Tracking Magick     Part 5    Testing & Optimization

I like to look at testing and optimization as a game. Tracking is the way you keep score, and the prize for playing well is lots more cash in your pocket.

So now that you're ready to play, let's talk about how to win the game.

There are two things you'll want to focus on:

Optimizing your traffic, and optimizing your marketing funnel.

Let's talk about optimizing your traffic first since it's pretty darn simple ...

If you're tracking things properly, your tracking system should be able to tell you the ROI of just about every source of traffic you have. The paid stuff at least.

From there it's really just a matter of cutting the losers and running with the winners.

I mean, unless you're purposely losing money as part of a customer acquisition strategy, some "branding" campaign, or a marketing test, it's pretty cut and dry:

If a source of traffic doesn't have a positive ROI, you should kill it.

When it comes to pay-per-click ads on Google or Bing, any individual keyword that isn't making you money should be paused or deleted.

Same thing with all of your individual interests and audiences on Facebook.

If you're running banners all over the web on lots of different sites, any individual placement with a negative ROI should be shut down.

Advertising on a blog but it's not generating enough sales to cover the ad? Kill it.

If you're buying solo ads from twenty different sellers you're certainly losing money on some of them, so you need to identify which ones and stop buying from them.

And even if you're not spending money on something, it still has a cost and you still need to track it so you know if it's actually worth doing or not ...

So if you're spending time posting in forums, commenting on blogs, writing "guest posts", or anything else like that as a way to get traffic, make sure you're tracking the results so you know which ones to focus on and which ones are a waste of time.

But in a nutshell, it really is that simple.

About the only thing there is to even think about here is how much money you're willing to lose on any individual traffic source before you shut it down ...

Technically speaking, there is a correct answer here - if you can stomach it.

The "correct" time to shut off a traffic source is when the negative ROI is considered to be statistically valid (and there's nothing you can do to increase it).

And while ClickMagick does all this complicated math for you and tells you when your conversion rates are statistically valid, so you know it's safe to actually make marketing decisions based on the stats you're looking at ...

Most other tracking systems don't, so if you're not using ClickMagick you'll just have to use a bit of common sense and good judgement.

For example you might have one new traffic source with an ROI of 30% after 200 clicks, and another traffic source with an ROI of -30% after 200 clicks ...

The first seems to be "winning" but neither result is statistically significant, and at this point you just don't have enough data to act on.

The traffic source that currently has a -30% ROI after 200 clicks could easily end up with a 50% ROI after 1,000 clicks and be a great long-term source of traffic for you.

There's lots of complicated math involved, but if you're looking at an individual source of traffic I would generally recommend that you don't consider things like conversion rates and ROI to be even remotely accurate until you have at least 500 clicks or so.

Now sometimes it's clear that a traffic source is just never going to work for you, but sometimes you can take a loser and turn it into a winner with a bit of testing.

Testing is really something you should be doing all the time in your business, and it's one of the only real "secrets" to success, so let's talk about testing.

Specifically "split testing." Also referred to as A/B testing, or multivariate testing.

Multivariate testing involves testing multiple changes at once. This is an advanced method of testing, and you don't even want to think about this now.

But regular split testing is very simple, and it can literally double your business.

This is where you simply show two or more different variations of something to different groups of visitors to see which one performs better.

To keep things simple I suggest you stick with basic A/B testing for awhile. This is where you literally just test "Page A" against "Page B" and see which one wins.

Page A is your currently best-performing page, also known as your "control", and Page B is simply the new variation you've created in an attempt to beat your control.

So in terms of optimizing your traffic you'll want to focus on split testing things like your ads, your banners, and your marketing message ...

But you can split test just about anything - headlines, copy, offers, images, colors, the location of elements on a page, button text, and just about anything else.

That's the cool thing about it ...

You can dramatically increase your business just by testing and optimizing a few key things, but at the same time there are an almost endless number of things you can continue to test and tweak to continually increase your profits over time.

As long as you have the right tools it's easy. Some people even find it "addicting."

I really hope you get into it, because I can confidently say it's one of the most profitable things you can do even without knowing a single thing about your business.

Now there are a handful of great standalone apps for split testing, but most people find it's a real pain to use multiple tools just to do this stuff.

