This article is part of a series about going from “I’d like to start a business” to having a small start-up running, and beyond. If you haven’t read the previous articles, I recommend you do so by going to the first one, which includes a small table of contents.
You’ve got your idea. You tell a few friends about it and they confirm what you were thinking: if you execute it, you will clearly become a millionaire. You lock yourself in a room and spend the next six months developing the product and everything related to turning your idea into a successful business (or just as tragic, you spend a few thousand dollars in someone to develop it for you). After six months, launch-day arrives. A day goes by and you don’t make a single sale. A week goes by, still not a single sale. After a month without a freakin’ single customer you realize what just happened: you worked for half a year on a product no one cares to buy. Good news: this post is about how to test your idea without even having a working prototype.
Even though the title of this post is “Validating your Idea”, it is actually not an idea what we wish to validate. We won’t go to people saying “hey, I got this awesome idea: selling t-shirts online.. what do you think about it?”. What we really want to confirm is our value proposition: the reason for which somebody would want to buy whatever it is we’re selling. We want to find out if there’s anybody out there who, just like us, sees something attractive in our product and are willing, at least, to use it. That’s what we mean with validation. In our example the value proposition might be “Buy basic, fine and quality T-Shirts without moving from your home”.
How to do it
Asking our friends what they think about it does NOT count as validation. We will look into two ways of validating our value proposition: first in a quantitative way, then a qualitative one.
Let’s get back to our t-shirts. It’s clear to us, especially having laid down our value proposition, that we will be selling them online. Now we want to find out if people might be interested in buying our product. The problem is, we don’t have a product yet but in our imagination, and we don’t want to waste time building a store1 (no matter how simple it is), nor contacting suppliers and doing other stuff, without first finding out if it has a minimum hope of attracting any customer. What’s the closest thing we can do to offering our t-shirts online and seeing how many customers we can get to pay us money? Let them pay us with their emails! This is going to be our hypothesis: if any user that arrives at our site gets interested in what we have to offer, they will at least be willing to give us their emails (in exchange for being the first to be notified we launch the real store). What we will explore now is how to carry an experiment to test this hypothesis, which will help us answer some really important questions. Among others: How many emails can we get? At what cost per email?
The Magic Combo: Unbounce + AdWords
In order to carry this experiment we will need two things:
- a website where we can introduce our value proposition and let users fill in their email address and send it to us.
- a way of buying traffic to that website
The former is called Landing Page: it’s where visitors coming from the traffic source we’re paying for land for the first time. Don’t worry if you’re not an expert web designer, there are various websites that will let you build a Landing Page without writing a single line of code. My favourite one, and the one I recommend, is Unbounce. Our site should have a title (our brand’s or product’s name), a short description (the value proposition) and what’s known as CTA, or Call To Action. The CTA is the action we want our users to do: in this case, we’re going to place an email field and a big and juicy button labeled “Submit”. A picture, if we have one, can help. It doesn’t matter that we don’t have a finished product yet: if we plan to sell an existing product, we sure can get a picture and use it (as in our t-shirt’s example); if we’re talking about something totally new, there’s always a way to illustrate it. Between the value proposition and the CTA, we should clarify that we’re getting ready to launch our venture, and that by submitting their emails users will be among the first ones to get notified when we’re ready. It’s very important to talk about the product as if it already existed to give our proposition some credibility. In order to do that, it’s also a good idea to register a .com domain right at this stage. Anyways, it’s not this article’s goal to dive deep into how to build a great Landing Page, so let’s go on2.
With the first step covered, it’s time to get into the “buying traffic” part. Although there are several places to look at, for the time being we will focus just on Google AdWords. What we want to do is relatively simple: every time someone searches for “t-shirt” or any query containing that word, we want our site to come out as one of the top results. AdWords allows us to do that. We just need to create an account, design an ad (which will look just like the rest of the results), choose which keywords in a query will make our ad appear on top of the results, and set a daily budget. The cost is set on a per-click basis, so we are literally buying clicks. As with the Landing Page, it’s not the goal of this post to explain how to create a great AdWords campaign3.
The next step is to activate the campaign. As a consequence, the following flow will take place: a lot of people will go to google.com and will search “t-shirts” or some string of text containing that keyword. Our ad will come out on top, a percentage of the people searching will click on it and will end on our Landing Page. After reading its title and our value proposition, they will decide if they want to give us their email address. Every email we get counts as an actual potential customer. With this system up and running, we can go ahead and to answer the questions we are concerned with.
