Brad Dwyer: Risks produce big rewards for Hatchlings
Dwyer (left, photo by Eric Rowley/Juice) is an entrepreneur based in Panora, Iowa, and runs Hatchlings, a Facebook application with over 3.5 million users. He is also interested in mentoring and investing in web ventures based out of the Silicon Prairie.
I have found that one of the most important things to do as an entrepreneur is to mix things up every once in a while. Most of the time your new ideas are going to flop, but every once in a while you'll hit a gold mine.
At Hatchlings, lots of our most popular (and most profitable) features have been ideas I came up with and threw out there not knowing what to expect. Our revenue is now made up of about 25 percent virtual goods sales and 40 percent membership subscriptions. Both of those were started out with a little spark of an idea I had.
Getting Stuck in a Rut
When we launched our virtual currency (called "shells") in the summer of 2008, we pegged our exchange rate to the U.S. dollar. We set one dollar equal to 100 shells. Along the way, we added some bonuses for using low-cost payment methods and some volume discounts, but the exchange rate stayed roughly the same.
Running a sale is something I had long been considering, but I had lots of worries. I worried that it would upset users who had already purchased currency. I worried that it would cause lots of extra work for our support staff. But most of all I worried that it would cannibalize future sales and cause people to hold off on their purchases in the future until the next sale.
Get Over It!
Long story short: we ran the sale and none of that happened. In fact, it succeeded beyond our wildest dreams. Worrying about all of those things delayed the experiment for far too long and, as it turns out, probably cost more in lost potential revenue over the years than I was worried about losing by running the promotion in the first place.
We ran a 25 percent bonus on our virtual currency for one week (e.g., if you paid $10 you got $12.50 worth of shells). This is a graph of our direct pay revenue before, during, and immediately after the sale:
- Before the sale, revenue was relatively constant (the average is represented by the green line, x).
- In a normal week we would have therefore expected revenue of 7x. During the sale it was 37x — more than five times our normal revenue!
- The reminder about the sale that we posted on March 26 generated as much buzz as the initial announcement (including the long-tail). This may mean we left extra money on the table. Next time we do something like this, we will probably run the sale a bit longer.
- So far, it doesn't appear to have cannibalized future sales at all! (As noted above, this was my biggest concern going in). It remains to be seen if there will be a long-term effect or not.
- Therefore, we can conclude that this promotion actually drove sales that would otherwise not have been made at all. In other words, we didn't just entice people to buy "now' instead of next week; we actually created new demand.
The Moral of the Story
Don't be afraid to experiment! Throw things out there, shake things up. You're not always going to get it right, but if you take enough swings, eventually you'll hit one out of the park.