It’s Lewis a.k.a. Nerd #2.
If you’re clueless about how to give your marketing more muscle with testing, you’ll be dangerous to your competitors by the time you finish reading this.
And if you’re well-versed, I imagine even you will find one killer idea to transfer to your practices in what you’ll see below. Yes, the information is THAT good.
Now let’s get this show on the road.
The fastest way to get more money in the door is to turn more of your visitors to your site into customers. And the only way to jack this number up is through . . .
Most business owners either know nothing about testing or what they know is wrong.
Most people are making changes to their site blindly and then guessing whether or not it had a good impact because they don’t know how to track.
A/B Split Testing
If you don’t know shit about testing, this is the place to start.
This is where you split your traffic 50/50 to two pages that have one difference on each of them that you’re testing and once you’ve pushed enough traffic through you end up with a valid result.
Most people screw up when they first learn about this by testing multiple changes at the same time. The only valid results come from testing one change at a time.
Software like Google Website Optimizer allows you test multiple combinations of changes at once.
So instead of only testing one change at a time, you can test 3 headlines, 3 different pictures, 3 different buy buttons which would be 27 different A/B split tests of a piece.
If you aren’t getting a shit-ton of orders, this is going to take you forever to come to a valid conclusion about which combination wins. He uses the term “Confidence Score” here which means you’ve had enough people go through the buying process that the numbers you’ve racked up is high enough in volume that you can believe it.
Adaptive Multi-Variant Testing
At the time he was doing this presentation this technology was so new that he didn’t trust it yet.
What this is supposed to do is quickly spot your promos that are horrible, kicks them out of the test, and re-diverts the traffic to other sites which reduces the time till you know the winning combination.
Derek Gehl’s Favorite Testing Tools
Beginner on a Budget – Google Analytics & Google Website Optimizer
For your stats Google Analytics can do pretty much everything a paid service can. For split testing he recommends Google Website Optimizer.
The things he likes about these is that they’re free, they’re simple to use and they have a ton of support you can access if you’re clueless.
The things he doesn’t like is that you can’t segment your visitors, there’s no heat mapping, and no paid support that allows you to solve problem with one call vs. scrounging through three forums digging for the solution.
Upgraded Services – Adobe Sitecatalyst (used to be omniture), www.vertster.com, and clicktale.com
Adobe isn’t super cheap because it’s a high end analytics tool but if you’re doing some heavy weight volume in traffic in sales it’s gonna pay off because you get paid support, you get the constant improving of it.
Vertster is what you’d use for your split testing. This software allows you to segment your lists. And it also measures results beyond conversion.
Clicktale is a great tool for heat mapping your site.
These services charge you according to how much traffic you’re ramming through them. They’re also kind of technical. But if you aren’t taking your numbers serious enough to track and eventually pay for, you’re leaving money on the table.
What Do You Need To Be Measuring?
1. Your Sales Funnel
Every action that you want people to take should be measured for it’s effectiveness in getting the result you want and constantly improved upon.
You need to know where your site is bleeding – which step is gaping open letting leads and money flow away from you. The numbers when measured tell you exactly where this is.
Metrics You Want To Testing
All google website optimizer measures is conversion. But that’s not good enough because your conversion can improve but your cost of sale can also rise with conversion because you’re buying more expensive traffic. Conversion stats don’t paint the broadest picture.
Gross Revenue Per Visitor
You want to know how much people are buying
Average Order Value
You can increase conversion but lower your average order value. If you only see conversion though you never see this stat.
Net Revenue Per Visitor (Minus Cost Of Goods Sold: click cost, commissions, product cost)
This is the number that shows you for real how much money you’re shoving into your pocket. This is ultra important to know.
Most Significant: Projected Net Revenue (version X vs. version Y for Z period)
This almost knowing which one is gonna make you the most money.
You have to see BEYOND conversion. Google Website Optimizer is handicapping so many businesses because they don’t give you these stats. It’s easy to make shitty decisions without having all the right data.
One Metric Alone Is Deceiving
Scenarios that can and do play out in your business . . .
Your conversion can rise but your average order value can plunge and this turns the increase into a loss. If you’ve only got google website optimizer you’re congratulating yourself not knowing the damage you’re doing.
Your conversion can dip but your gross revenue per visitor can rise which leads to an increase in sales. With Google Website Optimizer in place you’ll see this dip and think you need to go back to the control and bail on a site scheme that’s making you great money.
