7 replies to this topic

[hero]

Hi,

I am about to undertake a split run test using Duncan Carver's...

free script - The Scientific Internet Marketing Assistant (SIMA)

www.marketing-strategy.org

and I wonder what effect, if any, it might have on my search engine ranking.

Obviously the script will redirect my visitors from the primary index page to
one of two pages that I am testing. Will SE Spiders still index my pages or
will they assume that I am using some form of cloaking?

I don't want to start testing and then lose my rankings!

Thanks

James

[Jill]

Welcome, Hero!

I wouldn't run any tests like that except for pages that are excluded from the search engines through your robots.txt file. Use it for your landing pages, not the actual pages of your site that you want indexed.

At least, that's what I'm thinking...

Jill

[hero]

Thanks Jill,

Now that raises a real conundrum!

How do I test the headline and all the other key things on my index page?

All my traffic enters my site through my index page because it is a single page sales letter specialising in a small niche market.

Do you think it might be possible to get some guidance from the SE's?

Thanks

James

[Vertster]

Hi Hero,

One of the fundamental rules of testing is to control all variables expect the one you would like to test. I would suggest not testing your page as the homepage of your site. Website traffic is constantly in a state of flux- new links to the site arrive, old ones disappear. Maybe you get some press coverage, or your site gets mentioned on a discussion board.

The point is that all of these things can affect the accuracy of the test data. Some folks will respond to one headline better, others may respond to the other. When you don't have control of who is viewing it and when, you can be led to incorrect conclusions.

Splitting your traffic will not remove the effects of this variability, although it is certainly better than the "one after another" approach.

As Jill mentioned, it is not to your advantage from a SEO perspective either.

I would instead test from a contollable traffic source, such as Adwords, Overture, or a Popup campaign. When you feel you have determined a winner with targeted leads- only then move that creative over and make it your homepage.

Unfortunately, with organic SEO, you run into problems with getting the copy properly optimized for the engines, and maintaining a high conversion rate.

[dcmarketer]

Hi James,

Firstly - Unfortunately I have to agree to disagree with what Vertster mentioned here. Rather than explain it all again, I've already explained it in an Interview I participated in earlier on in the week. It will highlight why you need to test, regardless of where you're traffic is coming from - home page (index.htm page) included.

In regards to how it affects your rankings. It will have a slight negative impact, particularly if your sites poorly optimized overall. Check out this thread, after John's latest test result to see an actual example of the impact it had on my rankings (nil) and why.

As mentioned, I'm looking into ways to eliminate this negative impact entirely in the next update for those seriously concerned about it. I have a fair idea in mind as to how to do this in the best fashion. Can I ask how much traffic you're currently receiving from the search engines? If it's not going to have a huge impact right now, then I would recommend you go for it. If you want to get your testing done fast, use the pay-per-clicks and run several campaigns during the space of a week.

Anyways, hope this helps some...

Sincerely always,

Duncan Carver

Marketing Strategy Discussion Forums

Online Marketing Today

[Timothy Warnock]

Hello All,

This is my first post here, I just heard about this forum, and I immediately stumbled upon this topic that is so close to home, so I haven't even had the opportunity to lurk and get to know you all through your posts.

I hope I don't come crashing in like a bull in a china shop!

James poses a very interesting dilemma that is not easy to answer with a patch kind of fix. A whole view perspective is needed. A lot of people could find themselves in a real pickle on either side of this argument - Integrity of Optimized Page vs. Important Tracking Feature.

I think it is very important to keep in mind that split testing is just one function of an overall tracking approach towards successful online marketing. Granted that it is fast and arguably an accurate way to control one's promotional web page, I think, however, that we should put it in its proper context.

Let me use an illustration (in an offline setting) of what we see so commonly amongst businesses online that want to sell something...

Let's say that I want to sell lemonade. So I go to a crowded park on a hot day, and start putting up signs on all the busiest paths saying, "Tim's Lemonade at the playground", I set up a table, but unfortunately it has a broken leg, so it teeters, and the cloth I use to cover it is too big and droops on the ground, and the cups I brought are cheap plastic, and brown, and I charge way too little - 3 cents per glass. All day long I watch as hundreds walk by my table, but nobody, (except a couple of kids without money), stops to ask for a glass of my lemonade.

