Cracking The Secret Affiliate Combination For Any Market
Affiliate Marketing, Pay-Per-Click October 24th, 2007If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!
In my last post, which was sadly about a week ago, I gave an example of a particular affiliate offer that may or may not convert well with pay-per-click marketing. Really though, most markets have the potential to be profitable if you just figure out the secret combination. Contrary to what many might think, the way to find this secret is not through a bunch of expensive ebooks or membership sites. Instead, you must test for yourself in order to determine what does work and what doesn’t work.
Rather than talk about the theories and concepts of testing different campaign elements, this post is going to talk about what I do when I really want to ensure success in a new market. It requires a lot of additional effort, and there is no way to guarantee that it will make a profit, but by finding out exactly what works and what doesn’t work, it’s pretty hard not to.
Build Multiple Landing Pages
The first thing I would do, before even creating the pay-per-click campaign, is build multiple landing pages. If you’ve never seen the test results before, you would be surprised at what a difference design, color, and layout can make. Some markets might respond well to professional-looking landing pages while other may prefer amateur designs.
I would probably start off with two or three different design variations which would be wrapped around the same content. You don’t have to create separate pages if you know how to put together a simple template system. All you would need to do is create the new theme files and your pages would automatically inherit their design.
The variations in design I would use would include overall look and feel (amateur vs. professional), color, and content layout. Anything can make a difference, so try to make use of creative combinations. The only thing you need to be careful of is making too many landing pages because it would be an incredible amount of work and would be difficult to test.
Once you spit test your initial designs, you can introduce more completely new designs, or you can simply start changing things around on your winning design.
Swap Landing Page Copy
After you have concluded which of your landing page designs is the best, it’s time to start changing the words. Try adding a magazine-style headline, or if you already have one, try completely rewriting it. See how the use of textual call-to-actions vs. graphical call-to-actions effects your conversions. Rewrite your paragraph transitions to guide the reader down the correct path.
Many promotions could benefit from the use of pictures. See if you get more conversions with pictures or without pictures. Also try using stock promotional photos vs. homemade photos.
Most affiliate promotions that I do aren’t overly heavy with sales copy, so there isn’t as much to change around as if you were testing a long-form sales letter. Just keep in mind that they way you word things, and the emotions that you illicit, can be a big factor in the visitors decision to take the desired action.
Track Keywords By Group Or Individually
How precise you want to get from the start is up to you, but you are going to eventually want to track your keywords individually. I used to start out with ad group tracking, but lately have been using keyword-level tracking right from the start in an effort to speed up the entire process.
Once you have a system in place, it’s very easy to implement it on any new campaign that you are building. My favorite way to track keywords is to send the tracking data through the destination URL and then append it to a custom tracking id. I then get the conversion reports and process them with a custom script that gathers the information I need. I do this so that I don’t have to rely on affiliate managers placing my “conversion code” on their landing pages. This method takes some extra work, but is much more reliable.
Know When Visitors Came To The Page
In addition to tracking which keywords are producing conversions, I would also track the time of day that they landed on your page. This will allow you to increase your position during high conversion times and decrease your costs during low conversion times. Depending on your market, the time of day can be a big factor in determining how impulsive they are or how prepared they are to make a buying decision.
Write Multiple Text Ads
Anyone who knows anything about pay-per-click marketing will tell you to write multiple text ads so that you can get the highest click through rates and improve your positioning, but that’s not what it’s all about. Your text ads also play an important role in your conversion rates.
As an extreme example, if you write a text ad that appeals to many people and says “Get Free Movies Here”, you will probably get a pretty high click through rate. However, if they come to your page and see that they have to use their credit card, they probably won’t result in a conversion.
Make sure you know which text ads produced sales so that you can figure out which ones are matched correctly with your offer and are putting your visitors in the right state of mind. It makes no sense to have a high click through rate and a low cost per click if none of those clicks are generating a return on your investment.
Putting It All Together
None of this is necessary to make a profit in affiliate marketing, but I guarantee that it will increase both your chances of success and your potential profits exponentially. When I first started out, I didn’t do any of this, and I got pretty lucky. However, the pay-per-click game is a lot tougher now, and it makes sense to do all you can to optimize your efforts.
While all of this might seem extremely daunting, implementing it is relatively simple. The hard part is collecting statistically relevant data, interpreting it properly, and making the right decisions. If you can succeed at that, you will have a very good chance of profiting in any niche you choose.
October 24th, 2007 at 3:45 pm
“The hard part is collecting statistically relevant data, interpreting it properly, and making the right decisions. If you can succeed at that, you will have a very good chance of profiting in any niche you choose.”
Great topic for a future post (or series of posts)??
What is your thought process every step of the way?
How do you determine when something is worth pursuing and try to make it work and when to dump it?
October 24th, 2007 at 6:00 pm
Hey JL - I was actually thinking of possibly taking my custom programming and creating a version for end users that would help interpret tracking data. I don’t yet know how much work would be involved though.
I will definitely take your questions into consideration for a future post though. It’s always nice to get writing inspiration from readers!
October 24th, 2007 at 6:22 pm
Great blog Derek.
Just a quick question for you:
When you’re split testing landing pages and deciding whether an affiliate program is a “goer”, what sort of PPC budget do you have? Are we talking $100s or $1000s?
