Note: I no longer believe that product ideas should be validated. One should only build after one has pre-sold or collected signed letters of intent to buy at a specific price-point.
I wrote this for the rest of the team at Neo, where we built new lines of business for Fortune 500 companies. Our projects were 12 weeks long. We designed, built, and launched new products for our clients.
This article is an amalgamation of tactical advice, written for teams of at least one designer and one engineer who want to build a new product more quickly. It can be applied to building a new feature as well.
It's this strange feeling that you're not as productive as you could be because the flow of work and coordination and communication with your team is not at its best. It's the feeling that you are occasionally doing inefficient things, or that you don't know what's going to happen next, or taking steps backward. Your "chaos" as an engineer is different from that of a designer, something to be aware of.
Here we explain each area of chaos, along with some ways to reduce chaos.
As an engineer, it's difficult to know when to jump in and code. Has the design undergone enough feedback from users? Is it at a high-enough fidelity? Will the design change drastically soon? Is there a major flaw that won't be noticed until it gets coded up? These are difficult questions to answer–I'll outline a full flow, and then I'll outline an "ideal" flow that might help us move faster on future designs. Having this conversation with Alex (designer of the project) was very helpful for us.
Each bolded step includes some feedback from the Engineer, if needed. And possibly feedback from another Neo designer, but not too much! Too much feedback leads to more chaos. Advice can make you regress to a previous step in the flow, and it takes much longer to finish it. Also, not all advice is worth taking even though it may be "better," so you may have done yourself a disservice. It all depends on the situation.
It is difficult to choose the next experiment. It is difficult to assess how "confident" you are about the product. Are you truly confident enough to build the MVP right now? Here's the rough progression of experiments that eventually help you get comfortable enough to build the MVP. It is expected that you will run an experiment, it will fail, and you could return even to the first few bullet points! I don't recommend choosing to validate two ideas at once. Ask and share common conversion numbers for each step.
This is hard. No way around it. Think about it for a really long time. Here's a list of ways you can make your product simpler:
It's really best to 100% ignore the fact that your client has an existing audience. You certainly can send an email to a segment of their audience, but this will take weeks to make happen and won't make a significant difference to whether people use/need your product. Your other acquisition channels should already be working.
Growth Engineer at Credit Karma & consultant. Past clients include Aconex, Triplebyte, Neo, Brown Computer Science Department, Voxer, Cloudera, and the Veteran's Benefits Administration.