AdWords (https://adwords.google.com) is just the sort of advertising program that you might expect to roll out of the big brains at Google. The designers of the advertising system have innovated thoroughly to provide precise targeting at low cost with less work; it really is a completely new way of looking at advertising.
The "less work" part is something I take seriously. Focus on structuring a campaign in a robust way so that you can understand and adjust it. Don't beat yourself up by trying to get everything 100 percent perfect. A "robust" campaign has a well-planned category structure, or "ontology" (basically, themed campaigns and ad groups). And a robust ongoing advertising project includes easy-to-follow metrics such as cost per order, and an easy interface for reading those results. Remember, you might wind up handing this work off to someone else. Don't make it convoluted.
In the early days (AdWords 1.0), the platform offered few features and charged a fixed rate (in cost-per-thousand impressions, or CPM, format) for ads that would show up near search results, triggered by a user query on a keyword or phrase you placed in your AdWords account.
In 2002, Google came out with version 2.0. The pricing was based on a cost-per-click auction that allowed advertisers to bid, but also incorporated click-through rates (CTR) into the ad rank formula. The idea was to push more relevant ads higher on the page.
In August 2005 (with experiments and refinements taking place prior to and following that), Google changed the system fairly significantly. Instead of CTR, the new ad rank formula became Max CPC (your bid on a keyword or a group of keywords) multiplied by Quality Score (QS). QS is multidimensional. Google states that CTR is predominant in QS.
But other factors, including landing page relevancy and ad copy relevancy, can play a significant role in where your ad now ranks. It's largely a black box, but if you've followed Google closely, you may be aware of the types of issues it focuses on. On the organic search side, certain kinds of deceptiveness and poor user experiences are judged as "evil," and you could wind up paying the penalty. It's safe to say that some of Google's main pet peevessuch as pop-ups, deceptive redirects, and pages with nothing but graphicsare included in the list of potential QS criteria.
If a keyword in your AdWords account has a low quality score, you might be asked to bid very high$5 per click isn't uncommonjust to show up at all. Some advertisers are relatively unaffected by these changes. Others have been virtually wiped off the map.
Between this and the increasingly competitive environment, advertisers must become quite systematic with how they test and refine campaigns. Those with large accounts now demand more features to help them stay organized, to control ad delivery, and to report on their results. Irrespective of the fundamental "version" of AdWords that I'm calling 2.5, the past couple of years have seen Google add dozens of small features and refinements to the platform, sometimes seemingly on a weekly basis. Let me draw your attention to a few notable ones. You need to look elsewhere if you seek in-depth programming advice vis-\x88 -vis the AdWords API and the like. To borrow a phrase from Jim Sterne, "I'm a marketing guy."
Dynamic Keyword Insertion: Beware
Most of the time, users click on ads with titles and text that closely match what they're looking for. Lazy advertisers who want to take advantage of the ability to match the user's query in their ad title or ad text can use dynamic keyword insertion. The format is {KeyWord: Alternate text }yes, you need to include those squiggly brackets (called braces). So, if this is placed in your ad title and the user types Red Cactus, Red Cactus will be your ad title.
A typical use of this format is a large list of products that you don't want to create separate ads for; you want users to see something better than a generic title. Here's the problem with dynamic keyword insertion. Even factoring in the ad rank boost you get from the higher CTR that generally comes with matching the user's query, the return on investment of ads using such matching is ofteneven usuallylower than if you use a manual ad title that is slightly less Pavlovian in its appeal to searchers. You do pay a penalty in terms of ad rank when you write ad copy that filters out some prospects, but it's often worth it to pay that penalty.
Run Better Tests
Go into the "edit campaign settings" area of your account, for any given campaign. The default method of rotating ads is set to Google's advantage. Google will start crowding out the ad that gets fewer clicks (a lower CTR) and start showing your high-CTR ad more often. That puts more money in its pocket, but it doesn't allow you to run the test evenly to measure through to your revenue stream to be sure which ad is truly the "better performer."
To ensure even ad rotation, uncheck the "show better-performing ads more often" box. Incidentally, the even ad rotation continues to be an underused feature of AdWords. You can use it to test more than ads. If you have two or three alternate landing pages, you can send your AdWords traffic to them equally from a given group of keywords just by setting up more than one ad for an ad group, with identical ad copy, and changing only the destination URLs associated with the ad.
If you have tagged each ad URL with unique tracking code so your analytics package knows what's going on (or are using Google's Conversion Tracker), you should be able to compare the conversion rates on these landing pages without having to use any fancy content management techniques on your site. An example would be testing the home page against a tailored landing page, or a category page showing a selection of choices against a page describing a single product in depth. Don't guess, test.
Run Smarter Reports
You don't need to sign up for Google's Analytics service, or buy Webtrends Enterprise, Omniture SiteCatalyst, or other high-powered analytics services, to get useful tracking of your clicks right through to a sale. Google's entry-level Conversion Tracker gives you plenty of information if you set it up right. It's fairly similar to most analytics software in that you have to install the correct JavaScript code on your site and sometimes customize things to pass through revenue data, if desired.