I wholeheartedly agree, which is why I made sure that ClickMagick has a full-featured and fully-automated split testing system built right in to it.

With ClickMagick it's super easy to test stuff ...

For example if you wanted to test different PPC ads, all you have to do is use a different tracking link for each ad (or one tracking link with a unique Sub-ID for each ad) and ClickMagick will show you detailed stats for each ad individually.

If you want to test different elements on your own webpages, that's super easy too.

First you just create whatever you want to test. For example if you want to test a new opt-in page, you'd just make a copy of your existing opt-in page - your control - and then change the headline,  call to action, or whatever else you want to test.

Then in ClickMagick you'd simply set up a tracking link for your control page, click a button to start a split test, and then enter the URL of your new test page.

ClickMagick will then split up your incoming traffic accordingly, ensuring that repeat visitors always see the same variation so your stats are as accurate as possible ...

... and then let you know automatically when you have a statistically valid "winner."

If your test page wins you'll want to make this the new control in your marketing, then come up with a new variation for another test and do it all again.

It really is that simple with ClickMagick. Even total newbies can do this stuff.

Now let's talk a bit about optimizing your marketing funnel ...

There are entire books and courses on optimizing your marketing and your funnels, so obviously any type of in-depth training is beyond the scope of this simple guide.

I'm just going to talk about testing and optimizing the few key items on that short list you made earlier - which is fortunately all you need to focus on for now anyway.

We've just talked about optimizing your traffic, so now let's talk about optimizing your funnel in terms of sales and any other "actions" that are on that list of yours.

It's really all just a matter of methodical split testing.

One at a time you just go down your list, pick an item you want to test and optimize, work on it for awhile, then move on to the next item on the list.

Testing is never "done", but you can quickly reach a point of diminishing returns so at least initially you don't want to spend too much time on any one thing.

So work on testing and optimizing your opt-in page for a bit,  but then move on.

You can always come back and try to optimize your opt-in page some more later after you've already grabbed the "low hanging fruit" for each item on your list.

After your opt-in page, work on testing and optimizing any other desirable actions that you're tracking like webinar registrations, inbound calls, or anything else like that.

Optimizing these things should increase your overall sales, but if it's applicable in your business you should of course also split test and optimize your actual sales pages, order forms, your checkout process, and anything else directly related to generating sales.

I can't tell you exactly what you should test, or what to focus on specifically for the best ROI on your time, because it really does depend on your business.

But here's a short list to get you started. These are all things that apply to just about any page you want to improve, and can lead to huge conversion increases ...

1. The headline.

This is what people see first, and it can literally determine if someone spends more than three seconds on your page, so split testing your headlines can seriously increase your conversions - and profits.

Focus on simplicity and explain what your business does and how it can help the visitor in as few words as possible. Using specific numbers or figures and invoking curiosity are two things that can almost always improve a landing page headline.

Above all else, here's the secret to a great headline:

It needs to speak to the visitor and grab their attention. Fast. A good headline will cause just about anyone in your target audience to sit up straight and say to themselves:

"Hey, this is exactly what I want. They're talking to ME. I should keep reading!"

2. The images.

Aside from the headline, the next thing most people see when they quickly scan a webpage are the images "above the fold" (this means what the user can see on their screen right away, without having to scroll down the page).

While you need to test everything for yourself, images of people have been "proven" time and time again to increase conversions in all sorts of niches.

Especially people who seem to be looking at something important on the page, like a call to action - because the website visitor almost automatically follows their gaze.

If testing images of people doesn't work great for you, test product images instead.

There will probably be a very real difference between no product images, images of real people using your product, and images of just the product itself, but there's only one way to find out and profit from the one that'll work best for you.

3. The copy.

Hopefully it goes without saying that improving your sales copy can lead to significant conversion increases. You just have to do it systematically and split test properly, rather than just going with whatever you think sounds the best.

Writing good copy is obviously well beyond the scope of this little guide. I'd highly recommend you work with a good direct response copywriter if you can afford it.

Otherwise, my best advice is to focus on uncovering the specific problems your target market wants to solve, and then positioning your product as the solution.

