How I Used Lean Methodology to Build 99Bitcoins
Now you know how to create and validate an MVP. I wanted to take the opportunity of this final chapter to show you the whole process I went through with 99Bitcoins from start to finish so you can see how I implemented everything we covered up until now.
The idea I chose for 99Bitcoins originally was just an information website about how to buy Bitcoins with PayPal. I found a single exchange that allows you to do so and luckily enough they also had an affiliate program.
Coming up with my best idea through analysis
Of course I measured this idea against other ideas with the "Beigel Analysis" as I explained in previous chapters, and here's how I evaluated it:
Need
I came up with this idea out of my own personal need of buying Bitcoins with PayPal and seeing it was a difficult task. So I assumed (and proved this assumption later on) that this is a big enough issue for all new Bitcoiners. Overall need score I gave this was 4 out of 5.
Viability
Since I was pretty well versed in building websites and internet marketing techniques, setting up an information website was completely viable as far as I was concerned. The viability score was 5 out of 5.
Competition
Searching on Google didn't give me any results for competitors talking about this. However I did find out about this method from a different site that was doing the same thing. So there was little competition which is exactly what I wanted to find out. However I was afraid it may be too little, meaning there's not enough money to be made. I ended up giving the competition rating a 3.5 out of 5.
Profitability
This was by far the biggest enigma. Since I knew I could make money out of this but I didn't know if I can make large amounts of money. And I didn't want to spend my time on a project that will bring me $100/month. Seeing as Bitcoin wasn't still mainstream at the time, but I felt it was pretty close to getting there (since it was featured on TechCrunch) 1 gave profitability a score of 2.5 out of 5.
The total score I got for this project was 4+5+3.5+2.5 = 15. This was the highest score out of all of my other ideas (which included a Bitcoin news site, a local Bitcoin exchange and some other stuff). Like I said before, there is no exact science to this. It was all based on educated guesses and I may have been wrong. For example, Coindesk which was established the same time I started out is now a HUGE website.
Don't be afraid to be wrong about this, you're just at the beginning and it's better to test things out quickly and find out if you were wrong now instead of failing after you've invested a lot of time and money.
Now I had to craft my idea as laser-focused as possible. I could just say it's a "Bitcoin beginner's guide", but I want to be as focused as possible at this point. I ended up with this:
A Bitcoin website for beginners explaining how to buy Bitcoins with PayPal and a credit card.
Pretty focused right?
Now comes my favorite part - the assumptions.
If you remember correctly we have three assumptions we need to make - The problem, the solution and goal, and the business model.
Here were my assumptions:
The problem - People are looking for ways to buy Bitcoins with PayPal
The solution and goal - If I supply people with a written tutorial about how to buy Bitcoins with PayPal at least 2% will buy at my recommended exchange.
The business model - I will become an affiliate of the recommended exchange and generate commissions for each sale.
Let's prove each of these assumptions. We will start out with the problem assumption. In order to see if other people experience this problem as well I used the Google KW tool to see how many people are
searching for this term each month. Here's what I got:
If you remember what we talked about before, over 1,000 searches a month is a good enough indication for me. So I had my first go. But that wasn't enough for me, I then checked Google's autocomplete feature to see if people where actually typing this into google. Here are the results:
In order to prove my solution and goal assumption I had to set up some sort of a tutorial so I could test it and see if people actually followed my recommendations. I also integrated my business assumption into this step, because if people follow my recommendations and become customers of the relevant exchange, I will get my commission.
The obvious way to prove this assumption was to build a basic version of the website and test it. But I like to be super lazy if I can, so I decided only to write a Facebook note inside my Facebook profile. Here's how it looked (you can find the original here):
I also signed up to Virwox as an affiliate and made sure to put my affiliate links inside the post.
It took me about 15 minutes to write this post and I posted it in two relevant Israeli Bitcoin Facebook groups. At the time there were only 800 people total in both groups.
The problematic part about using Facebook notes is that you can't track them. I also couldn't track my posting on the FB groups. But what I could see was how many people saw the post (FB gives you that input as the original poster).
I checked my stats on FB and Virwox after four hours and saw the following:
- 30 people saw my post in total on both FB groups. This doesn't mean they necessarily read it, it only means they saw the link to my FB note.
- I got three conversions on Virwox (giving me a whopping $0.96)
So even though I didn't know my exact conversion rate, I could speculate that worst case scenario all 30 people who saw the post clicked on to Virwox and only three of them converted. That still gives me an amazing 10% conversion rate.
Now I know you're probably thinking about what I said earlier and about how you can't decide anything until you collect at least 300 samples and you're right. But sometimes you won't have the opportunity to collect 300 samples.
