How I learned to kill bad startup ideas

Mona Akmal
Jun 3, 2019 · 11 min read
Image for post
Image for post

6 months ago I finally decided to start my own company. I knew the team and culture I wanted to build, and the value I wanted to create.

I just needed to figure out The idea.

Many founders I spoke with shared that their idea came from existing domain expertise, or from deep passion. I couldn’t relate since my experiences & interests are varied. Through trial and error, I’ve learned that it takes at most 7 days to weed out obviously bad ideas, and years to craft a great one. This is about those 7 days that can save months if not years of wasted effort.

Day 0: Create the idea backlog

Start a spreadsheet and write down problems you face. Focus on your personal and professional life, as well as those you’ve heard from friends, family, customers and partners. This is difficult because we’re so used to living with pervasive problems that we can’t see them. Power through it. Also, there’s no problem that’s too small or too large.

Some examples:

  • No Siri, I didn’t want to say “what’s your ducking problem”
  • I want a world without usernames, passwords and other forms of ID
  • I wish meetings were shorter, with more follow-through and decisions

Step 2: Sort this list

We use 2 columns at this stage — level of excitement, level of expertise. Each is scored 1 to 3 (3 is most positive). Level 3 excitement means you’re excited to spend 5 years on it. Level 3 expertise means you’re an expert and have the right network to tackle this problem. This doesn’t mean you can’t step out of your comfort zone, but consider this: Building a successful company is a low probability scenario, why lower your odds even more by not playing to your strengths?

Cull problems that have an average score of 1 and most likely 2.

Day 1: Frame the idea

Image for post
Image for post

There’s a strong urge to start thinking of solutions or doing research on what’s out there. This is wasteful right now. Instead take an hour to frame each idea.

Step 1: Create a back-of-the-envelope pitch deck per idea

The purpose of this is to get clear in your thinking and make smart assumptions. Don’t use more than 150 words per slide.

The pitch deck is simply:

  • Customer problem (This also covers who the customer is)
  • Our solution (No need for minutiae)
  • Market size (Fermi approximations are a great frame for making good guesses. A good book to sharpen this skill)
  • Why is it better than existing solutions? (Faster? Cheaper? More reliable?)
  • Business model (who’s paying, how much and how often?)
  • Go-To-Market plan: Consider doing this as 3 phases. If it’s a consumer idea, Phase 1 should get you your first 100 customers, Phase 2 should be 10000, Phase 3 should be 1000000 (i.e. 100x growth). If it’s an enterprise idea, Phase 1 should get you your 1st customer, Phase 2 is 10 customers, Phase 3 is 100 customers (i.e. 10x growth)
  • Future growth (If you succeed at solving this problem, what adjacent problems are you positioned to solve?)

Step 2: Write down your assumptions

Go slide by slide and write down the assumptions you’re making. Examples of assumptions:

  • The customer problem is real for a sizable subset of the population
  • People will actually pay $x for your product
  • Your solution is actually 10x better than existing alternatives
  • Customer acquisition costs less than customer lifetime value

Day 2: Cull ideas with small markets

Image for post
Image for post

From here on out, ideas get culled each day and that’s great.

First and foremost, is the market big enough? Figuring this out requires clever guessing, since you are pre-product and pre-customers. We use a spreadsheet with one column for each step below.

Step 1: Figure out the size of your customer base

Google search, digging through census data, department of labor statistics, Pew Research, Statistica and others is helpful and fast.

Also do comparable analysis. Consider well-established adjacent products your customers use and research the user base of those products to approximate your own. Example: If you’re making backpacks for young urban women in the US, Ulta Beauty customer numbers are a good approximation since it’s a similar customer base.

Step 2: Guess their expected frequency of use

Example: Coffee is consumed multiple times a day, whereas wills are usually written once in a lifetime. This is important because customer acquisition strategy, customer lifetime value, churn prediction and pricing model are all impacted by frequency of use. Obvious example: Charging a monthly subscription service for a product that’s used a few times a year is probably not a great idea. Although gym memberships suggest otherwise ;-)

Step 3: Guess the level of pain experienced without the solution

Example: Not getting a ride to the ER when needed is highly painful, not having Coke zero when you want it is less painful. Most of us are not building companies that solve ER level pain points and that’s perfectly fine. But it’s important to be honest about it and price/position appropriately.