But what if you want to determine the return on investment on ad copy you're testing, as opposed to which keywords or groups of keywords are performing better? Many users aren't aware of how powerful the advanced reporting is in AdWords. Go to the Reports tab and, once on the Create Report page, select Text Ad Report. You can then select a time frame and limit the report to the campaigns you're interested in.
Finally, you need to customize the stats shown in the report. Conversion-related data isn't included by default, so you need to bring up the whole range of available report elements (under Advanced Options: Columns, further down the Create Report page) and then click some additional boxes to add them.
Run the report and view it either online or in your preferred downloadable format. Quite simply, if you've made enough sales off the groups of keywords in question, you should see comparative data showing exact cost-per-conversion numbers for ads you've been running head to head (no fancy tracking codes required on your destination URLs). If you determine that one ad is significantly better than another in its ultimate revenue generation effect, you've made the most important discovery you can make about ad performance, and next, you'll want to delete the nonperforming ad and perhaps run new tests. You'll also want to attempt to learn lessons from the test, which is easier if you've been testing certain elements of your copy based on industry theories and your own hypothesis.
Sometimes, these tests are amazing. I ran a test on one producta new ad competing with the oldfor the month of April 2006. Both generated about 600 clicks. The old ad generated 11 unique new customers at a cost of $9.73 per customer. The new ad generated zero transactions, so we're still waiting on that cost-per-acquisition figure as it is currently at "infinity"! 110...on 600 clicks! That's statistically significant to say the least. Goodbye to the new ad.
This helped us confirm that our call to action in the old ad, relating to free shipping and available inventory, was working well. Unfortunately, AdWords doesn't give me suggestions for how to explain to a client that I wasted his cash testing a "great new ad" that turned out to suck. Maybe we can chalk it up to "branding."
Control Your Ad Positions
You can now tell AdWordskeyword by keywordto show your ads only in certain positions in the advertising area of the search engine results page. This is available in Advanced Options under Edit Campaign Settings.
Let's say you want your ad to show up only if you can be no higher on the page than ad position 3 and no lower than 6. Enabling the ad position control provides a drop-down box in "keyword settings" for every keyword. The default is "any position," but if you say to show your ad no higher than third position, and no lower than sixth, Google simply doesn't show your ad if your ad rank forces you outside these positions. Be careful with this one.
Essentially, what it's telling the system is to keep your ads turned off unless your positioning criteria are met, so you could see sharply reduced ad impressions if you use the feature. This feature won't adjust your bid for you, so this is not "classic" bid-to-position functionality that is available with some third-party bid management tools. It looks like Google is testing the waters on this feature, including the revenue impact, in order to decide whether to implement a more robust set of bid management options.
For now, Google still has no bid-to-position feature and no day-parting feature. Advertisers seeking such exotica need to investigate third-party options or custom programming on their own through the AdWords API. Keep in mind, though, that the existing bid system is easy to use because it has an automatic bid discounter, and it subjects your ad delivery to whatever budgetary parameters you specify. Frequent (several times daily) bid changes are overrated, especially given that your ad rank is not solely dependent on your bid.
Don't Be a Slave to Automation
Google has been brilliant about automating editorial functions. If you're new to some of the legal and policy issues, the process can be daunting. One example is the automated ad copy checker that kicks in when you enter a new ad. Sometimes, the system identifies a misspelling when in fact you're using a niche term that's relevant to your region or industry.
Other times, a potential trademark violation comes up, such as using the word "enterprise" in your ad copy (which is the name of a rental car company). If you're not in the rental car business, it's highly unlikely that your use of a common word constitutes any kind of violation. You should politely appeal to Google in the box provided. Usually, your ad is approved within 48 hours.
Control Your Content Bidding
Some advanced advertisers might want to use custom programming to separate clicks from Google's "content partners" from clicks on the same keywords emanating from the search network. I prefer to use established analytics packages that do this. At the very least, you need to understand that most "content targeting" clicks are worth significantly less to the average advertiser than most clicks originating from a web search.
In Edit Campaign settings, make sure you enable "content bidding" if you have "content targeting" enabled. Then set different bids on content, ad group by ad group. If you're bidding around $.50 on a set of keywords, I'd recommend bidding $.15$.20 on content, or even less. The only way to know for sure is to use sophisticated analytics.
Control Your Geography
During campaign setup, you're offered "location targeting options." For existing campaigns, you can edit these settings. Specify countries and territories, regions and cities, or custom. Whatever you do, don't show your ad to the whole world unless that's your intention. Using custom geography, you can even target a radius of a few hundred miles, or a geographic "shape," if you're handy with the tools, which are now easier to use with a WYSIWYG-type interface on a map.
Boo-yah! This functionality works fairly well subject to the limitations of mapping users' IP addresses to exact locations. It's currently available in about 15 countries. Even national advertisers may find this highly useful, running campaigns in several dozen major metropolitan areas and adjusting their bids upwards in some of them to reflect the cities that are most responsive in terms of sales conversions.
A lot of these features can be a downright blast to use, especially for junkies. Don't forget your goals: acquiring new customers and communicating with them effectively and consistently. Ultimately, it's about customers, growth, and profitwhatever those concepts mean to you and your business. Many of the most advanced features have the effectwhen boiled downof causing you to buy too much media, or too little. Keep that in mind as you build and refine a campaign that runs consistently in an optimal range.
No comments:
Post a Comment