Your gross revenue per visitor can increase but your net revenue per visitor tanks because the test leads to higher pay per click traffic in order to convert and this leads to a lowered conversion of your low cost organic traffic. This is why you want a service that allows you to segment traffic streams. Every source of traffic costs a different amount and has a different conversion rate. The most profitable traffic is free but depending on only one source of traffic is dumb. So to shore up the business we go out and wrangle affiliate traffic, PPC traffic, banner traffic but when we send it all to the same landing page and track it all lump sum, this doesn’t give you valid stats.
Segmenting & Targeting Traffic
As of this recording Google Website Optimizer didn’t allow you to segment your traffic.
Something that you always want to remember is that all traffic is not created equally and different sources responds differently. For example, affiliate traffic responds differently than organic search traffic to the same exact landing page.
You need to measure each source of your traffic and do whatever you can to make sure it’s profitable. The big mistake information marketers make is being undisciplined as a result of being in such a high margin business where the beer flows like wine. So instead of looking at how high is high, they look at broad stats and cash in the bank and never question their approach.
Different sources of traffic should have different sets of funnels they run through. Vertster allows you to segment down all the major sources all the way down to individual referrers so you can see how each source is converting.
What Do You Test?
First you have to determine what your sales funnel is. You measure it. Then you look for what piece is broken.
Also make sure you aren’t ever testing 100% of your traffic on a change. At minimum, you want to do an A/B split.
If your bounce rates on your landing pages are high, this is a blatant signal that this page needs to be tested against another version.
Heat mapping and click tracking tell how people cruise your site. His favorite tool that does this is Clicktale. This is software that tracks where people’s mouse goes to on your page. Researchers have found that where people’s eyes go, their mouse goes which ends up making this software about as 85% as effective as actual eye tracking.
What you might find when you use Clicktale is that you’re drawing eyeballs to stuff you don’t want to. This allows you to rearrange what is where on your page so that what you want in a prime location is in a prime location.
Click Maps tell you what people are actually clicking on. He’s found that most people who are new to a site and it has an about page, that’s what they click first. Being that this is the case, you want to engineer your About page to assist in herding people into your sales funnel.
The evidence these two services give you great places to start testing.
Your First Fold Is The Most Important Place To Test On Your Site
This will inspire people to leave faster than anything else will on your site.
If you pay for traffic or put a ton of time into generating it, you want to make sure that what they see when they show up gets as many of the right people to do what you want them to do.
And this is why it’s crucial to cater and customize landing pages for each unique source of traffic. If you’re doing PPC you should have different landing pages for different keywords and groups. You want to match the mindset of the visitor showing up – affiliate traffic is warm – PPC is cold.
What To Test In Headlines
Unless you’re Amazon.com or some other mega brand even old people know about, your site needs to have a headline.
- Headlines the different sources of traffic see
- The Length
- Benefits/Emotional Hot Buttons
- “How To”
If someone came in from an ad that had a “How-To” headline, it’s important to continue that conversation by having the headline on your page be a “How To” as well. They clicked with this in mind so keep them so cater to this.
- Spelling Out Measurable Results
He’s found that headline that denominate a specific result you can expect to receive “Lose 10 pounds of belly fat in 90 days” will always out pull a broad and general claim like “Lose Weight Fast”.
- Negative Focus
People are trained to pay attention to negative headlines by the press.
After Your Headline Tests
Video vs. Graphics vs. Text
He’s found that no one modality completely dominates the other but that certain markets and certain traffic have their buttons pushed by different modalities.
Graphics can convey how something works easily and quickly.
He’s found that traffic that’s been warmed up by a referrer that comes from affiliates and such, respond better to video than any other modality. Also, he’s found that cold traffic that’s come in from a banner or a PPC ad don’t respond as well to video. The reason he believes this is so is because video takes more of time commitment to wait for it to load sometimes as well as to watch.
The company Basecamp increased their conversions on their site by switching from a headline with a sub-headline and a description of what they did to a graphic that showed in pictures what their technology makes possible for you . . .
mynotetakingnerd.com/blog/wp-content/uploads/2012/04/clip_image0024_thumb.jpg" alt="clip_image002" width="400" height="210" border="0" />
This graphic allows you to get what they do in one glance vs. reading an entire page. As a result of doing this, they increased their conversions by 14%.
Stuff To Split Test For Your For First Fold
- Limit options to two things – Leave or do the one thing you want them to do.
You have to thing about what you want them to do on that page – read, click a link, opt-in and then look for what is on the page that is derailing them from doing this and start pulling them out.
