A common question out there when it comes to testing is, can conversion optimization testing affect SEO?

Yes, I believe a bad test structure can have stronger side effects than most people realize - and not always for good. Depending on the scenario, embarking on the testing ship without due diligence can have harmful effects on SEO. 

Below are some of the dangerous testing scenarios that, in my opinion, could harm your SEO efforts. 

Note: The scenarios below are based on questions from past clients. I don't have solid data, have not tested this nor know of anyone who has, but I would like to open this discussion to what I think are solid arguments. If you see a flaw, please let me know.

First let's make one thing clear: Most Testing Tools Are Safe and Do Not Affect SEO 

Testing tools like Google Website Optimizer, Omniture's Test and Target and UnBounce to name a few, are safe when it comes to SEO.

Search engines can identify when a testing tool is being used on a page and will not confuse your variation with new or duplicate pages - should you set it up correctly, of course. But that's assuming all goes to plan, and as those of us who live on planet Earth know, it rarely does. 

So here are the different potential backlashes of testing, should you not know what you're doing.

Danger #1:  Tests That Run For Too Long and Hinder SEO Evolution

A very common mistake for those with little testing experience is to design a test that will take too long or will never finish. 

New SEO strategies, ideas and overall SEO evolution depends on elements that in many cases are considered in a test. Many times testing involves different copy, anchor links, headings, images, videos, and many other elements on important pages that impact SEO.  If those are not decided, your SEO team will have a hard time trying to figure out available assets. 

If both SEO and conversion optimization efforts that include dramatic changes are being done on a page, SEO efforts, like everything else, will have to wait until the test completes. Sometimes variations exclude or include whole headers, footers, menus, and other page assets that deeply impact linking structure, for example. 

For those in competitive industries, search trends evolve fairly quickly. Having a test running for three, four, five or more months hindering your SEO team/person, could translate into lost traffic. Also, if you've engaged a third party for SEO, expect for them to put some kind of halt to their services if the test is being done on an important one. 

Possible solutions:

• Hire a professional
• Be careful when considering the duration of the test. Use a calculator like the Google Website Optimizer Duration Calculator or do a search for "test duration calculator" and you'll get a few options.

Danger #2: Testing Along Social Media Efforts

Search engines take information in social media sites to create relevancy between a subject and the content on a page being linked to. If you're running social media efforts pointing users to a page being tested, social media enthusiasts will start exchanging information  - the nature of social media - depending on the variation they see. If the variations differ, so will the information these people will exchange.

Most of the times variations include different content, ideas, or headlines, and social media enthusiasts will start talking about your same subject or products in different ways, using different words, tone, talking about different features, different anchor text, which could dilute the SEO benefits of your social media efforts.

For example, if an anchor text on Page A pointing to Page B is close to the H1 of Page B, it will get more SEO brownie points that if they don't. Accordingly, if a famous industry blogger starts chatting about your company using information, text, h1s, h2s and features found on a variation that is not implemented, the SEO power will be diluted - even with a 301 after the test finishes (which you should do regardless if your test uses different URLs).

Possible Solution:

• Keep communication open between your SEOs, conversion optimization team and your social media team, and let everyone know possible SEO side effects.

Danger #3: Testing Phone Numbers

Testing phone numbers can be pretty tricky - or as I like to call it, a pain in the ***. It's one of those scenarios that look easy until you actually do it.

Most of the times, phone numbers mean a site-wide test. That's because a visitor could land on a variation page, be persuaded, but make that phone call conversion from a different page. So that means that accurately measuring most tests where a phone number is involved, you have to change the phone number on the landing page, the "Contact Us", footers, headers, the shopping cart, images, etc...

So how does this affect SEO? Local search correlates phone numbers out there and, the more consistency search engines see throughout web, the stronger your site's local impact. The danger is not in the tool. These can be configured to override this problem. But if your tests run for a long time, users will start to exchange information, complete or update profiles on sites like Google Maps, Yelp, Foursquare, which in time becomes an inconsistent data point for search engines - minus SEO brownie points.

Possible Solution:

• Hire a Professional.
• Ensure you're using the tool properly so that search engines identify the right phone number.
• Structure your test to finish quickly.

Sorry folks, but conversion optimization is not SEO

SEO and CRO can work together, but the first step is not to take testing lightly. It’s its own beast.

Just like you learned SEO, PPC, web design, or analytics, you have to hit bloggosphere and learn conversion optimization – because the actual test is just part of it, how you structure it can affect all your online efforts.

The thing is that just like SEO, because you have great free tools out there like Google Website Optimizer, it's hard to either believe or sell the idea that you have to pay for someone to use it.

But I'm a believer that you should pay to people who know what they're doing if you want to get the best out of your time.

Please let me know your thoughts on the comments below.

(Thank you Flickr users Uzbecka, Morten Liebac and daren_ck for the images.)