Google’s AdWords platform is one of the most popular ways to run ads across a huge display network, and with good reason. The barrier to entry is low, costs are manageable, and the wide variety of possible positions means there’s always somewhere your ad can be displayed.
With any large ad network, one of your primary concerns is going to be cost. You want to keep costs low without dropping into “penny click” spam territory. You want to keep quality high without paying 3x or more what the clicks are actually worth. What do you do to strike that balance?
1. Create many variations on your ads.
Ad variations in ad sets can help a lot with keeping costs down. Instead of targeting one massive audience with one ad that may or may not be very good, break your audience down. Target different segments of your audience with different, more tailored ads.
For instance, if you’re running a shoe store, you don’t want to target everyone looking for shoes with a generic ad for shoes. Break it down. Target people looking for dress shoes with an ad about dress shoes. Target people looking for heels with an ad for your best-selling heels. Target people looking for athletic shoes with your best-reviewed sports trainers. Taken all together, you’ll have a greater conversion rate than you would with one larger, generic ad.
2. Ignore the Google Keyword Tool.
The best way to target individual segments of an audience is to target the actionable keywords they’re using. Each ad should have a selection of keywords relating to that segment. For example, if you’re trying to sell athletic shoes, you might use keywords like “women’s athletic shoes” or “athletic shoes for women” or “cheap athletic shoes.” Using brand names, colors and genders can help segment your ad to the right kind of people. You can even target out of the way ads for specific conditions, like “best athletic shoes for plantar fasciitis.” It’s not a keyword with high volume, but if you can capitalize on that niche, it will work wonderfully for you.
So why do I say to ignore the Google keyword tool? Let me count the ways. Many long-tail keywords have no data available, even if they’re good keywords. Search volume is an estimate only and is inaccurate much of the time. Suggested pricing can be completely wrong. Google even ignores some incredibly popular keywords entirely.
What should you use instead? There are a bunch of different keyword planners and keyword suggestion engines. Use those, plus your own knowledge of what people are searching for when they find your site. Then test the keywords yourself.
3. Set up conversion tracking.
To successfully test keywords, you need a lot of different ads, and you need data that is tracked on a per-ad basis. If you’re running 10 ads and get $250 sales in a week, that’s cool. Which ad sent which sales? You have no idea. If two of those ads together sent $200 of those $250 sales, you would never know you’re paying for eight ads that are doing nothing for you.
Thankfully, AdWords does have worthwhile conversion tracking that is simple to set up and will funnel accurate data directly to your Google Analytics. You can read about how to implement it here.
4. Trim the fat.
Part of successful optimization, and thus successful cost-cutting, is to trim the fat. Run a lot of different ads, but run them for only a couple of weeks at a time. Two weeks, three weeks, a month; that’s how long it takes to gather reasonably accurate data for most keywords.
I say “most” keywords, because some keywords are going to be seasonal. People aren’t going to be shopping for Christmas ornaments in June, so running an ad for them isn’t going to give you accurate data for December.
Once you have a few weeks or a month’s worth of data, you can compare ads to one another. Once again; if you’re running 10 ads and only two of them are performing adequately, you can cut out those other eight. Some might be worth optimizing and salvaging, others will be worthless. Trim the fat, and keep records to see what you’ve done and avoid making the same mistakes over again.
If you have a large enough budget, you might consider letting Google optimize ads for conversions themselves. Google will do their best to get you as many conversions as possible, but they’ll spend what they need to do it. Once you’ve run with Google’s optimization for a few weeks, then you can go in and kill the bad keywords, optimize the good ones, and manually adjust your CPC. Sure, it burns some money, but it automates a process you would otherwise have to spend weeks doing manually.