Google Ads is so complicated that it often leads to incorrect assumptions as to how certain things work. When one authority site writes about one of these features incorrectly, it is often considered true and the misinformation quickly spreads and many marketers end up with poor information.
In today’s column, we’re going to look at 5 common pieces of misinformation we hear on a regular basis and set the record straight as to how you should evaluate and think about some of these incorrect assumptions.
Whenever you see some truly terrible search terms in your account, it’s either a result of Google not understanding your exact match keywords properly or the usage of true broad match.
While broad match deservedly gets a bad rap, there are times it’s very useful.
The most understood usage of broad match is when you have search terms that are in two different languages at once. We see in many countries search terms that are in English and Arabic, German and English, etc. When a search term is in multiple languages, the best way to target them is with broad match. For example, here’s a small sampling of a flower shop’s search terms that are using broad match to capture multi-lingual queries.
Broad match has a lot of good uses. It might not be your top converting keyword, but that doesn’t mean it can’t perform well for you. If you are using broad match, you will show for a lot of random queries, so you’ll want to ensure you are using n-grams to analyze your search terms and make that analysis part of your regular workflow.
Google uses a generalized second price auction that is based upon the Vickrey Auction model. In a true second price auction, you pay just above the next bidder. However, Google doesn’t just use your bid to determine your position. Google uses a mixture of your bid, quality score, and ad extension’s impact on CTR. These three factors make up ad rank.
The ad extensions are tricky, so let’s focus on the older model that just uses bid and quality score to determine ad rank so we can illustrate this point.
|Advertiser||Bid||Quality Score||Ad Rank|
In this example, the top position ad will go to advertiser 1 as they have the highest ad rank. As their max bid is much less than the next highest ad rank (advertiser 2), they can’t pay $0.01 more than their bid as that would have them paying a number higher than their max CPC. What they need to pay is $0.01 more than would be required for them to have a higher ad rank (8.1 ad rank).
In this case, if advertiser 1’s bid were $0.80, they would tie for ad rank with advertiser 2 with an 8 ad rank. Therefore, we add $0.01 to their bid to get $0.81, and this is what they end up paying for that click as it’s just above the ad rank of the advertiser’s below them.
You don’t pay $0.01 more than the next person’s bid. You pay $0.01 more than the ad rank you need to beat the next bid. This is why Quality Score optimization is just as important as bidding.
When modified broad match was introduced several years ago, the result over the next couple of years was a huge decline in phrase match usage. Modified broad generally performed just a bit worse, from a performance standpoint, than phrase match. However, modified broad match would match to many more search terms due to word order not being used in determining the matching. Advertisers would use exact match to capture their best terms and modified broad to capture the rest. If there were some matching issues, then negative keywords would usually clean up those poor search terms.
While phrase match looked nearly dead a year ago, the change to exact match caused a large resurgence in phrase match usage for advertiser’s who found out Google didn’t have any understanding of their industry.
For example, if you have the exact match term “Atlanta 2 for 1 tickets” as you are selling event tickets in Atlanta and offer a 2 for 1 special. Google’s new exact match will match this to “Atlanta 2 for 1” (dropping the word tickets). 2 for 1 (without the word tickets) is most commonly searched for people looking for 2 for 1 dinner specials. You just went from selling tickets to showing ads on irrelevant terms about restaurant specials. Deck paint and paint deck aren’t the same things when the word order is reversed. The worst issues are when you have multiple matching keywords in the account and Google picks the incorrect one with the new matching formulas.
With phrase match, Google matches your actual words, not the perceived intent of these words. If you are in one of these industries Google doesn’t understand, phrase match has become your go to match type. If you are still struggling with the change, here are some strategies that can help.
The bane of many advertisers is mobile devices. Some companies are doing better on mobile than on desktop. Others find they can’t get a conversion on mobile no matter what they do.
There are usually a few reasons someone is seeing poor mobile results.
Apps. Google used to make it easy to block your site from showing on apps, as most non-app marketers do not do well showing their ads on a random app game where many clicks are just accidental. Last year, Google retired adsenseformobileapps.com as being an easy negative placement to stop your ads from showing on apps. Now, you need to manage these closely and make sure you are using the new system to block app categories (even if you block them all, you’ll still get app impressions).
Complex Interfaces: When a user needs to use a complex interface, such as designing logo t-shirts, configuring SAAS software trials, or laying out how your new furniture will look in your living room using online designs, these companies usually do poor on mobile devices as the screen size is just too small to get a good feel for how everything fits together and to follow complex instructions. These companies need to use mobile as awareness and contact and desktops for conversions. If you are selling complicated software, especially ones that need configuration, then you need to use an attribution model to judge mobile.
Mobile research & computer/in-person buying: It’s still very common for someone to do their research on a phone (or showroom) and then buy through a different device or in-person. Even if you have an app, its often just easier to see options on a desktop. To understand mobile’s value in these instances, you also need to look at attribution models to properly determine the value of mobile.
Mobile isn’t going away. However, there will also always be a place for a larger interface. That might be a phone projector or VR glasses someday, but for today, that’s looking at how mobile and desktop devices play together as focusing on your attribution model so you are setting proper bid modifiers.
Automation is great when it works. Automation is terrible when it doesn’t work as it’s easy to miss problems that you’re not looking for due to automation.
Here are some common automation failures:
We love automation. However, humans are much better at ad testing, creating ads, employing new strategies, and auditing the machine than a machine is. Machines are great at managing large data sets, repeatable work (reports, bidding), and showing humans where there’s an anomaly so a human can use its strategic thinking to make the proper decision.
Humans and computers both have their strengths and weaknesses:
Markets won’t lose their jobs to machines, their jobs will just evolve into working with machines.
This is one of our focus points. Have the machine do the behind the scenes data crunching and recommendations, and then give advertisers the tools to easily evaluate and apply recommendations. If that sounds interesting to you, just check out Adalysis’s features.