Building a Customer Vending Machine

By Lewis Bassett, Bassett Providentia Ltd

In the first article of this series, I explained how you could use Google AdWords to build a customer or lead vending machine and buy new customers and leads whenever you need them. I showed you how to estimate the cost of building the machine, and the approximate return on investment you can expect from every dollar you put into the machine. This is determined by the cost of buying a new lead or customer - cost per action (CPA) - and the value of each new lead or customer - value per action (VPA).

Then in the second article, I demonstrated why it was important to always be testing two Google ads against each other: it constantly drives the cost per click (CPC) down, which helps to reduce the cost of buying a new lead or customer.

Another way to reduce the cost of a new lead or customer is to reduce the number of clicks that are required to get one signup or sale. And you do that by increasing the conversion rate of the landing page.

Conversion Rate
When a searcher clicks your ad, they're taken to your landing page. If the landing page does a good job of persuading them, they will buy something or complete a form for more information. This is called an action.

Not every person who clicks on your ad will complete an action. And that's entirely reasonable; your offer won't appeal to everyone. But a certain percentage will. We call this percentage the conversion rate. If 20% of the people who click on your ad complete the form on the landing page, the conversion rate is 20%.

Cost per Action
You can work out the cost per action by dividing the cost per click by the conversion rate of the landing page.

For example, let's say you have a landing page that collects leads for a whitepaper. If it costs you $1 per click and the conversion rate of your landing page is 5%, then it costs you $20 to get a new lead. That means the cost per action is $20.

You can set up your Google AdWords account to calculate and display the cost per action for you automatically. I'll come back to this later.

Crafting Your Landing Page
There are loads of different ways you can craft your landing page. And there are many different techniques for persuading people who click on your ads to take an action.

The most important thing - by far - is that you have a solid framework for measuring conversions and testing changes that you make to your landing page. After all, every question about how to craft your landing page comes down to the cost per action. Whichever method gives you the lowest cost per action is the one you should use, and you find that out through testing.

Content Experiments
Google Analytics has an excellent feature called Content Experiments, which allows you to run two versions of the same landing page side-by-side. It tracks the conversion rates for both landing pages and reports the results of the test once a statistically significant winner has been found. And it does all this automatically.

To use Content Experiments, you need three separate web pages.
- First, you need the original landing page. We call this the control.
- Second, you need a modified version of the original landing page. We call this the variant.
- Third, you need a conversion page. This is a page that is displayed to the visitor after they've performed the action. Regardless of whether that action is a sale or a signup, the visitor should be taken to a "thank you" page that is only displayed after an action has been completed.

Google Analytics gives you some scripts to insert into each page. When traffic hits the landing page, half of it remains on the control and half is redirected to the variant. And when visitors from either version perform the action, the script on the conversion page records that the action took place. So Google Analytics keeps track of how many visitors from each different landing page completed an the action.

After the experiment has been setup, you don't have to do anything. Google will keep track of the recorded actions and alert you when there is a winner.

If the variant has a higher conversion rate than the control, make the variant page the new control. Then create a new variant to test something new. And if the control has the higher conversion rate, make a new variant.

Just like with Google ads, make it a golden rule to always be testing two different landing pages. That way, your conversion rate will keep improving. This means the cost per action will keep going down, and you will be able to buy customers or leads for less.

The Most Important Elements of a Landing Page
Usually, when a visitor arrives on your landing page, they'll look at the headline. They need to see keywords and a message that matches the search they made and the Google ad they clicked on. This lets them know you're continuing the conversation you started on your ad.

If the benefit (or implied benefit) of the headline is strong enough, their eyes will move to the call to action. This is usually a button or a short form. And if the call to action is strong and congruent with the benefit of the headline, they'll complete the action that you're asking them to do.

So these two things - the headline and the call to action - are usually the most important part of a landing page. When I'm crafting a landing page, I put most of my effort into finding a headline and call to action that work together, and are strong enough to inspire the desired action on their own. I fill the rest of the page with "lorem ipsum" text. I only enter the real text after I've found a strong headline and a call to action that work well together.

These are the two things that you should concentrate most of your tests on. Experiment with different headlines and calls to action that communicate different benefits and offers.

The more traffic you can send to your landing page, the quicker you'll be able to complete these tests and improve your conversion rates. So it is very important that you are willing to spend money buying as many clicks as possible at this stage.

By increasing the click through rates (CTR) of your ads, you can purchase more (high quality) traffic for less. So a high CTR really helps with maintaining a good testing velocity.

When you first begin building your machine, the click through rates of your ads are going to be low, which means that less traffic will be hitting your landing pages and tests will take longer to complete. As you make progress, your testing velocity will increase. You can also increase testing velocity by spending more.

Cost per Action vs. Click Through Rate
Earlier, I mentioned how you can use AdWords's Conversion Tracking feature to keep track of actions that are performed when searchers reach the landing page. Google AdWords provides you with another script, which you put on the conversion page (keep the Content Experiments script in place too).

AdWords will count the number of actions that take place for each of the two ads that you're testing, and will automatically calculate the cost per action for each Google ad. This information is gold.

Different Google ads attract different types of people to your landing page. And these different types of people will convert differently, because their mindsets are different.

As your landing page becomes more efficient, you will find that the Google ad that brings the traffic often has more effect on the conversion rate that the landing page itself. This is why it is so important to measure the cost per action for each ad, as well as the click-through rate. The better ad is the one that provides customers or leads for the lowest cost - i.e., the one with the lowest cost per action. That might not necessarily be the one with the highest CTR.

Compound Testing
If you are always testing two versions of your Google ad and two versions of your landing page, the amount it costs you to buy a new customer or lead will keep going down. (This is because the cost per click is always going down, and the conversion rate of the landing page is always going up. And these improvements multiply together.)

After a few months, you should be able to buy a new lead or customer for significantly less than you make from them. When this happens, you effectively have a customer vending machine. You can buy new leads and customers whenever you need them, and at a fairly stable price.

Eventually though, your progress will slow and it will take you increasingly longer to improve the cost per action. So what happens if progress slows right down, before you're breaking even? Does this mean the market is not viable? Not necessarily.

Value per Action
This article has been concerned with only one side of the marketing coin: buying customers as cheaply as possible.
The other side of the coin is increasing the value of each customer, by getting the average customer to spend more with you over their lifetime as a customer. This means you make more profit from the average customer, which means you can spend more to buy a new customer or lead. (When the average customer is worth more to you, the average lead is too.)

You can increase the value of your average customer to you by building deeper relationships with them, and providing them with more value. In my opinion, this is the more powerful side of marketing. But it's also the one that gets the least attention. But increasing customer value is a topic for another day.

Building Your Customer Vending Machine
These principles - measuring cost per action and using testing to drive that cost down - apply to any method of generating new leads and customers. This includes direct mail, social media and print advertising.

This method of acquiring new business is both sustainable and scalable. You're limited only by the volume of your market.
If you need more customers, you just increase the amount of money you put into the machine. And when you want less, you just ease back. And if you're making more money from each customer than it costs you to acquire them, you've effectively got a license to print cash. You can spend $1 in marketing and get $2 back in profit.

For part 1, click here.

For part 2, click here

About the Author: Lewis Bassett is an online marketing consultant and speaker, and helps companies to increase their revenue. Bassett Providentia Ltd is his consulting practice.