Following the previous logic, we will consider the percentage of people who submitted their email address as the percentage of people interested in our value proposition. Before activating the campaign we should establish a percentage that will define if our value proposition is, at least at this stage, validated or not. As a general rule, any number above 10% constitutes a really good result4.
Getting a first estimation of our Cost per Acquisition (CPA)
This percentage, or conversion rate, is not the only thing that matters to us. Let’s imagine we get an astronomically high conversion rate… of 50%. That would mean that out of every two persons visiting our site, one wants to be notified as soon as our product is up for sale to buy it. It sounds too good, but if we are paying $100/click, then acquiring each potential customer is costing us, on average, $200. In order to make up for that amount of money, we would need really big margins. This is an extreme example, but it shows crystal clear that the CPA is a variable we can’t afford to ignore; it is, in fact, one of the most critical metric we will get out of the experiment when it comes to understanding the feasibility of our project. And we will come back to it in the next article of the series.
Comparing two or more ideas
These metrics I just described are not only useful to analyze an idea in and on itself, but also to compare several ideas. If we are trying to decide which of the various ideas we wish to pursue, these are two numbers we can’t overlook in the process of reaching that decision.
Getting out of the Building
Up until now we have covered a fundamentally quantitative method. In spite of its virtues, sometimes it might be insufficient or even unworkable. In order to understand why, let’s return to the example of Cena Plus: restaurant owners don’t spend their days googling “web based reservations system”. Specially before such a thing even exists! More often than not, we have no choice but to talk to potential customers. And that’s a good thing, because those conversations always give birth to ideas and knowledge we could never get from mere statistics.
When we were building Cena Plus, we didn’t even know that a concept like the one I just described existed. And having very limited funds, we had no intentions of paying someone to build the site without knowing if restaurant owners would be willing to use it. The first thing we did was to go to restaurants one by one asking to speak with the owner or whoever was in charge and conducting a short survey (of course, after describing the product). These surveys were just a bit useful, but we were not getting any commitment from restauranteurs. And lacking a product or concrete proposition, we were not taken too seriously. Meanwhile, we remembered how somebody (I can’t recall who it was) described the best market study: I go to a potential customer, I show him my product, I ask money in exchange for it. He says “no”, I go back to the drawing table to iterate.Bearing in mind this definition, we changed how we faced restaurant owners. We had already designed our wireframes (sketches), so we went once more restaurant by restaurant with little folders containing 3 things: a promotional flyer, the printed wireframes (yeah, that’s right) and a “pre-agreement” stating that the signing parts would work together. While we were telling the owners about our proposition, we would put the printed wireframes on the table and guide them through the different screens so that they could get a good idea of what we were talking about. When the “tour” was finished, we would explain that launch-day was coming soon and if they signed the pre-agreement we would give them the service for free for a set period of time. The pre-agreement was not really a legal document, put it was a way for us to get some commitment from restauranteurs and to become more confident about the idea of someone actually using our site (or else, no one would have bothered to sign anything in the first place!). When we signed the 20th pre-agreement we just took our wallets out, commited 200% with the project and the real development began.
Besides confirming there was a market for our product, those interviews taught us a lot about how restaurants handled their reservations, which aspects of our proposition they liked the most and which the least; and we even ended up doing several changes to our designs based on what we learnt. As Steve Blank says, “getting out of the building” and talking face to face with potential customers is one of the most important and enlightening things an entrepreneur can and should do in the beginnings of his adventure.
Conclusion and Next Steps
With these tools you can already begin investigation whether there’s people out there interested in your value proposition and learn a bit about them. Now, let’s get to work! I hope you will find the courage to pick those ideas you have been thinking of lately and test them. If you have any doubt, I will be glad to answer it in the comments. Otherwise, see you in the next post, where we we’ll see if the idea has potential to become a profitable business (people might like our value proposition, but if the financials don’t work…).
- Or coding an app or designing what we have in mind. ↩
- To cover that, I suggest you do some research on your own, the web is full of wonderful resources. If what you need is some inspiration, in this link there are 35 examples of good designs, each one of them with comments on what was done right and what could be done better. ↩
- I recommend reading the Google guide to this product before you activate the campaign. If you’ve never used AdWords before, it’s pretty easy to get a free ads coupon (generally worth between $50 and $100). You can get one of those when you register a .com domain with godaddy, for example. ↩
- This percentage is known as Conversion Rate. We should note that to start drawing some conclusions, we should get a relatively big data set. Personally, I feel comfortable with a minimum of 100 clicks (and I become even more comfortable with more than 300). This will also define our budget. ↩
- The general formula to get the CPA is= Cost Per Click (CPC) / Conversion Rate. In our example, $100 / 50% = $200 ↩