Now, how many do this with their websites? Thousands upon thousands.

Somehow they gather traffic, but they often insult the visitor, or bore them to death because they never learned the psychology of selling.

I think we will all agree that targeted traffic is essential towards online success, and that traffic from SEs is often an incredible source of targeted visitors. The problem is that so often, SEO becomes the first and main focus of those that create a web presence, without giving much attention to their particular proposal, they then discover afterwards that having traffic is not always enough towards converting visitors to clients. So, the copywriter, the opt-in newsletter, the JV's, etc., start coming into play in the hope that things improve, and they often do.

Now, without accurate tracking, it is not wise to do further modification with the hope of increasing sales, because your changes are blind changes, perhaps costly and inefficient. So finally, individuals and businesses begin to slowly catch on to the importance of tracking. But for one reason or another, it too often holds a back burner position. Marketing is really very scientific, but it surprises me to see how few actually apply the scientist/tester approach to their marketing.

Let me give you another illustration of how I think tracking should be approached in regards to online marketing .

Let's say I own a small but nice home, and one day I have the immense pleasure of being able to purchase a Monet painting. I certainly won't hang it up in the kitchen, on the only part of wall that is left nude, next to the calender.

No, I would empty out the most important room, find the most attractive and dignified wall, and hang the picture there. Then I would rearrange all the furniture based on the position of this painting - in other words, it becomes the central focus, not just another wall hanging.

This is where tracking should "hang" regarding your online marketing.

James, I should apologize, I am turning your simple question into a long discourse, I just wanted to make it clear that tracking is an essential element towards successful and long lasting online marketing, and shouldn't be a patch fix, before I answer your question.

To answer your simple question - do not risk sacrificing your good ranking by using a split tester, however, it can and should be used in other areas, with all the other forms of traffic, just not on optimized pages. And, by all means track! And track everything! A good tracking solution will NOT sacrifice your search engine rankings in any way. You should have already tracked before optimizing pages, but that is another discussion in itself.

I hope my rambling helped you in some way.

My best wishes,

Tim

[hero]

Everyone,

Thank you for such detailed responses.

I guess the answer is that if you have a relatively new site with no appreciable presence in the search engines you should take the time to test and tweak it as soon as possible.

If, however, you have a well established web site with established rankings then it is better to set up an independant test without jeapardising your existing index page. Obviously there is a price to pay in that you would need to pay for traffic through PPC etc.

James

[Timothy Warnock]

Hi James,

I guess the answer is that if you have a relatively new site with no appreciable presence in the search engines you should take the time to test and tweak it as soon as possible.

One good strategy to consider when starting a site with the scope of selling, is to first create a sales page with your particular offer, and then try out the PPC engines (Google Adwords is probably the best). Track everything that happens - use a split tester if you like, and keep tweaking for a while till you start to see some interesting conversion rates.

A good tracking program will also tell you which keywords bring you the best results (like ours ) - once you know this info, 500 - 1000 visitors are enough, then optimize pages for the keywords that brought you the best CR% - and funnel your traffic towards your converting sales page. Or, if you are a clever optimizer, and can manage to sneak in the keywords into your sales copy - do it directly onto your sales page, but this is tricky and takes some experience.

If, however, you have a well established web site with established rankings then it is better to set up an independant test without jeapardising your existing index page. Obviously there is a price to pay in that you would need to pay for traffic through PPC etc.

Only partially true. This applies specifically to tracking features and tests that risk hurting your rankings, such as a split test function that redirects. If you use a top notch tracking solution on your optimized pages, you will see how well they are really pulling, and exactly what keywords, and which search engines are bringing you results. This is essential knowledge.

A good tracking program will tell you how many are clicking over to your sales page and how well it is doing with ALL of your traffic - a split tester is just a function of this process, and helps you to optimize your sales page with greater ease, but it isn't essential. The PPC engines are useful and at times profitable, but not necessarily essential to know what is going on.

Again, I hope this helped.

Tim