Adnan Ebrahim profile picture Adnan Ebrahim 6 years ago 0

4 Big Lessons I Learned Running A Media Startup In 2014

As the year draws to a close, it's time I looked back and reflected on the last 365 days of running Car Throttle

Remind me later
4 Big Lessons I Learned Running A Media Startup In 2014 - Blog

When people ask me what it’s like to run Car Throttle, it’s hard to come up with an adequate one-line response. “It’s going amazing, we have, like, a bajillion users” or “FML, I want to cry” are two possible replies, and would probably make the questioner wish they’d never asked.

Obviously I’m (only very slightly) exaggerating, but there’s more than meets the eye when it comes to summing up the Car Throttle journey. Over the last year, we’ve quadrupled in size across all areas of the business, and in the last two years we’ve grown a staggering thirty times over. This, then, is a blast through 2014, and a chance for me to talk in-depth about lessons I’ve learned…

1. Comments aren’t dead. In fact, they’re vital

There’s been a huge shift in the value placed on comments this year. Popular blogs like Recode and vertical publishers like Bustle have denounced the value of comment sections, arguing that conversations have moved to social media where users are debating with their own circle of friends. Meanwhile, platforms like Medium didn’t even bother introducing comments to articles in the first place. Bustle CEO, Bryan Goldberg, argues that, “the real estate traditionally devoted to comments is too valuable to be wasted on a component that adds minimal value”.

I couldn’t disagree more.

What you might not know is that prior to Car Throttle I ran an averagely popular site called Blogtrepreneur. The community surrounding the topic I discussed, internet marketing, became the most valuable part of the site and helped land an early (but very small) acquisition. Most importantly, I made online friends who I continue to chat to today.

Car Throttle started life in a similar vein; in December 2008, I launched a simple Wordpress space which allowed me to write about the auto industry while striking up conversations with the few readers who wanted to hear my rambling thoughts. As traffic started to snowball, the number of comments also started to grow, and at that time we used Disqus to manage our interactions.

What I wanted to achieve with Car Throttle was to introduce an element of fun into this sometimes ridiculously lifeless industry, so we nurtured hilarious discussions, encouraging users to post funny images and memes and what I might crudely describe as ‘banter’ into the comments.

Comments per day on Car Throttle over the past year
Comments per day on Car Throttle over the past year

This was key. When we created our own comments platform, the number of interactions continued to climb, driven by the idea that the comments section could be just as informative and entertaining as the article itself. 212,000 comments later, and here are the main takeaways:

  • “Comments will cause lower pageviews per unique visitor” is, in my opinion, complete rubbish. FOMO (the fear of missing out) means our users visit more pages in a session than ever before, in part because they don’t want to miss a funny response someone else has made.
  • Each niche is different. Recode (that tech blog I mentioned earlier) services users interested in tech startups and funding. Can anyone honestly say that they’re passionate about investment announcements? I think not. Recode shut down comments because they received so few of them (this applies to TechCrunch, too) and hey presto, the story was spun that comments were dead. Venture Capitalist Fred Wilson has some great data on this.
  • Owning the technology powering comments is just as important. We require registration to comment not because we want more data, but because it emphasises a relationship between a publisher and an audience that many others are overlooking. Using notifications, we can alert users when they get replies or, maybe like Facebook, when they receive upvotes. And this leads me neatly onto…

2. The rise of the publisher turned platform

2014 saw arguably the largest growth in the funding of publishers we’ve ever seen. Vox Media raised $46.5m to continue growing vertical sites like The Verge, Moviepilot sold their German business for $20m to continue building IMDB for Gen-Y in USA, the aforementioned Bustle raised $15.5m to continue building the home of women’s news, and Buzzfeed raised $50m from VC bigwigs Andreeson Horowitz to change the landscape of online media. That’s a whole lot of money sloshing around the coffers of media companies.

But what’s interesting with all these big money announcements is how each of these sites isn’t just a pure media content play. They all fuse parts of traditional publishing with elements of tech expertise or distribution know-how.

But what’s so bad about the traditional publishing model?

Twenty years ago, the media world worked a little bit like this: Publishing houses like NewsCorp or Haymarket would employ dozens of journalists to carefully research stories, write them up, edit them and fit all the editorial puzzle pieces together to form a daily or weekly mag, which would then be delivered to a newsagent near you. And this wasn’t just limited to written content creation; the same rules applied to TV and film too. A few major players would own the distribution channels (cable companies for instance) and they in turn would strike deals with content creators to come up with some cracking reads, shows or films.

As more people became connected to the internet - that figure now stands at over 3 billion individuals, or 40% of the world’s total population - and as platforms like Twitter and Facebook allowed more and more people to create content, the media world’s lines of separation became fuzzy. Anyone could create great content and spread it to millions of people without needing TimeWarner or Rupert Murdoch.

Today, there is a glut of content. Every internet user has a voice, but as a result, the signal-to-noise ratio online has gone through the roof. Simply put, there’s too much content out there for any single human being to consume. Finite newspapers (with a limited number of pages) have been replaced by infinitely scrollable websites.

There's little value remaining in content creation and distribution
There's little value remaining in content creation and distribution

And where once gaining viral traction was difficult, that content can now gain millions of pageviews in twenty-four hours thanks to the ease of distribution provided by sites like Facebook. You only have to read about Emerson Spartz, the so-called ‘King of Clickbait’ to see how it can be done. I mean no disrespect to Emerson. In fact, exposing market imbalances and opportunities is what entrepreneurs do. It’s just that content creation and delivery have become commoditised. Anyone can do it.

