Should I build a prototype before building my app?

I’ve seen a lot of people dive right in to building without testing their ideas and use marketing to get people to their app. Is this a valid approach, or should I build a low-cost prototype first? Since either building the final product or the prototype will cost money, isn’t building the prototype a waste?

Is Building a Prototype a Waste of Resources?

It’s easy to see building a prototype as a waste of resources. Why not just go directly to building the final app? The real waste, however, is in building the wrong final product. If you keep your prototype lean and cheap (maybe by building a Wizard of Oz prototype), it will cost only a fraction the cost of building out the entire app. Not only that, but the prototype usually doesn’t need to be thrown out to start building the app from scratch. You can build the first version of your app on top of a prototype. You might need to scrap part of your prototype but much of it can live on in your final product.

Why Prototype?

The purpose of a prototype is to provide a low-fidelity version of what you’re building to show your customers. This can be anything from simple sketches of your app to a fully functioning app. Start on the low end of the spectrum, get feedback, and build up from there to maximize the effectiveness of your prototype.

Your customers can the register their feedback on what you’ve built. You go back, make modifications, and bring the new prototype back to customers. Before long, you’ve honed your idea enough to build the final product. If the prototype just isn’t working, you can scrap it an start again without having spent $50k on an app people don’t like. It’s a critical tool to help you build a successful startup. Maybe you could hit it out of the park on your first try, but it’s not likely. The risk is just too great.

Should I Prototype My Startup’s App?

A prototype or some other form of customer development is essential. Investing time and money into building on a hunch and then trying to market it to success is a great way to waste a lot of your money. Save yourself some heartache and incorporate your customer’s feedback into what you’re building. A prototype is a critical tool to help get you there.

How to Build a Super-Cheap Startup MVP

When you have a new business idea, you probably don’t have the budget to build your entire startup MVP. By being selective in what you build, you can deliver the full functionality of your app cheaply and start getting customers on-board. A Wizard of Oz prototype can help you get there.

Introducing the Wizard of Oz Prototype

The man behind the curtain from the Wizard of Oz. The metaphor behind the Wizard of Oz startup MVP.

A Wizard of Oz prototype is one in which most of the app’s heavy lifting is done by the “man behind the curtain.” (That’s you or someone who works for you.) Instead of building all the software up-front, you hold off on the parts that are hard to build, and you perform those functions yourself by hand. Once you have enough traction, build out the app’s key functionality. Until then you and your staff perform the key functions of the app.

How It Works

You build only the user interface and some mechanism to get jobs to you so that you can execute them. This could be as simple as a form where people submit their requests to you. Let’s put it in concrete terms with an example.

I frequently go back to the well of Derek Sivers’ music recommendation engine example, and it works great to illustrate this style of startup MVP. An entrepreneur asked Derek advice on how to get funded to build a music recommendation engine. Derek pointed out that he could make music recommendations face-to-face without any software (which is true). We can take this concept one tiny step further by making the recommendations look like a web app by way of a Wizard of Oz MVP.

Here’s what that prototype might look like in this example:

  1. Set up a WordPress site.
  2. Add a form where users can submit their name, email address, and a list of music they enjoy.
  3. You manually process these requests via email and reply with your recommendations. You apply your recommendation logic by hand and reply with the results.

It’s just that simple. It’s more work for you to process requests, but now you can start collecting critical validation. You can see how many people will actually sign up and get feedback from those who do.

Why Not Build Everything on Day One?

The Wizard of Oz startup MVP is cheap and easy to build. That’s why it’s a great choice for new startup founders with small budgets. You can easily save 80-90% of the cost of building your app outright. In fact, for the example above, you could set this up without a developer on a free WordPress.com site. Building an app like this would cost several thousand dollars even if you contracted a cheap overseas studio.

Besides that, if you build a nice Wizard of Oz MVP, you could even build out the software to back it later and swap it in without your users knowing the transition took place. That won’t work in every case because you may want a better interface to go along with the newly upgraded functionality. If your existing interface is working, though, why rip it out and start over? Just add in the automation, and let your users continue enjoying your app just as they always have.

Case Study: A Social Media Management App

I recently worked with a client bootstrapping a web app with a very limited budget. They had a vision for an app that would use AI to provide hands-off social media management — an extremely expensive app to build.

They were already manually providing this service to some clients, so we started small. We built an app that let users sign up for the service, fill out a survey, and start a paid subscription. At that point, the company kept operating just as it had. The CEO and COO played the part of the “A.I.” making sure their clients’ social media was filled with relevant posts. It didn’t change how the business ran very much, but it changed the perception of the business to both the clients and to potential investors. It allowed investors to see what an app that automated this would be like. Established revenue showed that users wanted it.

