sign in   |  about

Breue is a community for shaping products

Animated beer

Breue is a community of people who help craft better products. Breuers are designers, coders, product managers. We look at every detail of a product and collectively shape it into something more functional, good looking, valuable.

Next  Take a second to read the stories of our first breued products

Why do I want to be part of this?

This is your typical feedback :(

Looks interesting... I love the minimal design. This definitely has potential. Let me know how this project ends up... Good luck!

And this is feedback from a Breuer :)

tl;dr: Make recommendation quality near-perfect. Guide the user to the “right” answer.

(0) I'm going to give my 30% feedback.* But first, just a bit of context. My most recent role was running product at StumbleUpon. The value proposition of StumbleUpon is, roughly, "Answer a few questions, and we will magically give you the best recommended web page for you." Sound familiar? Before that, I was at Google working on AdSense, a platform that analyzed the entire web and chose the best recommended ads for the given user and page. See a pattern? OK, enough silly bio stuff. On to the feedback.

(1) I'm a very big fan of doing something simple, but do it better than anyone else in the world. Some examples of moderately successful companies that have done this: Google (type your search, voila!), Twitter (write 140, read 140), Facebook (see what your friends are up to), Tumblr (post something short), Instagram (post pictures), Evernote (take notes, store them in the cloud), etc. The point I'm making is, rarely does a company initially succeed because they tried to be too many things.

(2) Your platform falls into this category: pick a product, answer some questions, get an answer. So you have the "keep it simple" part of it down. The challenge for you is at least two things: how well you deliver on the perfect answer for the seeker, and the information architecture and user interface choices you make to bolster the user's confidence in the fact that you have given them the right answer.

(3) Let's take the first one: Get it right. There is a reason why that Google's search and ads quality team dwarfs most other engineering teams at Google. Google, to a large degree, lives and dies by the quality of search results and ads. This means doing everything you reasonably can to collect every implicit and explicit signal out there that might help you make your recommendation quality perfect. Analysis is nothing without lots of good, clean data. This includes all the sources you and others have mentioned, but should also include any other data source that could be useful: behavioral data, social graph data, heck, historical weather data if you think it might help (think umbrella recommendations). It can also mean that you adapt what kind of data you get based on previously acquired data. For example, if someone answered that their budget was limitless, I would think that would change the nature of the following questions you will ask, or even the order you ask them.

(4) Once you have good, clean data, then you can start to focus on how you structure and mine that data to find what you need. This is the part where I’ll let you do your thing. This can be a whole other conversation, but without knowing a lot of the details of what you’re doing, it doesn’t make sense to go back and forth on a forum like this. I will say this: right now, your product is Consumer Reports, but instead of experienced product testers giving their opinion, the crowd in the form of Amazon reviews, etc. are giving their opinion. Is that your only differentiator? Are there other specific ways to stand apart from CR? (Full disclosure: I’m a long-time subscriber of CR).

(5) Once you believe you have the right answer based on the signals you’ve taken in, how you take in and structure the display and experience of the answer matters a lot. Think of it as expectation vs. experience. What does the user expect in a product that will help them make purchase decisions by answering a handful of simple questions? How does their experience match what they were expecting? Is it worse, better, different? “Give me a product recommendation,” isn’t the same as, “Do all the thinking for me.” One of the only ways a person can assess the value of something is through comparison. If you give them a single product, they don’t know what to do with that. But give them two products, and now they can see that one is better than the other. Even better, give them three and they can feel like *they* made the decision, not some computer algorithm that gave them an either/or. I think that’s why some of the amazing people in the thread below are suggesting a list of recommendations, not one-by-one. Hard to compare that way.

(6) But that doesn’t have to be the end of it. You can guide your users to notice your top recommendation more. Make the top one slightly bigger, or have a larger thumbnail. Make some of the UI a slightly different color for the top choice. Google search results are displayed in a list (a single search result would just be the “I’m Feeling Lucky!” button which gets negligible traffic), and the most relevant result is at the top. Top = best. Top + different color = bestest. And what is above the search results in a slightly different color? Ads. Google ain’t no dummy.

(7) And of course, measure everything. EVERYTHING. One cannot change what one does not measure.

Apply and get your spot in line

Apply for breuership  |  Apply to have your product breued