So, You Want to Build Something

A friend recently outlined a problem he faced as an attorney and wanted my thoughts. He was curious if some kind of digital platform, app, or other solution could be built that solves the problem, thinking it might be worth pursuing as a side gig.

The idea seemed solid, but there’s some homework I always recommend before you break ground on a project.

Below is a quick outline of the process I recommend:

1. Identify a problem

You’re an experienced professional in your field and you’ve got a problem with the way something happens in your industry. Maybe it’s duplicate effort. Maybe you have a task that’s required of you, but isn’t really what you like doing in your job (think writing invoices if you’re a carpenter).

Regardless, you’ve got a problem and think that if you can solve it, maybe you can solve it for other people. Or maybe just solving it for yourself would save you so much time it’s worth building something to help.

Start by writing down the problem you have and why you think you have it.

2. Validate

Your hunch is a great start, but unless you want to start a vanity project you need to get out of your own head. Start talking to other people that might have the problem. Find 10–15 folks like you who might also face this problem and find out:

  • Do they have the same problem you’ve seen?

  • How big of a problem is it for them?

  • Are they investing money in a solution? Do they like the solution?

  • Does it take time away from them being better able to serve their customers?

  • How much does this time cost them?

  • Who is affected? Just them? Them and their customers/clients?

  • What are they doing to solve the problem? Have they created workarounds?

  • Does their solution create new problems?

Try to ask neutral, open-ended questions rather than pointed yes/no questions. You want honest input so, regardless of how you feel, keep the emotions dialed back.

3. Synthesize

Review the responses for similarities/differences. The more similar they are the more likely a digital solution can help them. Be careful! If many people share the problem, but have all created different (and divergent) solutions, there’s more examination needed. The solution might need to be so personal that a standardized solution might not work.

Assuming there are similarities start exploring how a digital solution can help.

In general digital solutions are useful to:

  • Coordinating asynchronous, non-colocated communication (Social Media)

  • Providing existing information in a new context, time or place to solve for information asymmetry (Zillow)

  • Compare, calculate, or compute a fixed quantity of data (Just about anything)

Start writing down what information needs to get moved around. Can you imagine how automating the most commons solutions you’ve uncovered could work?

4. Landscape Research

Who else is in the space? Do they have features aimed at the problems your target audience has? Where do they fall short?

Are there similar solutions in different industries that can be repositioned for your audience? What lessons can be learned there?

5. Define Solutions

By this point, you should have a good understanding of your audiences’ problems, how they’re currently being affected by them, and what some solutions might be.

Next look for low-hanging fruit. What’s the smallest portion of the problem you might be able to solve while still providing value? Be discerning here. Explore the smallest set of features that will actually make a difference in your users’ lives. Make your scope as small as possible, but no more.

6. Prototype

After these steps, you’ll be in a better space to start pulling in craftspeople.

Alternatively, can you prototype it yourself? A lot of processes can be automated via Google Forms. Try some process automation tools like IFTTT, or the Zoho Creator. Maybe Airtable.

Once you’ve been able to create a small-scale version of the system, you can start seeing if you’re on to something, or if there are additional hidden challenges in the way.

Next
Next

AI Interactions: Knowing When to Have a Smart AI Play Dumb