Back then I didn't know about all of the other Bitcoin traffic channels like BitcoinTalk and Reddit. I only knew these FB groups so that was my only source of traffic. Having said that, since I had such an off the charts conversion rate (usually it should be around 2%) I decided that even though I didn't collect enough samples, I would move forward with my idea.
So I've proven my solution and goal, and also my business model. But there was still one issue left -1 have shown I could make money out of this idea, but I didn't know if I could make enough money. That's why I needed one last bonus assumption which was:
Bitcoin is going to so big that I will be able to make at least $1000/month from this method.
But in order to prove this assumption I would have to get some more traffic sources and to wait until Bitcoin is big enough. What I did at that point was turn to Search Engine Optimization (SEO). I wanted my site to come up at the top of the search results when someone searched Google for "How to buy Bitcoins with PayPal".
I'm going to talk a lot more about SEO in the next part of this book, but for now I'll just say that because the competition was low I spent $80 on building a very lean and lazy website that got to the 1st page of Google in about four weeks. Here's what I spent the money on:
$10 - Domain name
$50 - Hiring a freelancer to write some content for the site (explained in the next chapter)
$20 - Hiring someone to build some links for me so the site will rise faster on Google (NOT RECOMMENDED!!!! I'll talk more about this later).
Here's a screenshot of my profits from April 2013 until October 2013. The profits are in a currency called SLL (Second Life Lindens) and if you divide them by roughly 252 you get to their USD value. But I already did that and wrote in red how much I earned in USD each month:
So there you have it. This is exactly the process I went through to get started with 99Bitcoins. Since then the profits have grown even further and reach even $10K/month in some cases. Of course today I have several income channels and not only Virwox.
In the next part of this book you'll learn exactly how I grew my website from generating $lK/month to $10K/month, but for now I want you to be crystal clear about how to use lean methodology to test out your own business ideas.
Bonus Chapter: Hiring Amazing Freelancers for Peanuts
Some people avoid starting their own online business since they lack the experience in fields like web design, web programming etc. The beauty of the 21st century is that thanks to the Internet you can now hire someone from the other side of the world to do all of those things for you and you will probably end up paying a fraction of what it would have cost you in the first place.
I use freelancers A LOT. And when I say a lot I mean I barely even deal with actually working on 99Bitcoins directly. Here are some of the things I delegate to my freelancers:
- Design of the site
- Writing news articles
- Promoting the website on social media
- Coding my 99Bitcoins Android app
- Coding my 99Bitcoins WordPress widget
- Fixing technical issues on the site
I almost don't do anything that isn't related to the strategy of the site - planning where the site is going. Once the strategy is laid out, I use freelancers to carry it out. This bonus chapter is here to help you wrap your head around the freelancer markets and give you practical tips on how to hire the best ones so you won't have to work so hard.
Today there are only four major freelancer marketplaces I use. Of course there are many others around the web but I found that through these four I can get most of my jobs done:
Upwork
Probably the site I use the most. Upwork is a great place to find programmers for your website. If you encounter any technical issues, this is the place to start figuring out how to fix them. I tend to use Upwork mainly for programming tasks.
Elance
I consider Elance to be somewhat of a more expensive Upwork. You can find many US contractors here and if you're looking for anything writing-related I'd definitely look here. Since Elance has higher quality freelancers, it's usually more expensive.
Fiverr
This is the flea market of freelancer marketplaces. Here you can find all sorts of tasks to be done for as little as $5. I use Fiverr when I need a quick fix for something, like taking care of a minor bug in my WordPress theme.
I call it the flea-market of freelancer marketplaces since you can find some totally mind-blowing bargains. I actually hired the voiceover for ALL of my instructional videos through Fiverr. Basically I got a two-minute, professional native English voiceover for only $15 - this would have cost me $200 anywhere else!
99Designs/tasks
This is a relatively new marketplace by 99Designs. 99Designs is a website that allows you to hold a contest for your next design idea. Let's say you want to design a website. You upload the wireframes to 99Designs and announce a prize for the winning design. You will then have several dozen designers competing on your contest and designing your idea. All you need to do is pick the winner. The awesome thing about 99Designs is that you get to see different design ideas and the final design before choosing a winner.
But I'm getting off track here. The main feature that I use in 99Designs is called "tasks". This is used for small design job such as fixing a banner design, changing the color of a logo, etc. All jobs are done for $19 and within 1 hour. So if you have a small design job you need to take care of quickly this is the place. You can still use Fiverr for these kinds of tasks but Fiverr usually has low quality designers.
Holding an auction for your next job
I'm going to focus mainly on Upwork and Elance in this chapter since these are the two biggest marketplaces you're going to use. They are also the place where you will probably spend the most money.
When you come to hire a freelancer you are basically holding an auction. You post your job and freelancers bid on your project. You choose the one you think is most suited. But you're not limited only to the people who bid on your auction, you can also invite specific freelancers which you think are qualified.
