Making your wish come true
(How to write for the Astrid Wish list and help us decide well)
What should we do next? At Astrid our team, investors and users love to share and ideas and opportunities about how we could make Astrid better. We love hearing ideas (keep them coming!) and we put them into wish lists for each of our products but we cannot do everything and are committed as a team to a larger mission. Namely we have come together to help over one billion people become happier, healthier and more productive. We currently have helped about 2.5 million people (who have tried our products) and hundreds of thousands who use them regularly, but going from hundreds of thousands to hundreds of millions won’t happen over night. Day to day we evaluate all ideas including new products, features, marketing campaigns or partnership opportunities by asking whether it will move us closer to this goal.
As a team we believe we need to do at least three things to get closer to our goal:
So whenever we hear an idea whether it is an intern, engineer or the CEO we want run it through a quick process to determine whether it will help us in one of these ways. Below is our process.
For every idea that isn’t simple UI polish (aligning images, fixing typos) or a bug (the app does something different than we think it should) we add it to a one of our wish lists in Astrid. As a team we have four (iPhone, Android, Web, Crazy Idea List). Every wish starts out as a priority 0 and has no due date.
Sponsoring an Idea
Before a wish is considered someone on the team has to be willing to sponsor it. Sponsorship simply means you are willing to take 5 to 10 minutes to think through and write down answers to the following questions.
Add a description with enough details that someone who is brand-new to the team would understand. The description should be in the title of the Astrid Task.
2) Personal Commitment: How much of your own time would you spend on this?
Would you be willing to spend time you don’t usually work to make this happen? If you are an engineer you would spend your weekends or early mornings working on this? If you aren’t an engineer how much time would you spend designing, drawing out, and in other ways clearly specifying what it might be? How much time would you devote to validating and testing mock-ups on users to confirm it is worth building? Please consider that you may or may not be asked to spend this non-work time to help us see that this is worth doing. Filling out the rest of these steps is the first test of your belief that this is worth doing.
If you are sponsoring an idea you probably think it will help the company in some way. Likely if it doesn’t help our team or product in a meaningful way it will be passed by for something that will. Before going farther with your suggestion first think which of the following buckets the idea fits and and how much impact it could make if we do it.
Why do you think this opportunity exists? What is your logic? (e.g. Chrome said they would feature us if we do x and it could get us 10-100k users if they. Or according to X study people are 10% more likely to sign up if you have 2 fewer fields or Facebook implemented X and 90% of user reviews say they love it... ).
How we can know if it is a good idea, before and after we build it? (e.g. Before: Survey, Observe comments on similar product, test on sample set of users, implement a false door... After A/B test in production on a subset of users, monitor metrics after change is introduced, customer support response).
How long / how much will it cost to build? Based on a clear written proposal as to what it is, how long do you think it will take to build? If you are not an engineer or the artist who would build it ask for help on this step. Hint: I sometimes find it helpful to preface this request with something like “if you really wanted this feature how long would it take you to build it.” I then take this estimate and double it. This helps us avoid estimates that are either overly optimistic or pessimistic.
Could something go wrong? How likely? How bad would it be? If we remove the feature or change it again how bad could it be? I one time had a user threaten to give us 1 star every day (to stay visible on the Android Market user ratings) if we didn’t add a feature back to our application.
Once every two weeks we review all the sponsored wish list items and decide which to test, build, defer or reject.
Example
Create an ugly-profile-to-cat converter
Description:
Filter ugly people’s faces automatically and replace them with images of cats run profile photos through HOT or NOT and photos that score less than 4 we automatically replace with a photo of a cat when viewed by others. The user would still see their own original photo.
Personal commitment:
I will spend 10 hours designing an testing a basic version on users and running 50 friends profile photos through this process to see what would happen.
Opportunity:
70% of people like looking at cats and only 20% of people like looking at ugly people according to search volume for “cats” vs “ugly people” (completely made up numbers). Therefore having more cats and less ugly people will help users spend more time spend more time on the site and increase user retention.
Basis:
a) Search volume of “cats” is 50% great than the search volume for ugly person. (made up).
b) Report from friend at a large social networking sight that they do this on their news feed and they probably measured this on their large userbase before doing this (made up but this sort of logic may find its way into our discussions).
Test:
Before: We should create screenshots of our pages of our application with a normal distribution of people including ugly people (Hot-or-not scores less than 2) and another screenshot with these same people replaced with cats. Test 50 people by showing them the screenshot and ask them after looking at the page for 30 seconds “from what you have just seen how likely would you be to recommend this application to a friend”).
After: If the initial test is promising for our new feature build it and test this on 5 percent of users and track their retention vs other users. Do people who are shown cats spend longer on the site? Do they complain less or more about this?
Costs:
Risk:
Because of the high risk and the unclear benefit in user retention we would likely choose to defer or run a some simple tests.