- Showcase Other People Saying You’re Awesome
- Customize According To Their Location
So one example of this is if you have a an offer for free shipping you’re leading with, you can say, “Free Shipping to “Their City”. He’s found that people LOVE them some free shipping and if this is directly relative to them because you’ve used technology to spot where they are, it hits home even stronger.
AVOID THE FOLLOWING MISTAKE LIKE IT WAS HERPES . . .
Kill any slow loading content that cold traffic will see. When you have only five seconds to suck someone into your vortex, you don’t wait those five seconds to blow by with them looking at a buffering signal because if that’s the case, they’re gone.
You’re going to better off with static stuff especially with people who have no clue who you are.
Order Button Tests
- Only Let Them Order At The End
People have found this work better with a video salesletter than if they made it available at the start or sooner. The only way to find out what your market likes is to test it.
- In Key Sections Of The Sales Letter
By layering your buy button throughout the sales letter you’re trial closing essentially asking them if they’re ready to order yet or not. If they aren’t they can keep reading. So you could test putting after all your features and benefits, after your social proof, and then at the end.
- Order Button In The Top Right Header
He’s found that an ecommerce site selling multiple products needs to have the order button right up top and easy to find.
- Shiny, Bright, and Loud
A prime example of this is the “Belcher Button” which you see on any of our sales pages…
What People Need To See When The Click To Order . . .
Make sure that at the top of your order page you sum up the promise of what awaits them, the discounts they’re receiving, and your guarantee. Another good idea is to have testimonials here.
You want to do this assuming that someone skimmed and missed everything you’re summing up here for them.
Pricing Aspects To Test
The only reason you wouldn’t test your prices is if you’re selling a commodity. Otherwise, you always need to be testing price because the market is always shifting.
Anytime he goes in to consult with a business, he can go in and increase the price – example he gave was going from $63 to $69 or $179 to $199 – and not have it hurt conversion at all.
- Premium Option: Sell less units, charge more
- Value/Volume Option: Sell more units, charge less
- Payment Plans
If you make payments available, you always want to make sure the full payment option is the most attractive unless you’re intentionally trying to make continuity income via a payment plan.
- Bundle Products Together – Sell Them At High Prices
If you’re selling single products, you always want to look for other products to bundle with it that you can offer them that add more value to the purchase once they’ve already made the decision to buy in the first place. Once people have made the decision to buy, it’s about a thousand times easier for them to decide to buy more if it’s only put of them.
- If you have none now get one. If you have 90 days, test extend it to twelve months or for a lifetime
His experience is that longer refunds convert better AND lower your refund. If you have a short guarantee period you give people the sense that you know something is gonna go wrong with this after 30 days from when they bought it. This means that if they aren’t absolutely dazzled by it before then, which in most cases of information products they aren’t because they haven’t even opened it, they’ve got to send it back before they find they’ve been burned . . . again. When you guarantee it for life, it implies you trust it’s going to be solid forever, otherwise you wouldn’t put yourself out there like that, so the buyer can just relax and get around to trying it someday.
- You can test only guaranteeing the bigger orders of what you sell
He’s found that in tests he’s done that people will pay more when you only offer the guarantee on the highest priced option.
- Find graphic guarantee seals to use
- Test showcasing your guarantee either in the top fold, right before you ask them to buy, or on your order page
- Facebook “Like” Button
- Tweet Button
- Digg Button
You want all of these buttons to have the counter next to them. Especially if it shows the number and then stalks their Facebook profile and shows you people they’re friends with there who have also ‘Liked” it. This allows for you to get the third party indirect endorsement from their friends.
- Testimonials: video vs. photo vs. written
Videos are the best ones to have here. Not as hard to get as they used to be with web cams, phone cams being cheaper to access than ever. If you can’t get a video one, you want to gun for a written testimonial with their picture next to it. If you can’t get a picture, then you just roll with the words.
- Celebrity In The Industry Endorsement
Derek Gehl is a nobody outside of an internet marketing discussion or seminar. But he’s a big somebody in these areas. Having him endorse anything internet marketing related would boost your cred with this market in a huge way. You always want to look for ways to make your celebrities relevant to your audience.
- McAfee, Hackersafe, etc.
He’s found that if you’re in the tech niche, these heavy duty endorsements are significant but he’s also known of people in non-tech niches, that you can make a statement on a seal stating that your site is 100% verified and secure (because it is if you’re using a shopping cart) and it has the same effect as if you paid the super high fees to McAfee.