Once this Wizard of Oz MVP was up and running and users were signing up with their credit cards, the company used its revenue to slowly build out other pieces of the app and bolt them on to the existing MVP. Each new feature took a little more manual work off them and got them closer to realizing their ultimate goal of full, hands-off automation. They were able to use these humble beginnings to secure an investment without spending a ton of their own cash building the app on day one.

Who Should Not Choose a Wizard of Oz MVP?

Like all things, the Wizard of Oz startup MVP won’t work for everyone. You should avoid it if

  • you already have a budget for developing the entire app

and

  • you’ve already done validation
  • you have zero time to be the man behind the curtain and prefer building the app to hiring a man behind the curtain

A Fast, Simple Way to Build a Startup MVP

By using the Wizard of Oz MVP, you can quickly start getting validation for your startup idea. It requires a very small budget and no experience in programming or web development. It’s labor intensive, but you can be up and going extremely quickly. Most founders skip right over this because it’s not glamorous. It’s far more glamorous than wasting money to build an unproven idea. Aside from running your “app” both manually and offline (which is also a great idea), this is one of the simplest ways to get your startup MVP up and running.

If you’re validating your startup idea, you’re doing it wrong

Students of the lean startup methodology raised an eyebrow as they read the headline. Validation is an important part of running lean. You should only build your startup if you know your idea is a valuable solution to your customer’s problem. That can only come through validation. So, how could validating your startup idea be “doing it wrong?”

Premature Ideation

The validation is not the problem; it’s when the validation happens. As founders, we’re pretty fond of ourselves. We need to be in order to make it through the tough times. We have to be confident and self-assured. When it comes to validation, those are exactly the traits that can cause us to move forward with misplaced optimism and end up building a product no one wants to pay for. We fall in love with our idea and, rather than trying to disprove our assumptions, we use the validation process to try to prove we were right all along. This phenomenon is called confirmation bias.

This is why our order of operation is important. Remember order of operation from math class? If you have an expression like this one:

2 + 3 × (9 − 4)

correctly applying the order of operation is the difference in evaluating it to 41 instead of the correct evaluation of 17. That’s a pretty big error. When you make that with your startup, it might cost you a lot of money building something you can’t sell.

What Comes First?

Instead of spending time developing an idea and trying to go back and validate it, wouldn’t it be nice to start with a valid idea? That’s why you should develop your idea during customer development. Instead of starting with the idea you want to build or even the problem you want to solve, go all the way back to the customer you want to serve.

Once you’ve found your customer, talk to them about the problems they have and how they currently solve them. Once you’ve got these juicy tidbits, you know a great deal about the business you should build. You can select a problem, build a solution around improving the existing system they use to solve it… you might even have an idea of the value of this solution to your customers which will help you understand how to price it later (and how much you can spend building it now).

This Beats Guesswork

The romantic dream of pulling the perfect idea out of the air, building it, and raking in millions is just that: a dream. Even the strategy of coming up with an idea before comparing notes with your customers gets it backwards. Take a shortcut and start with your customer. Your startup becomes a failed startup if it doesn’t meet their needs. Work with them to develop the idea, and you’ll short circuit all the guesswork and wasted resources to start working on a great solution from the get-go.

What’s the first thing I do to start working on my startup idea?

You have the perfect idea for a billion-dollar startup. You’re gung-ho, ready to get started building your dream. What do you do next?

Focus on the Right Thing

If you’re sure you have the perfect idea, you need to start by refocusing. Your idea is not the key to success. It’s the problem you’re trying to solve. Everything you do should center around that.

You don’t realize it right now, but your idea is built on a shaky foundations of assumptions about the problem and the best way to solve it. Some of those will end up being right, but most of them will be wrong. Your chances of success scale along with your willingness to change or even kill your precious, perfect idea. It’s not a pleasant thought, but it’s way more pleaseant than spending thousands of dollars trying to force a flawed solution onto the world.

Love the Problem

Fall in love with the problem not the solution, and the rest will follow. Uri Levine

Co-founder, Waze

Who are you trying to help? Whose problem are you solving? You need to find these people and find out if the problem is really a problem, and, perhaps more importantly, if it’s one they’d pay to solve. This process of talking with your customers to figure out the problem your startup will target is called customer development. (StartupGeist has a great guide to customer development if you want to read about it in more depth.)

You have to be really careful here. If you propose your hypothesized problem and your solution, most people are going to tell you it’s a great idea because they want you to like them. If you’re careful in the way you talk to them, you can get to the real truth. Don’t lead them into either your problem or your solution.

If your proposed problem is not a problem, that’s fine. Keep talking and find out what the problems are. Everyone has some sort of problem they need solved. The first step in building a startup is finding one of those you can solve.

Measure the Cost

If you can quantify the cost of the problem, that makes your job moving forward much easier (although the task itself can be extremely difficult). Here’s what that might look like:

Dave is in the widget manufacturing business. You know modern widget production has a problem: a significant percentage of finished widgets coming off the assembly line are defective. By talking to Dave, you learn that number is roughly 10% of the total widgets produced. Each widget makes $10 of profit for Dave’s company, and they produce 1,000 per day. Now, you can understand, in terms of cost, how painful this problem is.

