Guest post by Jason Cohen.
Some startups die because their business model is platform-dependent. Like this:
Used to be, Twitter let you post anything, even advertisements, even if you didn’t identify those tweets as ads. Companies like TweetUp and Ad.ly popped up as liaisons between advertisers and popular Twitter-ers. Specifically, an advertiser might be willing to pay $3 to have someone with 10,000 followers tweet about how great their widget is, as in this Ad.ly example:
It’s like AdSense for Twitter, if you like those kinds of analogies.
But in May, Twitter announced their own advertising platform and, in the same breath, disallowed “any third party to inject paid tweets into a timeline on any service that leverages the Twitter API” [TechCrunch]. Twitter’s Terms of Service now explicitly prohibits it.
So what happened to TweetUp and Ad.ly?
If you visit TweetUp’s website you’ll find their name is now PostUp, and it’s no longer AdSense for Twitter. Now what they do is… ummm… a pile of content-free marketing buzzwords. Their slogan is “World’s Best Tweeters… and more,” which the long-time readers already know I think is dumb, and their “About Us” copy is similarly useless:
“We’ve embarked on an exciting quest to let even more people realize the tremendous potential of social networks as a real-time information platforms.”
Oh happy day! Just want I’ve been waiting for! Wait, what is it exactly?
Oh well, maybe they don’t know either.
(P.S. Did you spot their grammatical error? Zen koan for you: If a sentence has an article-count mismatch and no one cares to read it, is it still grammatically incorrect?)
Ad.ly is still around at least, but wisely pivoted into matching celebrities (like Mandy Moore) with major brands (like NBC). They “survived.”
Both of these companies have millions in funding, so they can afford to try new things or just flounder around. Bootstrapped companies have no such luxury.
Of course you can’t conclude that using another platform is automatically imprudent. All technology depends on “platforms” — operating systems, browsers, libraries, even pre-existing customer prejudices. But you can say that some platforms are still undergoing tumultuous change and are harder to depend on, particularly if your entire company depends on it.
The Apple App Store is perhaps the best example of a tempestuous platform, famous for screwing innovators without notice. Scandals fly almost weekly about some iPhone app inexplicably prohibited (or allowed). One telephony app is banned, the next is sanctioned. One day you’re an awesome analytics platform like Flurry, the next day Steve Jobs calls you out by name because you’re “pissing us off.”
If an iPhone app is a fun, useful addition to your offering but it wouldn’t kill you if it vanished tomorrow, that’s great. If your entire company is based on an iPhone app, this is critical.
In my opinion SEO is the same way. It’s an ever-shifting landscape of shit you chase without ever really knowing what the rules are. If you’re doing SEO as one of ten tactics to get people to show up to your site, that’s great. If you’re betting the entire business on getting traffic through search results, at any moment a competitor or the platform itself could oust you.
Some things look like scary platform lock-in but aren’t really. Take Heroku, a fully-managed Ruby on Rails platform. You “git push” your code and it’s automatically deployed across multiple redundant servers, as scalable as your wallet. Great, but they have all this black-box magic infrastructure that you’re now depending on, so isn’t that dangerous?
No, because it still just Rails and Postgresql — all the tools are freely available and you’ll have them installed on your laptop anyway. So if you had to leave Heroku, you’d just get a hosted server somewhere, “git clone” your code, install a database, and restore from a backup. Of course it would be a pain in the butt and could even mean a day of down-time and lost data, but that’s true no matter where you host. Your company is still alive. That means you’re exploiting Heroku as a tool; your business’s success does not depend on them.
My company WPEngine depends on WordPress, which is also a platform. (In fact, many people call us “The Heroku of WordPress,” an analogy I love but most WordPress folks don’t know what Heroku is, so unfortunately it’s not a good home page marketing message.)
But WordPress is different again because it’s an ecosystem, meaning not just a single technology controlled by a single company but a guild of many thousands of consultants, products, companies, and power users, all of whom depend on each other for revenue, attention, and credibility. Unlike Twitter, where social media marketing consultants don’t care which Twitter clients are most popular. Unlike Apple, where they rule with an iron fist no matter the consequences.
WordPress itself is open source; Automattic (the company behind WordPress) couldn’t kill off this community if they tried. But also we have five years of evidence that, quite the contrary, they love, support, and encourage this community, up to and including helping companies like WPEngine, like when we cut a deal with them that landed both of us on TechCrunch.
So even if you want to argue there’s no difference between a “platform” and this tangled knot of forces I’m calling a “ecosystem,” I’d still argue this one is stable and supportive with clear direction.
Salesforce is one of the greatest, smartest curators of their ecosystem. They intentionally don’t add really obvious features to the base product, exactly because it would harm companies who make money adding those very features. Salesforce knows they’re the shark, and ensuring the survival of the remora fish is vital for their own health and continuation.
So one way to avoid the platform-dependance problem is to not base your company around something that you already know is dynamic and out of your control. The other way is to diversify. Like with SEO for lead-gen, you still do it, but it’s not your only means of lead-gen. If it disappears you still have nine other ways to get people to your website. If you’ve tried nine things and none of them are working… I guess the problem is with you, not the ten platforms!
In the end you have to depend on something, and it’s always possible the rug could be pulled our from beneath you. At least don’t pick a rug that’s already moved around the house three times.
How do you avoid marketing platform dependence? Or do you think this argument is overblown? Let’s continue in the comments.
This essay was originally posted on November 29, 2010 on Jason’s blog, A Smart Bear.
Jason Cohen founded Smart Bear Software, maker of Code Collaborator, a tool for peer code review and recent winner of the Jolt Award. He took Smart Bear from start to multiple millions in revenue and 50 percent profit margin without debt or VC, then sold it for cash. He is also a founding member of ITWatchdogs, another bootstrapped startup which became profitable and was sold. He’s also a mentor at Capital Factory (like TechStars or Y-Combinator in Austin). And he’s the author of Best Kept Secrets of Peer Code Review, the most popular book (35,000 copies) on modern, lightweight methods for doing peer code review effectively without everyone hating life. He blogs at “a smart bear.” Email him: jason (at) asmartbear (dot) com.