The 80/20 Principle

Tim Ferriss — "Pricing can be very tricky to figure out. I never guess, and stick with a guess. I’m happy to guess and then test. There are a lot of tools now that you can use for testing (landing pages), and I encourage people to do split-testing when possible. You don’t have to be an amazing engineer and designer to do this, you can use a service like Optimizely, you can use Unbalance.com, you can use something like Hubspot also to do this type of drag-and-drop testing. I would just say, even if it means you can’t do a simultaneous test because it’s too complicated (or whatever reason), that’s fine. Do a test for a couple of days or a week with one price, then swap it out and do another, swap it out and do another, and see which performs best.

Once you arrive at that optimal price (which might not move the most units, this is really important) if you shift to a higher price and sell half as many units you might make more money, but I would emphasize that all things being equal - I would rather have fewer customers making me the same amount of money. It means less headache. Less customer service, fewer returns, and generally speaking: the people who pay more money = less talky, less headache.

I apply the 80/20 principle to everything. The 80/20 principle, in brief, means that 20% of your actions or inputs or products or services will create 80% of what you what, whatever that happens to be. If I have ten products and those products are taking up all of my time - working 100 hour weeks. Let’s say one hundred products that are creating hundred-hour weeks. Chances are if I did an analysis I would find that twenty of those products are producing 80% of my profits, which means, hypothetically, I could cut out 80% of my products - just get rid of them. Pull the trigger and get rid of them. Make 80% of what I am making now, and work 20 hours a week instead of 100 hours a week. I do that for everything. Who are the customers who complain? I’ll spec it out and try to figure it out, what is the profile of the 20% of customers who create 80% or 90% or 100% of my headache and then get rid of them. Figure out a way. I apply that to marketing spend, I apply that to advertising, everything. I do that on a once-weekly, or once every two weeks basis, it’s very, very important.

There are a number of ways to calculate the effectiveness of a given ad or the success of a given product. You can have a one-time sale, or you can have lifetime value, or you can have everything in between. Let’s say I’m going to sell you a candy-bar. Well I can calculate the success of that based on how many candy-bars I sell once, I can also calculate that by how many candy-bars I sell to multiple customers. I encourage people to look at a couple of metrics that are very important: Average Revenue per User (ARPU) on a one-time basis. This will determine your short-term ability to spend money. For instance, if I know that my average user spends $10 on my site and that if I drive 100 people to my site that 10% will buy. I can very quickly figure out how much I can afford to spend on each customer. As long as it’s less than $10 I am okay.

When you get a little fancier, you might have a recurring model. You might say ‘Okay, I’ve got a subscription model. We can afford to give away 3 months for free because we retain 50% of those people, and then spend X amount.‘ Of course, if you are cash-poor like I was when I started my business (I didn’t take any outside financing and I only have two credit cards, maybe five thousand dollar limit), well maybe I can’t wait three months. Maybe I could afford to do a $9.95 trial for a $50/month subscription with automatic renewal. I use something like Recurly, I think that is one of the start-ups out there that does recurring payments. These are all things that you tweak over time. Only test one variable at a time. If you test three variables? Forget it. You can’t keep track of anything.

The time to bring in specialists is when you have tested your product and you are ready to commit. Before you’ve tested the market, the product, the pricing, before you are generating any real revenue, unless you get a lot of PR early (which is an exception), you don’t need to bring in a lot of specialists. As soon as you decide ‘I am going to form a company’, as soon as you decide ‘I need to file for a trademark that I can protect this product (or whatever it might be) with’, that’s when you need professionals.

The first step might be going to a site like LegalZoom and downloading documents. Not many people know this, but the company Automattic which runs Wordpress.com run about 16% of the internet. It’s huge! Most people who are in the blogging world know Wordpress. So Matt Mullenweg (who is a buddy of mine), when he incorporated this he did it with documents of off LegalZoom! I would just say that when it turns into a revenue generating business and you are going to commit more chips or time to it - spend a few hundred bucks to get the company property incorporated, or form an LLC, which in the US is a very flexible format. If you are going to raise a bunch of money and do something fancy then you are going to need a C-corp probably. You do not want to skimp on that kind of professional help in the beginning and wait until you have the tiger by the tail or a bull in a china shop, or your business is like a train-wreck in slow-motion to then bring in help. It will be too late. Get good advice when you need it, but not before.”