The magazines in the checkout aisle at the grocery store have mastered the art of capturing interest within a glance. If you are pressed for inspiration, go check out the headlines on these mags.
Peel and Stick is the tactic of peeling off keywords that are doing well in an adgroup and making a unique adgroup for that keyword. Also you should always be split testing multiple variants of the same ad. It works like Darwinian selection so the campaigns with the highest CTR rates survive. You constantly mutate your ads over time and go with the winning strain on each iteration. This is the key takeaway from this book- this knowledge is priceless and becomes your invaluable IP. Months of split testing yield a rock solid campaign with unbeatable effectiveness.
This is a trick I thought was reserved for companies like eBay and Target. Turns out anyone can insert the keyword search text dynamically in ad copy using the braces syntax. From my experience this technique yields a high volume of mediocre traffic. If you can get away with lower CTR, this can definitely bring you loads of traffic. Just be sure you have a daily cap on adspend before you try it. also I would limit it to the Google search initially.
The content network is composed of sites that run Adsense. By default new campaigns will run on the content network. It is better to start your campaign confined to just the standard Google search pages to refine your messaging and then expand into the content network. Typically the content network has lower CTR. It is also a different psychology of reaching the person who is already on a web page where the ads are appearing vs. the person who is actively searching for what you have.
You should write ads differently for the content network because the situation is different. You are essentially stealing the user away from the site on which the ad appears so you have to interrupt them with something that is compelling enough to sidetrack them. This is different than writing clear copy that speaks to the person who is actively searching for your offering on the Google search engine.
When you are split testing image ads it is important to control all variables but one. Different ad sizes will yield different CTR so you will get tainted results if you are testing different versions of an ad and simultaneously varying the banner size. Either test the same ad with different sizes or vice versa.
Just about all products can be ultimately commoditized over time. Your brand and style is defensible. Inject personality into your ads – sterile ads and newsletters are easily forgotten. People will come back to something that is authentic and speaks to them. Think Joel Spolsky, Kathy Sierra and Paul Graham- people that infuse their writings with their personality and tell it like it is.
The analogy that many people tried to make a flying machine by putting a bigass engine on a craft and focusing on the locomotion aspect first. These attempts all failed- the Wright Bros were successful because they made a glider that would fly under its own power first and then adding the engine afterwards. This reminded me a lot of Clayton Christensen and his idea of being patient for growth but impatient for profitability. Start small with a campaign and refine your messaging until you have the formula working- no sense in dumping money down a hole and growing the campaign if it is losing money. Use split testing to nail the execution and then it will be a no brainer to dump fuel into the engine.
Expanding universe theory is the idea that there are orbits of systems of diminishing importance. The core system that likely has the greatest ability to generate traffic is Google Adwords. Master that first and then branch out into other systems to capture more market. Overture does not currently have some of the nuances that Adwords offers that can give you an advantage. Overture is like roulette and Adwords is like blackjack in that once you know how to play, you can stack the odds in your favor with Google.
This is a key concept- split testing different aspects over time yields compounded returns. Consider the phases of the funnel: person sees ad, person clicks on ad, person visits your landing page, person takes the call to action, person receives email and returns for future purchases. Provided you conduct split testing and refine each phase of that funnel you will see massive improvements. So for example doubling effectiveness of each step of a 4-step process will yield a 16x return.
The idea of a marketing budget is an arcane carryover from old media where there were significant fixed costs. New media like adwords is entirely marginal costs and therefore if you can tune your adwords process so putting in one dollar returns two dollars, you would be an idiot to cap your marketing budget at a fixed amount. There is a ceiling to adspend but it is defined by the size of the total market at what you are able to capture. Either an Adwords campaign works and makes more money than it costs or it fails and burns money. There is little reason to continue a failing Adwords campaign unless the primary motivation is pure branding or some objective other than profit.
Ad campaigns default to run in optimized mode where Google will intelligently display the more effective ads more frequently. If you are in the early stages of tuning a campaign and making regular tweaks, you should turn this off and have them rotate round robin to expedite the split testing. Use the optimized mode once your campaign is humming and you are no longer babysitting it.
If you have keywords that generate few searches they will fall off the map and become deactivated. Google will tell you at this point you need to up the bid price to reactivate them but you can actually revive them purely by breaking them out into their own adgroup with the keyword in the title of the ad. This is a cheaper alternative than upping the bid plus it has a higher likelihood of producing a better CTR when the search term appears in the ad.
There are a lot of countries that generate tire kickers, people that will click through and never buy. It depends on what you are advertising but it is likely that when you are in the initial stage of refining the campaign that you want to confine your geographic targeting to higher quality markets and exclude countries that sap your conversion rate. This is something you can enable once you have things refined and want to get more volume of traffic.