Step 4: Figure out customer’s willingness to pay

This is partly a function of the value customers perceive, but also depends on existing user expectations. In a world where social networking is “free”, can an upstart enter with a paid model and gain adoption? So do some research to understand pre-defined models for your type of product. Saying “we’ll make money through advertising” doesn’t work. In order to be a credible destination for online advertising, you have to get to massive scale. For first time founders that’s a hard story to sell.

Step 5: Cull

Ideally your idea serves a large customer base who will use your product frequently, will be in extreme pain without it, and have an existing expectation to pay for it.

Rule out anything that looks small and free. That eliminates some amazing ideas as evidenced by breakout successes like TikTok, but I don’t know how reproducible that is with this method. For those types of success stories, a wander in the woods is required with funds to keep trying until something truly entertaining is discovered. Unfortunately, most of us don’t have the resources for that.

Day 3: Cull ideas based on weak business model

Step 1: Customer acquisition brainstorm:

You made many assumptions about your Go-To-Market phases on Day 0. Now you get to figure out how you plan to get these customers and what that will cost. Have a freeform session about where your customers hang out.

  • Are they taking their dog for a walk to the park?
  • Do they follow certain influencers?
  • Do they use an existing product that’s complementary to yours?
  • Are they in specific Facebook Groups?

These are your potential channels to reach your customers. Be creative. Rank these channels based on the degree to which they are crowded, expensive and scalable. For example, recruiting dog owners at the dog park as Rover customers is a great Phase 1 channel (i.e. gets Rover to 100 customers), and doing a partnership with Petco is a great Phase 3 channel (scalable). Generally, for Phase 1, opt for channels that are less expensive, less scalable and less noisy. For Phase 3, you need channels that scale very effectively (so partnerships, paid marketing are strong candidates).

Step 2: Make a business model

Pick any template (many available online) and make sure you cover fixed costs and how you expect them to change over time, customer growth ramp, customer acquisition cost (based on your go-to-market channels for Phase 1/2/3), pricing (including attach rates for premium options), R&D and how you expect it to change over time, churn rate (a quick Google search on typical churn rate for your type of experience can help) and returns/refunds.

Create 3 versions of the model — conservative, reasonable, optimistic.

In the conservative model (note: doesn’t mean doomsday), set your customer acquisition cost on the higher end, your customer lifetime value on the lower end and customer growth on a slower ramp. Optimistic plan is the opposite, and so on.

If an idea is only viable in the optimistic path, it’s less interesting. If the idea is viable in the conservative, that’s great.

Day 4 & 5: Cull ideas if the problem is not real

Image for post
Image for post

Not all problems are real. They may already be solved, or customers don’t perceive it to be important (i.e. it’s a latent problem). Convincing a customer of a problem they don’t perceive is impossible. So focus on problems that are already palpable, and throw out everything else. Here’s how:

Step 1: Do Google AdWord research

Use the Google AdWord planner and find keywords that describe the problem or solution you’re researching. Look at the number of searches per month as well as the bid cost for each keyword. What do these numbers mean?

If the number of times people are searching for these keywords is low, you know people won’t find your product by searching for it. So at a minimum, this may not be a good channel right away. Interpreted more broadly, it may also imply lack of demand or an undeveloped market.

If the bid price for your keywords is really high, you know the competition is fierce (which is a validation that the problem and demand is real) and that your customer acquisition in this channel will be expensive (so it only makes sense if your customers are willing to pay a lot more for your product than you are willing to pay to acquire them).

Step 2: Google search

Research existing solutions by going down Google rabbit holes. Be warned — there’s no problem that hasn’t been “solved” so don’t be disheartened by what you find. Be curious. Dig deeper. What are the customer reviews for solutions you find? What is the level of adoption? Can you immediately see how your solution is 10x better?

If you’re exploring enterprise solutions, be wary of what the websites say because all enterprise products say they solve every problem. Gartner/Forrester can be a better resource (though they’re not unbiased). Speaking with people that use the market leader products is often the best way to do enterprise research.