The problem is that most people avoid freelance marketplaces. Too many times I hear people say that they tried a certain marketplace and didn't get what they expected so they decided to ditch the idea of using freelancers.
The problem isn’t the marketplace
If your job didn't turn out the way you thought it should, you did something wrong along the way. I've been working with freelancers for seven years and have spent over $50,000 on these marketplaces so I have some reference. In this bonus chapter you will learn the exact process I use when outsourcing my work so that you too will be able to get your desired results as quick and as cheap as possible.
In order to achieve the best results we will focus on the following factors:
- Posting a job like a pro.
- Screening your applicants effectively.
- Maintaining the best foreign relationships possible.
Posting like a pro
As always...the title is crucial.
There are so many job posts on Elance and Upwork that it's hard to stand out. Contractors browse through hundreds of job offers each day and you need to catch their attention. Try to make your title as descriptive as possible. So for example the title "Website needed" is much less understandable than, "WordPress website programming needed". In the latter, although not
really detailed, you can still understand that we are talking about WordPress, and about the programming part and not the design. So make sure to fill your title with as much descriptive words for your job posts so that the relevant freelancers will instantly identify that they have the skills for it.
Create a post description that six-year-olds can understand
When you post a job on a freelance marketplace you usually have a vision of what you want to be created, whether it's an article, a website or a design. The problem is that you're forgetting that the contractors on that marketplace don't know what's in your mind. I know it's sound stupid, but so many people don't characterize their projects from A to Z. Here are some examples:
Let's take a look at this job post for example. Imagine that you are a contractor looking at this job post, what questions might you have? Maybe, how long this booklet should be? How should a page look like? Is it in color or not, etc.
Give as much data as possible
If you want to a logo for a website, give the website's URL. If you need an article written supply the bullet points of what the article should be about.
Give live examples
I can't emphasis this enough. If you have examples of things you like and are similar to your job request, use them. So if you are looking for someone to program an app for you and you have a similar app that you like its interface use it as a reference.
Have specific demands
Don't be afraid to request samples of work or for contractors to write down how they are planning to get the job done. One of the sentences I like to use when posting a simple job is, "This should be an easy task for someone who knows what he's doing. Please do not bid if this isn't your area of expertise. " Having demands shows that you mean business and will attract better contractors with better job applications.
See how he states exactly what he's looking for? He spells out how many words each article should be about, what the topics are, and he also gives an example and requests sample work in the job application. After giving the specifics for his job, he then gives a wider view of what would be expected of the contractor. Nice work...
So now that you've got the job description nailed, it's time to tweak your job post. Here are some things that you can do to get better applicants:
- Make sure to request applicants who have done previous work on that marketplace (even one hour is enough).
- Request applicant with a feedback of 4.5 and up (there are more than enough).
- If applicable - ask specific questions about the job. e.g. "How are you going to implement this?", "Where have you written before", etc...
Bonus Chapter: Screening out dozens of freelancers in just five minutes
Now comes the fun part, screening applications. After writing a good job post this is probably the most important step of the process.
On average each programming/design job post will receive 30+ candidates
This is not taken from some Wikipedia statistic, this is taken from my own personal experience. Who has time to go over all of these candidates? There has to be an easy way to screen at least the majority of them...Good news: there is. Throughout the past few years I've created and enhanced a simple methodology that allows me to pin-pick amazing freelancers every time. This is important because it save your money, and moreover it save you time (which is probably your most valuable resource).
Step 1 - A rough cut of everyone who doesn't follow the rules
You already know that you should have some specific demands from your freelancers. For example you might want them to explain in their application how they are going to solve your problem, or give you specific references to similar jobs. The first step would be to screen out everyone who didn't seem to actually read your job posting in full. This means:
- "Copy-Paste" applications that don't have any personal reference to your posting - these are almost 80% of your applications.
- Applications that didn't answer a specific question.
- Applications from new candidates who don't have any feedback or history on the website.
- Applications who are drastically outside your price range
When you screen out an application I prefer "hiding" it than "declining" an application. This is because of the fact that when you decline an application you will need to enter a reason why and then you will probably get that freelancer applying again with a "corrected" application. I don't want this. I gave him his chance and he blew it.
The only exception to this rule would be if you see a really good application which is priced too high and you want to lower it. I would then contact the freelancer and ask if there's a chance to lower the price.
Step 2 - Screen for feedback and response time.
At this point my guess is that you'll be left with up to 10 candidates. Now it's time for some in depth screening. The first action will be to cut off anyone who doesn't have a 4.7 feedback score and up. This may sound harsh but we're looking for a superstar here. The reason I'm not looking for a 5-star only is because I take into account that people make mistakes sometimes.