- Better Business Bureau Seal
His testing has proven that the only places these security seals ever have any impact on conversion is when they’re tested on your order page. Don’t put these seals everywhere.
- Entry Pop-up - Popupdomination.com
These guys use light boxes. We love this! It’s kicks ass! Watch for using these on sales pages though because they can decrease your response.
- Opt-in in the top fold on left vs. right margin
Top right has usually tested best for him.
- Newsletter update vs. One Time Offer
A one-time offer like an app, a report, a video, etc. is what he’s finding is out pulling Newsletter offers. And the cool thing is, you can let people know when they opt in to get their report that you’ll be staying in touch with a newsletter so you get the best of both worlds.
You’re looking for the one mega benefit they’re going to get for opting in. Can’t be complicated. Needs to be sexy.
- Submit Button
This is your call to action so it’s definitely something to test.
He believes everyone should be using video somewhere – whether it be your salesletter or your testimonials or these other ways. He’s found that you always want to go with auto-play, video that start automatically when a the page loads.
This is a technology that allows for a person to walk out of the side of your site and give an introduction to your visitor. You go to their site, and for $500 bucks you can pick out a model, give them a script, they shoot the video professionally and then send you all the info you need to be able to plug it into your site. He’s found that in a big broad market this can be effective but not in the small world.
- Sundaysky.com – multi-SKU ecommerce sites
Information marketers have long used video more than ecommerce sites. There’s sites now that are recording demonstration/pitch videos for each of the products in their line (shoes, equipment, etc.). This service allows you to have someone shoot the videos for each of your products for you. If you’re selling a set of skis for $1,000 bucks it’d make sense to do this more than it would for the $5 dollar pocket hand warmers.
- Video salesletter vs. Traditional salesletter
You always want to split test this.
- Video salesletter with a traditional salesletter underneath it
- Test the length: Different markets and different traffic have a different attention span
- Screen Capture vs. Direct To Camera (infomercial style)
Screen capture works well for selling information. But if you’ve got an object like a sham wow you’re trying to sell, letting people see it in action (infomercial style) is a thousand times more compelling than talking about what it can do. If you do go with the infomercial style, you gotta make this look professional all the way down to the person doing the demo to the quality of the video (easy now to do HD).
Random Stuff To Test
- Free shipping
People love them some free shipping. You can raise the price of the product to cover you paying for shipping and reap the benefits of more sales.
- Adding shipping insurance (insureship.com)
This is an actual legit insurance company that will replace something that got lost when it was shipped. It’s something the customer can just check off and add to their order and you end up making a couple of bucks each time they do.
- Having an image of your digital products and bonuses (ebooks, box for software, CD covers, etc.)
- Sans-serif fonts
Font you see in print – books, newspapers, magazines – is in serif. Font you should be seeing online is Sans-serif. The older your market is, the bigger you want your font to be.
- Light/white background colors free of patterns
He’s found this always works better. Test it.
- Add a dollar value to anything you’re giving away
- Add as much detail as you can to your testimonials
Name, city, profession, picture, etc.
- Hand Drawn Graphics (copydoodles.com)
This can work in your screen capture videos as well as your traditional sales letters.
- If you sell multiple variations of something, show it
If you’re selling socks and they come in three colors, show the three pairs of socks on the same page instead of thinking they’ll go to your drop down menu and do it themselves.
ALWAYS TEST – NEVER STOP TESTING – NO MATTER IF SOMEONE TELLS YOU IT’S THE DEFINTE TRUTH
If You Redesign Your Site Do You Do It One Element At a Time or Redesign Fully and Test?
If you’re going to do a complete site re-design using best practices, do it all at one time, but you still want to split test the traffic you’re getting to the new and the old site. If the new one wins you’re golden. If not, you know you’ve got work to do.
What kind of role does web chat have in the conversion process?
It’s something to test but at this point in time, he didn’t believe web chat would increase conversion because he feels he’d rather get serious potential buyers on the phone and take the order and answer questions that way otherwise you might have a bunch of people who end up wasting a lot of your time going back and forth.
He’s been testing live chat for years on different businesses and been able to make it work.
That’s it’s for today. Go put at least one of the aha’s you picked up here to work for yourself. And once again, feel free to stick aside a few thousand dollars of your profit that came from what you learned here so that you can buy me lap dances and champagne at the Olympic Gardens strip club when you come to Vegas.
Lewis LaLanne a.k.a. Note Taking Nerd #2 a.k.a. L. L. Cool Nerd