Founder Friday is a weekly guest post written by a founder who is based in or hails from the Silicon Prairie. Each month, a topic relevant to startups is presented and founders share lessons learned or best practices utilized on that topic. September’s topic is how to market your startup without a full-time marketing team.
About the author: Kyle Rogers is the co-founder of Knoda, a startup that aims to bring accountability to the world by tracking predictions from professional pundits and average Joe’s. Their prediction-tracking apps are available on iPhone, Android and the web through Knoda.com. Knoda predictions can be seen all over the web on 3rd party websites from major daily newspapers to personal blogs.
My co-founder James Flexman and I both hail from marketing backgrounds. We worked together at Sporting Kansas City under some brilliant marketers, but promoting a previously unheard of consumer internet company requires very different tactics from those required to promote a sports team. Today we’re happy to share some of the lessons we have learned over the past nine months since Knoda launched.
Things that have worked for us
When we tell our story. Some of our best growth days have been days when we got up and spoke in front of groups. We’ve done this in six different 1 Million Cups cities. We’ve spoken to high school and college groups. We’ve given talks at regional conferences and startup events. When people hear our mission, our founding story or see a demo, they are far more likely to download our app or visit our website. Other opportunities to tell the story include several appearances on local talk radio stations. We’ve been fortunate enough to talk about Knoda on 96.5 The Buzz, 610 Sports, Sports Radio 810 and 980 KMBZ among others. Each of these segments leads to a spike in our registration rates and activity numbers.
When other people tell our story. This one is obvious and seems cliched at this point. Well, it’s cliched for a reason. When people love your product or service, they want others to join them in experiencing it. When a friend tells you about something, you’re far more likely to try it out for yourself. This is harder to quantify until after the fact, but frequently when we talk to power users, they start listing off all the people they know who they brought to Knoda. This is the magic that we must bottle up and replicate over and over again to keep Knoda growing.
When people see our content. The interesting, prescient predictions created by our users are what make Knoda special. We had an unbelievable example of this in February when one of our users correctly predicted the winner of the Esurance $1.5 million cash giveaway that followed a huge Super Bowl ad. He figured out who among the 3 million contest entries would win the massive prize and put that on Knoda 36 hours before the public announcement! It was a great example of the power of our system. As the media caught wind of this, shared the prediction and some info about Knoda, we saw our registration rates climb.
Since then, we have worked to make this process even easier. Knoda predictions can be easily shared out to Facebook and Twitter in a manner very similar to how Instagram images are shared. On days with especially high share rates, our registration rates also climb.
We’ve also created the ability for online publishers to embed predictions within their websites, as SPN profiled a few weeks ago. Publishers are responding extremely well. The Kansas City Star is embedding multiple predictions every week, including regular Sporting KC game predictions on The Full 90 and weekly predictions from columnist Sam Mellinger as part of his Twitter Tuesday routine.
Knoda predictions, like the one below, are easily embeddable in various types of web content.
Our pitch to publishers is simple: you already make predictions (especially those covering sports, award shows and politics), Knoda just allows your readers to weigh in alongside your writers. The publishers often agree, and we see registration rate spikes when Knoda predictions are embedded on sites with high traffic volume.
Things that haven’t worked for us
Paying for users. We’ve tried all kinds of small experiments on the paid side. Facebook, Twitter, Google Adwords, pay-per-install networks. So far we haven’t been able to get the cost down to a palatable level on Facebook, Twitter and Google. The pay-per-install networks create users, but those users don’t return. I suspect this is something that might change at a different stage in Knoda’s life cycle.
You see major companies running ads for app installs, but those are usually companies with serious brand recognition. We don’t currently have that, especially outside of KC and the Silicon Prairie. The magic number for an average person to hear about your product or service and then act on it is seven. We don’t have the kind of funds required to spread the word so that potential users around the country hear about us seven times. Instead we focus on the types of mentions that will drive users to download before the seventh encounter: hearing our story and seeing our content.
We certainly don’t have it all figured out. Every day we work to refine our story and our product such that our growth continues to accelerate. This is the challenge of every consumer Internet company. Hopefully some of our experiences will be helpful to others interested in going down a similar path.