Also, something seems a bit unnatural in a freelancer that has a 5-star rating in each and every job he completed. That's not to say that he's not good, it's just saying that different people have different standards. So if someone has done 500 jobs on Upwork for example, and has a 5-star rating, I would be suspicious.
After doing this, it's time to check for response time. The idea is to see how attentive the freelancer is to his message inbox. What I do is write a short message to each of the candidates with an additional question. For example:
"Hey [freelancer name],
I wont to thank you for bidding on my project. Before we process I just have one more question.
[insert question here].
Thanks again.
Ofir."
Now I see who responds and how long it takes them. Keep in mind that freelancers come from different time zones so you might be emailing them in the middle of the night. Check out the country of the specific freelancer and make sure he should be awake (I don't care if it's a Sunday, I expect a 24/7 service).
Once you get your replies, read them and this will give you a better idea of who is suitable for the job. Also, screen out all of the freelancers who took too long to answer (I consider over two hours to be too long if it's daytime).
Step 3- In depth reviews
My guess is that you are now left with a handful of freelancers. Most of them should have some bad feedback, that's the interesting part. Go inside each of their profiles and read their reviews. Find out why people didn't rate them 5 stars before. This will give you a better idea of who is the best freelancer for you.
VERY IMPORTANT:
Sometimes projects go "bad" and things get ugly in the reviews. These reviews are extremely helpful in order to assess the freelancer's quality. Read his feedback to these reviews and see if he just gets mad and disrespectful or does he answer in a dignified manner.
You should now be able to choose your best freelancer. If you're still debating between two or three candidates I suggest setting up a Skype interview (10 minutes max with each).
Step 4 - Personal interviews (optional)
When interviewing a freelancer on Skype, the idea is just to see if your prospect knows what he's doing and that he is ready to get the job done. Remember that your prospect's English level may not be that good so he might not feel comfortable talking to you and would prefer chatting. Personally I prefer chatting also since it's less embracing for both sides and that is how you will correspond mainly throughout your work.
The whole interview should take around 10 minutes. I usually take this time to ask the freelancer about his plans for completing this project and also add some more background questions to see that he is actually qualified. If you are still debating between two different prospects after the interview stage it's all about gut feeling from that point on.
Remember, that no matter how thoroughly you will conduct your screening process; there is always a chance to miscalculate and choose someone who isn't right for you.
Bonus Chapter: Getting the Most Out Of Your Freelancer
So by now you probably have chosen your freelancer, and it's time to maintain a good working relationship with him. The problem is that most people ignore this part and don't understand why their contractor under-delivers or doesn't continue working with them.
Your contractor will treat you the same way you treat him...
You need to think of your contractor as your co-worker and not your "slave". Just because you paid someone 5, 50 or even 500 bucks doesn't mean they don't deserve to be treated with respect. If you're rude, annoying or disrespectful it will come back to haunt you. And although you don't think it affects him because he ends each sentence with "yes sir" or "thank you sir", trust me...it does. So here are some guidelines to follow:
Your contractor matches your response rate
The faster you respond to your contractor, the faster he'll respond back. From my experience if it takes you two days to reply to a message he will be just as lazy. So make sure that even the unimportant messages get answered. Even if it's just an "Okay" or "Thanks" - it will calibrate your contractor's response time.
Never pay all of the amount before the job is completed 100%
This is a crucial one. Always leave some leverage on your part so the contractor doesn't slack off. I know sometimes people want to get the payment issue out of the way so they pay early. This is just human nature. Remember, once someone reaches their goal (i.e. money), they won't work as hard as when they're striving for something. So always make sure to keep even a small amount that is unpaid until you're a 100% satisfied with the job.
ALWAYS pay on time
This is more of a "karma" issue, but I believe that paying on time will get freelancers wanting to work with you again. Also, if someday you will actually have an issue and won't be able to pay on the agreed date, the freelancer may still work for you knowing that you're good for your money. Paying on time also helps your employer score by boosting your feedback (more on this in a sec).
Feedback is almost as important as money
Most professional freelancers care about their feedback almost as much as they care about getting paid. Mainly because of the fact that competition on Elance and Upwork is so fierce. This acts as another leverage point for us as employers. Always make sure to leave feedback, but always leave it only after the job is done completely. Just like paying all of the amount in advance, once the freelancer gets what he needs from you, he's more likely to become lazy and less responsive.
You get feedback too you know...
Many freelance marketplaces supply employer feedback also. This is a good reason to start treating your contractors well. If you have poor employer feedback it will be difficult for you to find quality freelancers.
So you need to keep in mind that the following things matter for your client profile:
- How many jobs you've posted
- How many job you've awarded (your "award ratio")
- The amount of money you've spent on the marketplace
- Number of feedbacks you've given
- Feedbacks you've received