This usually means explaining your product or service in terms of benefits, rather than just features. So if you're like most "self-taught" copywriters who tend to focus on just features, this is definitely a good place to start tweaking and testing.

Definitely test your sub-heads, testimonials, and all the other important elements too.

4. Button copy and other "call to action" text.

You may have seen someone claim to have increased their conversions 132% simply by changing the color of a button.

You know what they say about if something sounds too good to be true, and in a case like this most likely they didn't do the test properly, or they're flat out lying.

Tweaking button colors can certainly make a difference (Supposedly Facebook tested over 50 shades of blue!), but it's rarely anything to write home about unless the color you're starting with is ridiculous e.g. a light yellow button on a white page.

The main thing with button color is to just make sure it stands out well on the page.

What can make an even bigger and often very meaningful difference though is your button copy, and any other "call to action" text you might have on your page.

Buy Now? Checkout? Purchase? Add To Cart? One will work better for you.

All I can say is test your button copy as much as you test your headlines.

Tell people what to do and why they should do it. Every word counts in a button.

If you can use the word "free" you're almost guaranteed an increase in clicks. Also test calls to action with just text against CTAs with text and images or icons.

5. Social proof.

Social proof can make a huge difference with some businesses and some products, so it makes sense to spend some time split testing this too.

Test a page with a testimonial against a page without one. Test just one or a few testimonials versus a whole bunch of them. Test "better" more specific testimonials against usually weaker more general testimonials.

Try to acquire and test testimonials that address common questions or concerns that your prospects may have. A testimonial from an existing customer that reassures the prospect about something they're concerned with can really boost conversions.

You can also test how they're presented, and where they appear on the page.

6. Reassurance Copy

Sometimes people have concerns, and it's your job to address them while you still can. The way to do this is by using what's generally called "reassurance copy."

This includes things like strong guarantees that remove risk for the customer, letting people know you won't spam them or sell their email when they opt-in, showing people that your checkout process is safe and secure, and things like this.

This type of thing is more important with some businesses than others, and no one on the planet can tell you what will help your specific business, so you just have to test.

7. Video

Some people respond especially well to video, especially in certain niches, so you'll definitely want to do some testing with videos.

Test a page with a video versus the same page with no video. Test a short "explainer" video versus a longer demo video if appropriate. Test an animated video versus an "over the shoulder" screen capture video if it makes sense for your product.

Test videos that start playing automatically versus videos the user has to click on to watch. Test different types of voices for your videos, especially male versus female.

Testing videos is usually more work and can be a pain in the butt which is why I've added it last, but it can have dramatic results and it could be worth testing once you've already tested all the "easy" stuff.

Anyway, I think you'll agree that's more than enough to keep you busy for awhile ...

And "just" optimizing these basic things can have a serious impact on your bottom line, because there's low hanging fruit everywhere and it all adds up quickly.

For example, say you work on optimizing your traffic and increase your ROI 30% there.

Then you split test a bunch of opt-in pages and end up with another 30% boost there.

Next you work on your actual sales copy and see another 30% increase ...

And then you get another 10% bump optimizing your order form and checkout process.

Individually these are actually all very conservative numbers. Most people are able to see even larger increases if they have never done any testing or optimization before.

But even with these conservative numbers you've just doubled your business. It's actually more than that because the conversion bumps compound one another.

It's not going to happen overnight, but I promise you that it's entirely possible.

In fact it's quite realistic if you're willing to put in a bit of effort and do the work.

Compared to most of what it takes to grow a successful business this is probably some of the easiest, most profitable, and most fun work you'll ever do - so I encourage you to jump in with both feet and strive to become a master of the optimization game.

I'll leave you with this one final thought about testing ...

Make sure that you actually do it, and NEVER guess about anything important.

No matter how long you've been marketing, how "smart" you think you are, or how well you think you know your customers or your business, you WILL often be wrong.

And don't ever make changes based on what some other marketer says either. Just because a blue button works for them doesn't mean it'll work for you.

It's great to see what other marketers are doing and what they're testing in order to get ideas for your own testing, but never, ever assume you'll get the same results as anyone else and NEVER make any meaningful changes to anything without testing.

If you do, you won't make as much money as you could be making. End of story.