Comments on: Marketing science Q&A with Sean Ellis

By: Tristan Kromer

Tristan Kromer — Sat, 26 Dec 2009 19:10:53 +0000

Hi Sean,

Thanks for the link and the response. Guess I should start going through the back posts.

The survey approach is interesting to me. You imply, but don’t explicitly mention that people accurately self report on their purchasing behavior in surveys. This surprises me. I would have thought there would be over reporting and that when it comes to putting the credit card info in the rates would drop off although the shape of the demand curve predicted would be the same. Is there any further information available on that? Is there any difference between type of product and reliability of reporting?

Thanks again.

By: Nivi

Nivi — Sat, 26 Dec 2009 07:18:38 +0000

Jim, I don’t totally understand your distinction between users and customers. If users are coming to the website “for information,” I wouldn’t survey them for fit. Survey the people who are experiencing your value proposition. In your case, that’s probably the customers.

By: Jim Taylor

Jim Taylor — Sat, 26 Dec 2009 02:23:24 +0000

Nivi- Makes sense to first focus our survey on the “users” to get the right product fit. Once established, implement the business model by surveying the “customers” to make sure there is the right fit for them.

We will work to implement both pieces as we get closer to our launch. Any additional thoughts or ideas on the customer surveys would greatly be appreciated. Likewise, we will let you know how we end of surveying both groups…

By: Nivi

Nivi — Fri, 25 Dec 2009 04:47:04 +0000

Jim, I think Sean would say to survey the users who’ve experienced the core value you offer and have been back recently, say, in the last few weeks. Later, implement the business model.

However, I’m interested in how you later survey paying vs non-paying users and what you look for.

By: Jim Taylor

Jim Taylor — Fri, 25 Dec 2009 03:58:50 +0000

Sean – What are your thoughts about surveying customers versus users. Our customers will be the ones eventually paying for the service, and our users will be those coming to the website for information.

Should the survey be the same for both groups, and if so, should we then get to the 40%+ must-have for both groups… with the same answer of being very disappointed?

By: Sean Ellis

Sean Ellis — Thu, 24 Dec 2009 21:04:52 +0000

Hi Tristan – I wrote a blog post on this about a year ago. First I try to understand which parts of the product experience users consider most valuable for determining the overall biz model. Then I hone in on max yield pricing by surveying the purchase likelihood at different price points (surveying separate groups each time). Generally surveys give me the same demand curve as actual bucket price tests. This process works great for most software or web services. Of course pricing is tougher with hardware or web services that have a high marginal cost. Here’s a link to the full blog post on pricing: http://startup-marketing.com/how-to-determine-the-optimal-price-for-your-web-service/

By: Tristan Kromer

Tristan Kromer — Wed, 23 Dec 2009 23:16:02 +0000

Thanks for the interview and posting the nice follow up questions. Every bit of insight I think is a great help.

I have not seen much regarding how to best implement pricing and establish market value for products once you have a market fit. Are there any best practices in this area that you prefer? Bucket tests? Or just start at the top and work down until you can establish the outline of a demand curve?