I know Amit at Super Affiliate Mindset says he spends $1000s trialling offers - can’t wait until I can do the same!
October 24th, 2007 at 6:32 pm
Hey Hadi - It’s really dependent on the cost per click. If you have a smaller budget, you will probably want to test fewer variables and fewer versions of each variables. The more you are tracking, the more data you need to make sure it is reliable.
On the other hand, if you can get it to a break even point or slightly profitable point from the start… what does the budget matter?
October 24th, 2007 at 7:55 pm
great post - the part I find I am running into challenges with is simply designing the landing pages. I understand the concept of building out affiliate authority sites - it seems where I am running into trouble is creating a landing page with enough quality for 1000’s of keywords to make google happy. I am modifying the content dynamically based on the keyword so I can have templates as opposed to having to create individual pages, but i still run into problems. Love to hear more about how you manage this, and what you do with the root level domain that you hand your landing pages off that you are testing with.
October 24th, 2007 at 8:38 pm
Chris - I pretty much always buy dedicated domain names rather than using a testing domain. They are too cheap not to.
October 25th, 2007 at 5:58 am
It’s a great idea to make something like “affiliate marketer tracking kit”. I think many marketers that start with their ppc efforts would say a BIG thank you.
October 25th, 2007 at 12:32 pm
Derek - great info again. Question…when setting up your landing pages for these new markets do you follow the same format/strategy as creating PPC landing pages for authority sites. Single and Multiple offer landing pages and using the PPC revenues from these new market campaings to develop authority sites?
October 25th, 2007 at 12:41 pm
Craig - it depends on the market, but yes, a lot of them follow the same format: a comparison review page and individual review pages.
October 25th, 2007 at 3:39 pm
Derek, if you are putting each new offer on a new domain name, what is the minimum you build out the site to get a solid QS ranking for your landing page? And, how aggressive are you on the landing page of creating content tied to the keyword?
October 25th, 2007 at 4:33 pm
Hi Derek, another great post.
I would also be interested in posts on gathering and processing stats, as well as any code you could offer to jump start the process. I can program, and I would rather have my own custom solution than buy a shrinkwrapped product (if one exists). But it seems like a waste to do the whole thing from scratch seeing as there must be others (like yourself) out there who have already created solutions. What do you think of an open source solution?
October 25th, 2007 at 10:01 pm
Derek,I like your blog.I’ve learned so much from it.
But,the feeds seems doesn’t work.Can you have a check and fix it? Thank you.
I tested it here:This feed does not validate.
http://feedvalidator.org/check.cgi?url=http%3A%2F%2Fwww.derekbeau.com%2Ffeed%2F
October 26th, 2007 at 6:14 pm
Nice….I guess a comprehensive split testing/taguchi post is in the works?
PS: Are you still on your 30 post per month goal? Personally i think quality overrules quantity any day of the week, whether it’s traffic or blog posts…
October 29th, 2007 at 1:10 am
Hi Derek,
Great post. What software do you use to split test your landing pages?
Ronnie
October 30th, 2007 at 12:09 am
Wow, this could have easily been several posts … I reviewed your site a while back and think I called in Derek beau University … I wonder what’s higher in the education world than university.
Everything you write about I have seen somewhere before but nowhere does anyone have the “whys and wherefores” of what you are doing the way you do.
A question which ties in with the comment by Chris, above. let’s say a guy wanted to build a campaign to sell some “how to” books and courses that were all interrelated … in the same genre like, for example, internet marketing. In terms of QS and just plain keeping up with traffic and administrative duties, are you saying it’s better to have 10 sites: sellingbook1, sellinbook2, sellingcourse2, etc as opposed to a structure like www.mothlodeofmarketing.com and book1.mothload, book2.motherlode, course1.motherlode … etc.? Especially with the way Google eagerly “laps up”subdomains I kind of lean toward the one site per area idea … but then I haven’t done this much and you have.
November 1st, 2007 at 10:10 am
Great post, Have you ever thought about doing a start to finish Case Study. I’m sure that would have a HUGE audience. Basically taking a niche you think could be good. Picking the Keywords, making the Landing Pages, Sharing your bids, and then tracking progress. The tracking posts would be easy to make and give you something to blog about everyday.
Also if you’d like any help coding that conversion tracker let me know. I have a class written for Directtrack already to pull data automated, then have it populate a db where my clicks are also stored. I have some basic matching and click->conversion data. However would be nice to be able to get gclickid reports on click prices to get a full analytics setup going. If you have any interest in working together on this let me know.
November 4th, 2007 at 9:17 am
[…] [2] Cracking the Secret Affiliate Combination For Any Market - This is a solid, hands-on post that shows exactly what Derek Beau does when he wants to ensure success in a new market. […]
November 24th, 2007 at 7:49 am
great post Derek. When it comes to split testing , I go for usfreeads and cragislist before using ppc. I won’t use my money to testing purposes as I am not in a position to afford. SO i usually use USFreeads and test the conversion ratio and then use adwords. What do you think ?
December 13th, 2007 at 1:11 am
In the past, I tested multiple landing pages & it took me quite a lot of time. Recently I started testing dynamic landing pages & the results totally rock! http://www.theultimatepublishing.com/blog/one-landing-page-multiple-traffic-sources