More than a tweet, less than a blog.

So there’s an interesting article at Quora on Amazon’s process for product development, and a link to another great piece on the same subject. It goes a little something like this:

1. Come up with a new product/feature.
2. Express the suggested product/feature in the form of a press release.
3. Test the press release on prospective customers (and I assume, press) to see if they get excited.
4. If they get excited, continue. If not, I assume there’s some kind of iterative process to improve 1, 2 or both.
5. Use the press release as a spec to make sure that the development team is focused on the features that got customers fired up.
6. Deploy and send out press release.

They call it “working backwards”, or “starting with the customers”. Interesting enough, I see a lot of ad agencies that use the same process for crafting campaigns - the assumption being that a good press release key to get “earned media”.

I love the idea of “working backwards” to keep priorities in check, but I think that it can be applied beyond the press release, which is only helpful when PR is in fact the main distribution channel. Another issue is that it’s really hard to test a press release in any other way than actually distributing it. 

Asking a customer “what do you think?” isn’t a great indicator for how well your idea will spread, and since a journalist/blogger will interpret your release it’s hard to guarantee that your story will be replicated with a high enough fidelity.

Press release or no press release - the takeaway here is to simulate “the moment of truth”. In some cases this is getting a spot to present on DEMO - if so, mockup a 5 minute product demo. In some cases it’s getting Crunched - if so, write your idea in the form of an email to Arrington. If your moment of truth is a landing page after people have clicked a search link at Google, build a landing page to do a ghetto pretest your idea.

In some cases (most cases?) the moment of truth is a sales call - if so, build it in the form of a sales deck with screenshots, complete with a problem description, pricing and some corporate bullshit chrome.

The last example - the sales pitch and a deck - is my favorite (and what I do all the time at Burt), since you can actually test it for real by “getting out of the building”, sell it to a live customer and see if they buy it.

The worst thing that can happen is that customers say yes and want it right away.