If 10% of the widgets are defective, Dave is losing 100 widgets per day. Since he’s making $10 on each widget, Dave is losing $1,000 per day.

Find a Solution

This is still part of your customer development process. You’re still working with the customer to find a solution. What have they tried? How are they working around it? Instead of pitching your ideas, get them talking.

Rinse, Repeat

Once you’ve done this with one customer, you’re ready to start building! Oh, wait. Actually, that isn’t right. Once you’ve done this with one customer, you’re just getting started!

You want to sell to lots of customers, so you should talk to lots of them. You don’t want your startup to fail because you got a single false positive. You also don’t want to fail because you got a genuine positive… but it was the only positive and you could only sell to that one customer. As you talk to different customers, you’ll learn new things that will help you build the best solution to help the most people.

Start Building!

Only now that you’ve talked to many customers and validated both your problem and your solution are you ready to jump in and start building.

If my .com is taken, should I buy a .co (or .net, .org, .io, .us)?

I’m afraid I’ll have problems if I choose a different TLD for my startup.

This is a common concern because “.com” is far and away the most common TLD (top-level domain). When normal everyday internet users type in a URL, they expect it to end in “.com.” As an owner of some alternative TLDs (including this site), I’ve experienced confusion when I try to tell regular people my domain. I can’t count the number of times I’ve heard, after giving someone one of my “.io” domains, “Is it dot i-o dot com?”

Some ‘Splainin’ to Do

“No, it’s just dot i-o.”

In general, it’s not great if you have to explain something to your prospective customers about your domain. This is why I caution people against using articles (i.e. “a”, “an”, “the”) or special characters (e.g. “-“) in their domains or getting cute by substituting numerals for words (“2” for “to”). Now, in additon to remembering your domain, the person you’re talking to also has to remember a silly rule about why the domain isn’t what they think it should be. Since most people think your domain should end in “.com,” this is a further example of that problem. You should get a “.com” domain if you can (and if it makes sense for you).

So, I Should Avoid Other TLDs At All Costs, Right?

Buying a domain with an alternative TLD is not the end of the world. It’s not going to destroy your business in the same way getting the “.com” doesn’t guarantee your success. The amount of impact this has on you depends on a few of factors:

  • Who is visiting your site?
  • What is your site for?
  • How are they getting there?

Your Audience

If the audience you’re targetting with your startup is very familiar with the internet, they’ve probably seen and used (or at least chuckled about) some of the newer TLDs like “.christmas” and “.baby.” They’re predisposed to hearing and understanding your alternative TLD domain. If they’re not sure of your domain, even they might be more likely to try “.com” before something else, but they’re much more likely to just search for you from the start.

If your audience is not web developers or otherwise extremely internet-savvy, anything besides “.com” is going to through a tiny wrench in their understanding of domains. Once they get past that barrier, it’s not a big deal… but they have to get through that barrier. They may still have to fight with their understanding when recalling your domain, but they’ll get there… eventually.

The Site’s Purpose

Some people point to vine.co as proof that the domain doesn’t matter. Many other recent apps have launched with domains that use the “.com” TLD but add words to the name of the product (e.g. get___.com, use___.com). This would seem to be a bad idea for a domain since it requires users to memorize another arbitrary word to to get to the site. It doesn’t seem to have much effect on these sites though. In the case of Vine and many of the others, this is because the site is secondary to an app. When Vine launched, the app was the central value proposition. The only way to access the content was through the app. Additionally, the app was downloaded via one of the app stores. The “vine.co” domain was home only to a landing page for the app, which isn’t make-or-break in this case since it wasn’t even necessary in getting users to the app.

If your site is not central to your strategy, it won’t matter much what your domain is. Otherwise, you should give your domain selection more weight.

How They Get There

This is closely related to the site’s purpose. How are your users getting to the site? If they’re not typing in the URL to get to the site, the domain is obviously less important. If it’s just a landing page for your mobile app, most users probably aren’t getting there at all. If your web site is the app and users access it through their browser, be extra careful in your domain choice. You’ll want something that is short and easy to recall. For those of you who remember social bookmarking site Delicious, for as clever as  their original domain name was (del.icio.us), it wasn’t good when they decided they wanted to attract users who were not early adopter internet nerds like myself. They ended up changing it to “delicious.com” instead. (They still failed, but surely this is more the fault of the Yahoo curse than of the newly more familiar domain name.)

Important But…

Your domain name can be extremely important, but it can also be less important depending on your circumstances. Whatever you choose, don’t let your choice of domain be a showstopper. Pick something and move on. You can always circle back around and make improvements later. If the domain you really want is up for sale, you can always make a bid for it later once you have either revenue or investment dollars to work with.