6 ms·

That's very cool, Michael. Really great to see your business doing well, congrats! Really interesting to see that adding a metal box and making the product mor

That's very cool, Michael. Really great to see your business doing well, congrats! Really interesting to see that adding a metal box and making the product more premium made a huge impact on revenue. Have you run any pricing experiments after that?

mtlynch 3y ago
Thanks for reading! >Have you run any pricing experiments after that? No, I haven't tinkered with the price much for the past few months. I have a hard time doing price experiments because at my volumes, I have to collect a few weeks of data before I think the evidence is strong, and I don't have a good way to A/B test different prices in parallel, so it's easy for two different time periods to have different results independent of price. I've also found that there's a lot of friction around changing prices too frequently. When I reduce prices by a significant amount, we get emails for a couple of weeks after from people who are annoyed they paid a higher price just before a reduction, so I've been reticent to invite those complaints. I don't have strong evidence for it, but $399 for base and $499 for premium "feels" right because they're round numbers and they're close to optimal based on different prices we've tried.
- dheera 3y ago
  > we get emails for a couple of weeks after from people who are annoyed they paid a higher price just before a reduction What if you just explicitly guarantee a price match for say 30 days? People would feel safe and you can do your experiments.
  - mtlynch 3y ago
    Yeah, I think there are ways we can do it. It's just that with every hoop to jump through and layer of friction, it makes it a little less appealing than other paths to increase revenue, so I end up not doing it. I think it's very possible that my prices are suboptimal, so I will think more about how I can do some price experiments while minimizing the downsides.

mtlynch 3y ago

Thanks for reading! >Have you run any pricing experiments after that? No, I haven't tinkered with the price much for the past few months. I have a hard time doing price experiments because at my volumes, I have to collect a few weeks of data before I think the evidence is strong, and I don't have a good way to A/B test different prices in parallel, so it's easy for two different time periods to have different results independent of price. I've also found that there's a lot of friction around changing prices too frequently. When I reduce prices by a significant amount, we get emails for a couple of weeks after from people who are annoyed they paid a higher price just before a reduction, so I've been reticent to invite those complaints. I don't have strong evidence for it, but $399 for base and $499 for premium "feels" right because they're round numbers and they're close to optimal based on different prices we've tried.

dheera 3y ago
> we get emails for a couple of weeks after from people who are annoyed they paid a higher price just before a reduction What if you just explicitly guarantee a price match for say 30 days? People would feel safe and you can do your experiments.
- mtlynch 3y ago
  Yeah, I think there are ways we can do it. It's just that with every hoop to jump through and layer of friction, it makes it a little less appealing than other paths to increase revenue, so I end up not doing it. I think it's very possible that my prices are suboptimal, so I will think more about how I can do some price experiments while minimizing the downsides.