Google Ads: Most Frequently Asked Questions / Part 2

October 13, 2020
Posted in Paid Search
October 13, 2020 Jen Lopez

Google Ads: Most Frequently Asked Questions / Part 2

Here at Eboost Consulting, we’ve seen hundreds of Google Ads accounts – from ecommerce accounts to lead-gen accounts, both B2B and B2C.

If you’ve been following along in our Google Ads: Most Frequently Asked Questions series, you’ll remember Part 1 where we answered questions related to concepts about Brand, Generic and Conquesting campaigns, as well as the importance of account segmentation.

In Part 2, we’ll be discussing impression share, negative keywords and how to determine a clear ad copy winner.

Q: What is “Impression Share” and why doesn’t it add up to 100%?

A: Impression share has been a hot metric within Google Ads for a while, particularly since Average Position is no longer a metric recorded in the engine. Your impression share is the estimated percentage of impressions that your ads received compared to the total number of realistic impressions that your ad could have gotten. The way Google calculates the “realistic impressions” that your ad could have gotten is based on many factors like bid, quality score, targeting settings, approval statuses and more. Impression share can also be reviewed at multiple levels, like campaign, adgroup and keyword level for Search campaigns and at the campaign and product level for Shopping campaigns.

When trying to understand your impression share, it helps to review two additional metrics – Lost IS (impression share) due to Budget, and Lost IS due to Rank – both of which are columns available in your Google Ads reporting. These metrics will tell you if you’re missing out on impressions because of a limited budget, or because of your ad’s rank which can be partly determined by bid and quality score, or a combination of both.

So, why doesn’t my impression share ever reach 100%? Even in most brand campaigns, the impression share only ads up to about 96-98% and I haven’t personally seen 100% impression share on any campaign type. It is because this metric is an estimate by Google – and frankly, who estimates 100%?

We have managed a few campaigns where our goal was to manage to a certain impression share level, but, that can be a dangerous goal if it isn’t fully understood by all parties. One broad or competitive keyword can lower your IS significantly. In fact, when we have accounts with impression share levels of 30% on Generic campaigns, depending on the keyword portfolio – that may be entirely intentional. It is very expensive to “own” keywords on their own, much less an entire portfolio of them without BIG bucks, particularly if you are purchasing the keyword in large geos, like the US, versus just a small zip code. At any given time we can be testing a ton of Generic keywords that we wouldn’t want to work toward “owning” unless they were surefire winners and we had the budget to support that endeavor.

And so, we recommend using impression share to help diagnose issues, rather that work to achieve the highest possible impression share when looking at a full keyword portfolio. For example, if your impression share is dwindling over time – what’s going on? Are more advertisers entering the auction with higher bids? Is search volume picking up on these terms beyond what you expected making your budget too slim? These are some smart questions to be asking if you notice impression share suffering.

If you do have a handful (and I mean only a few) keywords where impression share is really important, consider narrowing the settings, like having strict geo-target, and only targeting certain audiences and demographics if you’re allowed to go that route – as it will help make that impression share level be more attainable and targeted. You can also test the Target Impression Share bid algorithm with Google Ads that will aim to help optimize your bids in real-time in an attempt to hit the target IS you set.

Q: What are negative keywords and how often should they be added to an account?

A: Depending on how your Google Ads account is managed, you likely have a few keyword match types within your account (hopefully segmented into different campaigns or adgroups at the very least). The goal of using varied match types is to help balance efficiency with scale. We typically use exact match keywords for efficiency, but rely on other match types like Phrase, Broad or Broad Match Modified for scale. Outside of Exact (save for Exact Match Variants), the other keyword match types all allow for some freedom with how keywords are shown or interpreted by Google, which could mean that unintentional search terms are triggering your ads to show.

For example, if we are a restaurant open for the holidays and wanted to show an ad about our delicious Thanksgiving Turkeys, we may bid on the term +Thanksgiving +turkey (which is Broad Match Modified). With this match type, the keywords Thanksgiving and turkey will both be required to be in the search term for your ad to trigger, but unless you have negatives, that’s really the only rule. So, you can expect your ad to serve on searches for “kids thanksgiving turkey crafts” or “homemade thanksgiving turkey recipe” or “how long to defrost a thanksgiving turkey” or even “why does eating thanksgiving turkey make you tired.” Yep, all those search terms could trigger one of your ads, and so, what we do as PPC Manager is before we launch a campaign, we come up with a list of all the keywords we DON’T want to show for – what we don’t want to be associated with. In this case, we might add: recipes, crafts, kids, tired, sleepy, chemicals, cooking times, how long, defrost, temperature, etc.

While you can add as many negatives as you’d like up front, there’s always going to be something you haven’t accounted for – and, clients should really give their PPC managers a break on this one! Even Google says, that 15% of the trillion of searches they receive per year have never been searched before. Because of this, PPC Managers have no way of coming up with every possible negative from the onset. Instead, we often pull Search Term Reports to review the actual search terms matching to keywords, and add negative keywords based on that real data. This is typically done weekly (depending on how large or how new the account is) and is reviewed over the last 7 days. Over time, we review that Search Terms report over longer and longer periods of time though, to catch keywords that may only trigger a few times and not appear significant within a 7 day view, even though they may add up to be. However, this may start to change with the recent news of Google Ads limiting their Search Terms Report, creating a new blindspot for PPC professionals.

Q: Ad testing is something we definitely want to do. But, how do you determine a winner?

A: This one we get a lot. Ad testing is always a hot topic, and rightfully so. While Google Ads (even when it was AdWords) has long had features like Experiments to help determine which ads are more effective, it is hard to say with certainty if you don’t know what to measure. Many PPC pros want to measure the CTR. After all, we are looking at the ad’s effectiveness, and if more people are clicking on one versus another, shouldn’t that ad be the winner? Well, some might disagree if the other ad has a higher conversion rate. A client might think, well I care about conversions more, so even if the CTR is higher on that ad, this one is converting at a better rate. To get around that tricky conundrum, we began using Brad Geddes definition of success here, which is “impressions until conversion,” as it neatly blends the ideas of CTR and conversion rate. So, once you know the amount of impressions until conversion, then what? It is fairly easy to pull up a free significance calculator online, which I often use. But, don’t be afraid to go with your gut if significance is taking too long. We’ve learned that tests which take months to prove significance because of low volume are likely a huge waste of time in the grand scheme of managing the account, and there are likely other matters that can have a much more immediate impact. You should strive to look for high impact tests, where you can learn quickly and apply changes to multiple parts of an account, when it makes sense.

If you have any questions, or have a particular PPC question you’d like answered in our next series, feel free to comment below.

We will see you again next month for Part 3 of our Google Ads: Most Frequently Asked Questions series.