So how does this relate to Car Throttle? We believe that we’re doing what traditional publishers are not. While they create and distribute content and compete with more and more media upstarts every single day, we realise that creating the publishing tools themselves and encouraging stickiness outside of Facebook and Twitter is what counts. We think that significant scale can be achieved inside seemingly small verticals (like automotive) by allowing user-generated content to sit alongside in-house editorial content.

We’re not the only ones addressing this problem. Gawker has invested heavily in Kinja to surface user discussions. Reddit has editorial staff to help with celebrity AMAs. And The Lad Bible thrives on hilarious user contributions. This is content created by fans and power users who love these brands and want to feel part of a community of like-minded individuals.

And so our mantra has, and will always be:

“Why would you want to share this car content on Facebook or Twitter, when you could share it on Car Throttle with fellow petrolheads who’ll be far more likely to engage with you?”

3. Never stop testing. Never stop learning

If you’ve been following Car Throttle for the last year, you’ll have noticed the rate at which the site has changed. Last December, we were still using Wordpress as our Content Management System, still using Disqus to manage our comments, still scratching the surface of Facebook and still struggling to find our content fit on YouTube.

So how did we manage to grow our metrics four times over in just twelve months? Simple. We never stopped testing.

Some of those tests turned out to be valuable data exercises. Take the Garage, for instance. This was added at the last-minute by the product team to see how much users cared about showing off their rides. 10,000 uploaded cars later and we consider this product to be a relative success; it never gained wide traction, but the users who did upload their cars loved the process and continue to vote on the latest cars using what is almost a bastardised version of Tinder.

Why didn’t the Garage gain large traction? From a product point of view, it’s easy to see that we didn’t create an incentive for a user to return back to their car and keep telling us more. For example, what did you do with your car this weekend? Where did you drive? What mods did you just install? Had we added the ability to keep a journal or diary for your vehicle, I think we would’ve seen the Garage used far more frequently. But we prioritised Discuss instead.

And speaking of Discuss: this is, essentially, a product still in beta phase. It’s been an amazing test bed for us and it’s almost shocking to think that just before Christmas 2014, the Discuss feed was receiving more direct visitor entrances than the main homepage!

4 Big Lessons I Learned Running A Media Startup In 2014 - Blog

We’ve also screwed up countless times over the last year, too. From video ideas gone wrong (with a higher than usual like to dislike ratio), to Facebook pages killed off and buried. We’ve invested time into platforms which have never returned any real benefit (cough Pinterest cough), but we’ve been quick to analyse the results and change course as and when needed.

The moment you stop trying new ideas and testing new theories is the moment when your company breathes its last breath. Jeff Bezos, founder of Amazon, recently said that he has wasted billions of dollars on failed experiments. But if he hadn’t spent that much time or money, he’d have never released Kindle, or started the fantastically lucrative Amazon Web Service.

Don’t be afraid to fail, and when you do, learn everything you can about how you screwed up, and then move on.

4. A company is only as good as its team

Now for the mushy part. Tissues at the ready. This past year has taught me a lot about managing people. To many, management is a dirty word and, actually, it’s not a word I’m a big fan of either. However, what you quickly learn is there’s only so much you can do on your own, and only so many hours in the day you can rely on. And more importantly, there are skills which I, as a founder, simply don’t have.

The easiest way for me to demonstrate this is to talk about our team and how each member helps drive us forward:

  • I’d like to think I can make a mean YouTube thumbnail, but our designer James has some truly incredible responsive design and Photoshop tekkers
  • I can understand basic HTML and CSS, but our CTO George has single-handedly architected and managed the whole platform build using languages I’ve never even heard of
  • I can write (poorly, if you’ve got this far), but Staff Writers Darren and Matt can craft some of the best features you’ll read online with their journalistic talents
  • People often (wrongly) credit me with our social reach, but Gabor started and continues to build our social marketing strategy. He knows Facebook like no other man in London (minus Zuck when he comes here for a visit)
  • I can edit and attempt to stand in front of a camera, but Alex chops words together like a literary Mr Miyagi and rocks reviews and car comedy which over 30 million of you have come to watch and love over the past year
  • I can open up Final Cut Pro and use the chopping tool, but Ethan has become a videography and video editing superstar this year. If you’re a YouTube subscriber, I need not say any more…
4 Big Lessons I Learned Running A Media Startup In 2014 - Blog

People make companies. And the only way companies succeed is by having people who have drive and ambition, who aren’t afraid to take chances and who revel in roles of responsiblity. I’m grateful that I have a team of friends who I can rely on to go that extra mile and as we grow in 2015, we’ll be looking to bring in individuals who have the same level of dedication as we all do.

We’re hiring! If you want to change the face of the media industry, then we want to hear from you. Head on over to our Jobs page.

Where to next?

And so that’s it. 24.5 million of you have read over 123 million pages this year (up from a paltry 18 million pageviews last year) and we couldn’t be more proud of how far we’ve come. But there’s still a heck of a long way to go.

What’s on the menu for 2015? There’ll be an exciting financial announcement in January and we’re also on the lookout for talented individuals who want to join our growing team.

But the most important question I know you’re all gagging to ask: will Car Throttle finally release a mobile app? Well, you’ll just have to wait and see…