Step 3: Talk to 5 customers

Identify people (that are not your close friends) who can be your customers. Ask each one the same questions based on a structured interview plan. Don’t pitch your solution. Focus on vetting the problem and its severity and frequency, how customers workaround it today, and their willingness to pay to solve the problem. Also invite customers to share other problems that are adjacent to yours. Tabulate the data so you can clearly see what you’re hearing.

Step 3: Do a short ad campaign (Fake door testing)

Pretend that your solution already exists. Create a nice landing page describing it, how it will make the customer’s life better etc. This is a really great exercise to succinctly describe the value of your solution in simple terms. Put in a simple email capture form (this can be phrased as “Learn more”, “Join the Beta” etc.). We use Instapage for landing pages (because it’s really easy and mobile friendly).

Create an ad on a channel that’s appropriate for your customers and product — common players: Facebook, Google search, Instagram or LinkedIn, USPS. The ad should direct visitors to your landing page where they learn more and provide you their email address.

Run the ad with a tiny budget ($20) and measure the cost per ad click, clickthrough rate for the ad, how many people end up on the landing page, and how many of those people give you their email address. Every channel has its own average clickthrough rates (you can Google search for those). If your clickthrough rate is 2.5x or higher than average, that’s great. At the end of the campaign, count how many email addresses you got and divide your ad spend budget by that. This is your initial customer acquisition cost. For enterprise products, I suggest running the ad to download a “white paper” on a topic. Anyone who gives you their email address for it is telling you they’re interested in solving the problem, which is a sign that they’re a good lead. Customer acquisition cost calculation in enterprise cases is bogus so not worth calculating.

Step 4: The big purge

At the end of these 2 days, there should be a serious culling of ideas due to weak demand (validated through customer conversations and rudimentary data collection).

Day 6: Figure out technical feasibility

Image for post
Image for post

Feasibility is a spectrum. As a first-time founder, I’ve chosen to stay on the more feasible side of things. I believe serial founders with more capital can peg themselves on the less feasible part of the spectrum because additional capital can help tremendously.

Roadmap the product. Give yourself a constraint that you can’t disappear for 6 months to “build” your product. Screw the Minimum Viable Product. What’s the non-viable first version of your product that lets you get it out there with 1 customer? What skills are needed to build it? What will you learn after you put it out there?

My roadmap preference is to front load elements of the product that test real demand and final output of the product. The process for getting there can be faked. Example: If my product is a candy bar, can I create a version of it in my kitchen just to see if my customer likes the candy bar (even though it took me 10 hours to make 1 and it cost $100).

What’s the next version after that and what am I hoping to learn from it? How many customers would I like to serve with this version?

And so on. Do this until you believe you’ve achieved your “scale” solution.

For each product milestone, write out the skills and cost of build (including people, materials, logistics etc.). Also identify fixed costs with each phase of the build.

It can be a humbling experience when you see the funds you need to raise for the solution you want to build.

Finally, if you’re betting on key technologies to work as you think they do, it’s worth a 1-hour search to figure out if you’re blatantly wrong (e.g. your product assumes that there are 3rd party solutions that do great sentiment analysis on text — there aren’t).

Day 7: Rest & reflect

Image for post
Image for post

Borrowing from divine wisdom here (even though our day count started at 0).

Your mind needs a break. Instead of asking yourself deep questions, go for a walk or a bike ride. Your unconscious mind has lots of information to chew on, many dots to connect. Let that happen.

On the night of the 7th day, look at the final short list of ideas. Do any of them speak to you? If yes, close all doors and jump in. No more idea wanderlust.

If not, REJECT ALL OF THEM. Start another sprint.

There are thousands of amazing companies yet to be built. There’s an abundance of opportunity.

Ultimately, the decision to build a company is irrational. And even though rigor in the vetting process is essential to figure out what NOT to do, it the decision to dedicate the next 5 years of your life to an idea must come from the heart.

The Startup

Medium's largest active publication, followed by +732K people. Follow to join our community.

Medium is an open platform where 170 million readers come to find insightful and dynamic thinking. Here, expert and undiscovered voices alike dive into the heart of any topic and bring new ideas to the surface. Learn more

Follow the writers, publications, and topics that matter to you, and you’ll see them on your homepage and in your inbox. Explore

If you have a story to tell, knowledge to share, or a perspective to offer — welcome home. It’s easy and free to post your thinking on any topic. Write on Medium

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store