How can one apply North American search marketing insights to a very small market in Europe? I was reminded of this question when I recently spoke at a conference in Norway.
Norway’s e-commerce market is a slim 1 percent of that in the United States. I sometimes assume I instinctively “get” the small market equation, being from Canada. But in fact, our market is not particularly small, not even counting the fact that some Canadian companies have U.S. divisions.
Canada’s digital universe, by contrast with Norway’s, is 10 to 15 percent of that in the United States. Despite some shortcomings in the adoption of e-commerce best practices, Canada is often cited in the top three to five nations globally for broadband penetration, digital savvy, and things like Facebook usage per capita. (Cory Doctorow is Canadian. So is the founder of Friendster. Need I say more?) Among other things, Canada has sizeable operations from companies like Amazon, eBay, Google, Sears, Rogers-owned The Shopping Channel, and The Gap operating within its borders. Canada’s population is 34.2 million; Norway’s is 4.9 million.
In terms of logistics and scale, Canada is much more urbanized, with over 20 million people living in or near metropolitan areas of 500,000 people all the way up to 5 million. Norway’s fifth-largest city has only 80,000 people. There are no large cities in Norway; breathtaking, historic Oslo is medium-sized at 600,000 inhabitants.
Here’s a reminder of why this is such a huge issue: the hallmark of great paid search campaign management is data, feedback, and nimble testing to ROI-based metrics. Feedback cycles are much faster when there is more sales conversion data to work with. It’s going to take at least 10 times longer for a test to become statistically significant if the market is 10 percent the size of the larger market. (A qualified statistician can show you it might even be longer than that.) Allowing for certain adjustments and lessened competition, a test that might take one month to conclude in the U.S. market might take six months in Canada, and two to three years in Norway. Unless you do something differently, you can kiss that idea of rapid feedback goodbye.
No, you can’t manage these campaigns the same way. So how do you manage them at all?
It’s important to try. Wasting money on advertising sucks. This can divert scarce cash from places it might be better used.
In “Why We Make Mistakes,” a well-respected book on human cognition by Joseph T. Hallinan, we learn that most people are dangerously overconfident because they are poorly “calibrated.” Calibration means the relationship between reality and self-perception. In the mass media advertising industry à la Don Draper, people thought they were doing great because they didn’t close the data loop. Accountability was all by feel. By contrast, TV weathermen are well-calibrated. If their predictions aren’t accurate, viewers’ lives are affected, and there are complaints. The profitability (or even safety) associated with good calibration forces professionals to figure out how to use data to provide better feedback on how they’re doing, so they aren’t dangerously or recklessly overconfident.
Clients expect paid search professionals to be plugged into actionable data. Does working in a small market let you off the hook? No way! Here are 10 ways you can get better results, even working within the limitations of a market that provides you with less feedback.
- No whinging! Also known as whining, moaning, and complaining. No one is immune to solving small market challenges. Data can be limited in local campaigns, B2B campaigns, and small startups, even in the biggest markets in the world. So we all need strategies to deal with this problem.
- Best practices first. Where there is limited data, first get the fundamentals right. This includes eliminating jargon and difficult-to-scan abbreviations from ads, writing compelling headlines, and focusing on good calls to action and your company’s unique positioning.
- Moderate granularity. It might not be a good idea to chop campaigns up into too many little pieces. If you do so, you’re going on faith that this will perform better because it will take even longer to gather click and conversion data for those ad groups. Granularity is good; too much is making life into a guessing game.
- Experience matters, so get help. Kickstart your keyword selection efforts by asking for feedback from people who have managed many campaigns. Attend clinic sessions at conferences, or seek a second pair of eyes willing to do a campaign audit.
- Use broader matching options. Narrower match types starve you of data in small markets. You may get less out of phrase match (my personal favorite) than you should. Try instead using broad matching and check out the search keywords report to see what actual phrases it’s matching to effectively. Don’t forget to exclude wildly inappropriate matches as they come up. Better yet, try the new modified broad match in Google AdWords.
- Try enhanced CPC. This scaled-down version of Google’s Conversion Optimizer tries to bid predictively to aim your spending towards the best performing keywords, ads, and overall situations (such as time of day, geography, and other user factors). Consider this as a last resort. This is available at the Campaign Settings level in Google AdWords.
- Take advantage of your context! If you’re a home-grown company, make sure everyone knows it. The country TLD (top-level domain), localized messaging (“Italy’s favorite”), local shipping details, and local dialects can make you an attractive choice to consumers wondering who is really going to bat for them. This can work even in a mid-sized market such as Canada, where consumers are wary of high-cost, cross-border shipping.
- Adjust expectations. If your manager or your client is asking you to create more complex tests, they may be extending the life of those tests indefinitely. You might be forced to explain the nature of statistically significant feedback. Rotating more ads and more landing page permutations isn’t going to do you any good if you don’t have enough volume to work with.
- Look for trends across accounts. For example, attempt to merge the kinds of feedback you receive across multiple languages, especially when all indications are that similar triggers are working in different languages. Look for common threads such as ad copy and landing pages that convert better across many different campaigns and ad groups, and extrapolate those trends to other parts of the account. If you wait for every individual ad group to reach a statistically significant conclusion independently, you may never get there.
- Soft conversion events are a must. Don’t always wait for a revenue conversion event to make bidding and other account management decisions. Deliberately set other goals in your analytics, such as reaching a check-out page, or clicking on the Dealer Locator. Watch user engagement metrics, such as good old bounce rates by segments such as keyword.
You may still be flying in a light fog at this point, but at least you won’t be flying blind! Good luck.