Navigating Google’s Double-edge Optiscore Sword: The Best and Worst of Their Recommendations

I’ve been working in Google Ads since 2015, so I’ve seen many new Google innovations in the space. The shift away from manual bidding to Google’s (IMO – far superior in many situations) automated bidding was landscape changing.

In 2019, Google introduced its optiscore that was essentially a ranking system to determine how “compliant” and healthy your campaigns were based on Google’s weighted optimizations. If you were at 100%, you applied most recommendations by Google, or (do with this as you will) just clicking to “dismiss” them. I don’t suggest this without at least taking the recommendation into consideration, as many are actually very helpful.

Some of Google’s recommendations that impact your Optiscore (weighted from 0 – 100%) include:

• Raise Your Budget: Seems pretty simple right?
• Add Dynamic Image Extensions: Google offers image extensions for Search ads and will suggest you add these to your search ads.
• Remove Conflicting Negative Keywords: Google will let you know if a negative keyword you’ve added is affecting another keyword in your campaign.
• Improve Your Responsive Search Ads: Gives you more headlines and descriptions to bolster your ads.

Having worked with Google’s Optiscore since the beginning, I’ve found that many recommendations are beneficial for your campaigns, but many require a lot more caution and review before implementing. This blog post reviews the Search/Shopping recommendations that I found to be the most helpful in terms of improving the possible outcome of your campaigns, as well as the recommendations I would evaluate with extreme caution.

Before I continue, I’d like to express that, like the Optiscore, I view it as it is: “recommendations.” They aren’t the “end all be all,” and Google doesn’t know your campaigns as well as you do, but they do help PPC-ers go the extra mile with a few campaign areas.

Go To Recommendations:

Remove Conflicting Keywords: I like this one because when working with possibly 5-10 campaigns and upwards of 50-100 keywords, there can be some overlap and what might be worthwhile to exclude in one campaign/ad group could cause conflicts with other campaigns. Review this when you see it in your recommendations tab because you could have a negative keyword that is affecting one of your best performing ad groups. Here is a good example:

Excluding the term “auto” from your campaign when one of your keywords is “Local Costco auto center.”  You may have excluded this accidentally, or you excluded it not realizing its effect, but Google will catch this for you if you excluded “auto” in this situation.

Improve Your Responsive Ads: I find this one very helpful because, sometimes, my mind goes blank, and I am unable to think of anything additional to add to my responsive search ads. I usually try to add at least 5 headlines and 2 descriptions per ad, and this is where Google comes into great use. I’ve had it create 1 or 2 new descriptions or headlines for my ads that, wait for it…actually make sense.

I highly suggest you review the improvements that Google suggests you make for your RSAs. If you’re skeptical about its benefits, make some benchmark notes for your RSAs before and after you applied Google’s recommendations.

Add New Keywords: A simple recommendation by Google: just add new keywords. Oftentimes, when building out campaigns, you think you’ve grabbed every possible keyword. Google’s recommendation will tell you what keywords it deems you’ve missed that searchers are searching.

When BMM was more popular, this was not really a concern because Broad Match Modified allowed for broad search queries to be triggered by your low character keywords. With Phrase Match now being the non-Exact match norm, there is more grey area on what Phrase Match covers. Review this and see if you already have any keywords that cover these recommended keywords.

Cautious Recommendations:

Upgrade Your Existing Keywords to Broad Match: I don’t run Broad Match, I never have and I doubt I ever will. While I am not saying “Do not run Broad Match” keywords, I would as strongly as I can say:  be cautious before you just click “apply” on anything that includes “broad match.” I have not tested Google’s algorithm to see how well they match search queries to your broad match keywords but I’ve seen enough accounts get burned by Broad Match randomness in the past.

The current way that Google recommends you complete this swap is to take your (possibly) high performing keywords and change them to Broad Match. This is dangerous for 2 reasons:

1. You’re basically ignoring the strong performance you’ve had on those keywords with Phrase Match
2. You have no clue how these keywords will perform and are entering a PPC’ers worst nightmare; the unknown.

If you would like to try Broad Match keywords, run an experiment by clicking “Apply as an Experiment” on the recommendation:

Increase Your Budget: This seems pretty self-explanatory but I do not click “Apply” without making sure that these changes fit into your or your client’s budget. Google does not know how much your client wants to spend monthly, quarterly or annually and hitting “Apply” could have disastrous effects.

Adjust Your ROAS Targets: I can’t deny that Google has some pretty excellent automation when it comes to ROAS bidding. Some campaigns that had below a 100% ROAS increased to 300% or 400% just by switching to Google’s automated bidding. What I want people to be cautious about is just agreeing to whatever ROAS target Google would like you to lower/or raise.

Occasionally, they will suggest you decrease your ROAS bid, with the trade off being more conversion value (revenue) gained. The problem is, often the cost of doing these changes outways the improved conversion value. Take for example this recommendation by Google:

It wants me to lower my target ROAS by 17% and, if I do that, I will see a $146 increase in conversion value. I’m sure you see the problem, but it’s saying the weekly cost increase will go up by MORE than the increase in conversion value. Always review the details before clicking “apply.”

I wanted to end this blog by saying I am not “anti-Optiscore” and in fact I actually appreciate most of the recommendations it provides. I’m just a realist and caution you to make data-centric decisions based on any recommendations placed on your lap. If you feel “this is the right move for me to apply this recommendation” I am glad – but I want it to be one that you will not regret!

May your Optiscore be high and your ROAS even higher!