I’m a testing junkie. I believe in testing so much that I not only do it in AdWords campaigns and internet marketing, but I actually did a two week live-in test before I let my boyfriend move in with me (at both his apartment and mine, to get something akin to statistically valid results over the month-long trial period). I have split-test cat food. I am a believer.
Testing can get really complicated, but it doesn’t have to be. The longer I’ve been testing, the simpler I tend to make it. Simple A/B split tests sometimes work better than elaborate multivariate experiments. The simple Google Website Optimizer tool is MUCH easier to use (in my humble experience) than the far more robust Vertster testing platform. Simple is really good in testing. There is so much inherit resistance to testing to begin with, making it as easy as possible is critical – it makes the difference between doing it and not doing it.
Over time, the difference between “doing it” and “not doing it” in testing can create a 100%+ increase in sales. Aside from tracking, I know of no other way to get great returns for so little work.
Of all testing, adcopy testing in a Google AdWords adgroup is the easiest. There’s no code to install, and to interpret the data you don’t need anything more complicated than these two sites: www.splittester.com or (for the more advanced) www.supersplittester.com.
It takes less than 30 seconds (seriously – 30 seconds) to test adcopy. And there are so many little things you can test. A good start would be to knock-off alternate ads every Tuesday, let them run through the week, then come back the next Tuesday and kill the loser, then knock out another variation to test against.
Everything I’ve outlined below are little tests. Think of them as testing the “leaves” on a tree. Testing “branches” on the tree would be checking out your competitor’s ppc adcopy and borrowing or modifying some phrases they use. Testing the entire tree would be the difference between a simply-worded, factual ad versus that kind of ad that screams at you. For example the difference of “Copper Semiconductor Wires” versus “Stick This Up Your Nose”. I have not tested the latter headline. It would get noticed, but would it get clicks? Would it convert? Only a test would tell.
Here are 25 things to test in adcopy:
1) In the display url: use “www.” versus no “www.” So “www.ForexMagic.com” is tested against “ForexMagic.com”.
2) In the display url: use initial caps for each word versus no initial caps for each word. So “www.ForexMagic.com” is tested againist “www.forexmagic.com”.
3) In the display url: use a fake subdirectory versus no subdirectory. So “www.ForexMagic.com/free” is tested against “ForexMagic.com”.
4) In the display url: use one fake subdirectory versus a different subdirectory. So “www.ForexMagic.com/free” is tested against “ForexMagic.com/14daysfree”.
5) In the display url: use caps in one fake subdirectory versus no caps in the other subdirectory. So “www.ForexMagic.com/14DaysFree” is tested against “ForexMagic.com/14daysfree”.
6) Test entirely different display urls. You will have to create a second adgroup (or a second campaign, if you want to go that route) because Google doesn’t let us use more than one display url per adgroup now. Deep sigh. This can make campaign management a big ole pain, but at least early on in a campagn’s life, we still need to do it. Display urls can make a HUGE difference in both CTR and conversion.
So, for example, you would test “www.ForexMagic.com” versus “www.ForexSecrets.com”. Ideally, you would test up to 10 different urls if you were pre-launch for a company. That means ten different adgroups for each set of keywords, each adgroup using a different display url. You want to do it early in your campaign development because it becomes a mighty headache to do after you’ve split your adgroups down into single keywords and now have 100+ adgroups even before you start testing display urls. But I digress.
7) In the headline: Test initial caps versus no initial caps.
In the headline: Test a question mark at the end versus no question mark.
9) In the headline: Test dynamic keyword insertion code versus no dynamic keyword insertion code.
10) In the headline: Test using “Official” versus not using it.
11) In the headline: Test swapping out the verb. For example “Search Driving Records” versus “Find Driving Records”.
12) In the headline: Test plurals versus no plurals.
13) In the headline: Test your competitor’s headlines versus your headline.
14) In the body copy: Test initial caps versus only capitalizing the first word of each line.
15) In the body copy: Test flipping the top and bottom lines. This works best if each line has a contained phrase, with punctuation.
16) In the body copy: Test breaking the words into the shortest possible sentences. For example, test “Free $50,000 Practice Account With News & Research, Live Phone Support” versus “Free $50,000 Practice Account. With News & Research. Live Phone Support”
17) Test switching out any one of the words in that adcopy. For example, test “Free $50,000 Practice Account” versus “Free $50,000 Experiment Account”
18) Test abbreviating words. So test “Free $50,000 Practice Account” versus “”Free $50,000 Practice Acct”
19) Test dashes versus other kinds of punctuation. Test “Free $50,000 Practice Account. With News & Research. Live Phone Support” versus “”Free $50,000 Practice Account-With News & Research-Live Phone Support”
20) In the headline or body copy: Test ™ and ® symbols versus not using them.
21) Test using free versus not using free.
22) Test including prices versus not including them.
23) Test using a dynamic keyword in the entire first line of body copy versus not using it.
24) Test using threat language (ie “don’t buy any zyx service” until you’ve read this”) versus just promoting benefits and features.
25) Test using “100%” versus not using it to emphasize something. For example, test “100% Accurate” versus “Accurate”.