day were developed by me. The methods were primitive and crude, but they worked and would still mostly work today. That complete and utter reliance on a new, untried way of doing things on the web, which we developed ourselves, meant we truly believed in it. The end result was that we absolutely savaged the print guys, being able to print advanced new copy to any schedule we liked and doing so largely completely free of charge to the reader, bar the delivery of some ads to your browser as you read.

It’s hard to understate what the twin effect of free-to-read and constant daily publishing had on the guys with a print run. I pick up a print mag from time-to-time, to check how those that are left are getting on. It’s not pretty; all the best writers now live on the web.

So why have I felt the need to reminisce, and now in particular, if everything is great, organic and something I can be really proud I was a part of, where the original sites are still around with the founding teams for the most part? Clearly they’ve been able to sustain a living, even though I failed to figure out how at Beyond3D.

The reason I feel like writing it down now is I think that enough has changed in the last few years for it all to be at significant risk. The main reasons are hardware related and well covered by the industry; PCs have been partially marginalised by the rise of the smartphone and tablet, resulting in less of an interest by the average person. Whereas a decade ago the majority of someone’s computing was done on a traditional PC, these days that’s increasingly false.

Then there’s the PC hardware itself, which has struggled to justify its continued pace of advancement in the PC mainstream. There are lots of fantastic reasons to be enthusiastic about the onward march of PC technology in components like CPUs, GPUs, storage and memory. But there are increasingly few reasons to be enthusiastic about it in the mainstream, especially outside of gaming. One of the reasons I marginalised myself into the graphics technology space, away from general PC technology, is I couldn’t see a serious need for ever more performance given most modern software, and where software was heading in terms of what it provided the PC end user. I had that view in 2006 and I still hold it today.

The computer I use to type this out was manufactured in late 2009. It got an SSD upgrade a couple of years ago, transforming its every day performance. Every day performance for my needs that is still fantastic today, causing me to have very little reason to upgrade it. I don’t even know what CPU or GPU it has without asking the system information thing. So while I love to keep on top of modern technology as an interest and enthusiast, I’m happy using technology that was cutting edge 5 and a bit years ago for the most part. My PC can’t really game, but then I switched to doing that on the least compromised gaming platform that I feel wastes my time the least: consoles.

I have a much more powerful computer sitting to the left of this one, actually. It’s reasonably modern, with no component more than 2 years old, but it’s no different in feel and performance to this computer for the majority of things I want to do with one. Mostly because my older machine has an SSD. I got it for writing GPU software, because the GPU is great and much better than the one in this machine. But it’s noisier and consumes more power, which are more important metrics than more CPU and GPU performance in the general case, for me at least.

So in 2010, when I bought the computer, I stopped being an active PC enthusiast and became completely passive instead. The rise of the smartphone and tablet pushed me down that path in part. Heck, if using my phone or tablet for writing my blog was a great experience, I’m not sure I’d even switch this computer on that much.

Looking around me at the world, I tend to believe that my experience is mirrored in the average person. I don’t actually know anyone, bar people in the computer industries, that actively uses a desktop or laptop computer as their primary computing machine.

That leads me into why I think technical journalism is in trouble, at least for analysis in traditional general purpose computing terms. The word traditional is important. If I believe the world has moved on to different kinds of computing device, I have to believe the audience for reading about general purpose computers has too.

I don’t even really have to believe, either. I just have to look out onto the Web. In the last five years we’ve seen the rise of a new breed of website that covers computing technology, completely different to the Hexuses, Tech Reports and Anandtechs of the world. Sites like The Verge and Engadget, enormous in size with huge staffs of full-time and freelance staff, covering everything out there in terms of modern personal computing. When I say modern personal computing, I mean things like smartphones, tablets, advanced wearable computers, and car technology. I also mean, in no small part thanks to Edward Snowden in recent years, sites that now spend a lot of page area covering the pertinent issues around the use and application of modern computing technology to people’s lives today, alongside the raw technology itself.

You don’t see The Verge or Engadget review GPUs or CPUs or mainboards. They don’t care that SSD technology has marched on in leaps and bounds in the last few years, all while dropping drastically in price, causing the biggest beneficial changes in using the PC by a single component in the last decade.

Coverage of that kind of technology — the bits and pieces that the sites I grew up on were, and still are, devoted to — doesn’t mean anything to the layperson any more. Understanding it is marginalised now. If it did matter, those huge new publishers that have sprung up to dominate the technology reporting landscape, almost out of nowhere, would cover that stuff.

And it’s not just marginalised in terms of “mobile vs PC”, either. Those publications are even marginalising the core technologies inside of the new mobile devices they spend most of their word budgets on. Look at them and tell me the last time any of them wrote more than a handful of words about the GPU inside the phone or tablet they just reviewed. A sore point, because I help develop the GPU technology in those devices, but it’s true. The GPU is one of the key things that makes the experience of a modern mobile device as magical as it is today, yet nobody in the mainstream technology press cares. They don’t care because the user largely doesn’t care. They just care there’s a GPU of some kind with some performance, where that performance is better in their new device than the device it replaces.

The layperson just doesn’t want to understand. And that’s fine, there’s a lot more to care about in the world of technology today than the performance of the IP blocks in the SoC.

Circling back on the founding technology sites of the modern web, I think Anand called it quite well when he decided to give up his life’s work to go work at Apple. He’d built the biggest technology site on the web by miles, then watched it be overtaken by others who looked less closely at technology and how it worked, and more closely at technology and how it’s used and experienced, especially in the amazing growth areas of personal mobile devices like the smartphone.

What of the smaller sites, where the founders can’t as easily cash out and move on to do something else, because of the economics of riding in Anand’s wake, along with the increasingly difficulty to find a core audience?

It’s hard to know. The Tech Report have tried a new model in the last year or so, with subscriptions. You can now pay, although you don’t have to, for the site. You get a few nice perks, but really the core premise of the subscription is to give the site a revenue stream it can trust, leaving the team accountable not to advertisers, but to you the reader. It’s the relationship that technology publications like TR, Anandtech and Hexus should always have had. In hindsight, I don’t know why I didn’t see that back in 2000 when I started in tech journalism.

I used to buy print magazines regularly. I bought them with fascination at one point because I was helping to kill them, and I needed to understand exactly where I needed to be better, to get their audiences to come and read my stuff and make Hexus successful, as the Internet grew and grew. So why didn’t I suggest to David that he charge people a bit less than a magazine subscription every month, in order to read the site and let us focus on their needs exclusively? Web publications inherited the same broken financial models as the print guys, and didn’t innovate. We innovated on the publishing, but not how it was paid for.

So while TR’s model isn’t perfect — somewhat because people are are cheap (and rightly so, I’m not knocking anyone who can access something legitimately for free and doesn’t pay for it), but also because readers don’t have a truly compelling hook yet to make it a must for them to take out a subscription — it’s the best basic model I can think of.

I believe it’s the only model that can really work for technical analysis on the web that’s still absolutely necessary to certain audiences, where the publisher can remain independent, do great work, and survive.

There will always be people interested in hearing about how the latest CPUs perform in key, niche workloads that drive their use of PCs. There’ll always be a desire for PC gamers to know whether the new GPU they have their eye on is going to be great value for money. There’ll always be someone who wants to know whether a certain SSD brand tends to make reliable products, so the new SSD they have their eye on with a new type of NAND might be worth a shot, because the prices keep getting really attractive. It’s just harder to find those people I believe.

The kinds of reporting and journalism that I think it’s key to support in a different way now, aren’t in need of that support because they got worse and declined in quality badly, or did something else fundamentally wrong and disadvantageous to their readership. It’s because the whole world moved on a bit, making their work less relevant to the masses. They became niche, rather than mainstream. Sites like The Verge aren’t doing anything wrong either, pushing the previously mainstream sites to one side, they’re just reacting to the newer mainstream views of modern, popular technology and covering it in kind.

It’s now therefore not unreasonable to ask those people still interested in niche reporting to directly pay for the analysis that lets them make their purchasing decision, or inform them about a product in a way they find compelling. I believe there are other ways to do it than direct subscriptions that need to be explored, but I believe the core point is clear and compelling:

Either actively pay for the less mainstream technical journalism out there on the market today, rather than consume it passively via adverts you probably block because you’re tech savvy, or that journalism will die completely and you’ll have no way to easily get that information any more.

That’s especially true where it’s really great, in-depth analysis that really tells you something new about the product that you couldn’t easily find somewhere else, because the journalists that wrote it really put hard work and effort into analysing as many facets as they could, and where you can get a real sense of the personality behind the copy and build up a feeling of trust and rapport with the authors.

That kind of niche reporting has to be supported directly these days. It’s a bit of a strange thing to call Anandtech a niche website in terms of modern computing, but it increasingly is. If we want to see websites like it survive, I think we now have to engage with the sites directly and support them ourselves, and not hope that the adverts that we probably block are going to do it for us.