Skip to content

Feeding the Google Ads robot: the future of ad management

feeding the google ads robot

Here are a few phrases you’ve probably heard lately, ad nauseam

Performance Max. Conversational UI. Generative AI. Privacy. GA4. ML. AI. More AI. Even more AI.

Whether you’re brand new to Google Ads or you’ve been staring at that plain gray interface for decades, it changes so much that we can never truly be 100% certain when it comes to tactics, focus, and the proper roadmap for your accounts.

As I mentioned in an SMX Advanced panel on Google Marketing Live this week, there are some changes coming in the Google Ads ecosystem. Not just any small changes: a tsunami of changes. 

Google Ads is evolving rapidly, and it’s more important than ever to keep track of the changes, and how they can be used to help improve campaign performance.

But first, let’s look at how Google Ads used to work (like, way back in 2021).

A few years ago, Google Ads was about control

Extensive keyword research, optimizing for devices, manual CPC bids, exact match, huge negative keyword lists, single-topic ad groups, A/B testing different ETAs, then A/B testing different RSAs, audience and demographic targeting, et cetera – all these things could turn a campaign into a profit explosion.

And it manifested very clearly. A few years ago, since Google Ads was so receptive to granular control, we could easily take over a faltering ad account and turn it around in a matter of weeks. We would dial in, make very specific changes, and see it turn around immediately.

And these campaigns needed constant babying. A few years ago, if you set up a campaign and let it run without any adjustments, there were only two options: if you were lucky, it would stay about the same and never improve, or more likely it would slowly just fade away into oblivion.

Now, Google Ads is about data

One of the biggest changes we’ve seen is that campaigns can now actually perform better over time…and trying to tweak too much can actually damage performance.

If all your other ducks are in a row (an important if) we have seen that campaigns can improve, especially if you’re not trying to over-optimize. As long as your website is performing well, your creative/branding is on point, your pricing, product quality, and customer service is excellent…your campaigns will tend to perform even better.

And sometimes, if you leave campaigns alone long enough, they get way better. The patience required to wait the 1, 2, or even 3 months to start seeing better performance is a barrier that’s very difficult to accept. And hard for small advertisers to afford. But it’s there.

Changes to match types have already hit (we used to never use broad match…now we do use broad match) and that’s only going to continue.

And I wouldn’t even be surprised if the keyword goes away completely in a year or so (as much as it pains me to say it).

Instead, the algorithm vacuums up the data and our job is soon going to be shepherding that data in order to curate the best results (sort of like knowing how to write the prompt to generate a specific image in Midjourney, like the pink robot at the beginning of the article).

Google is now encouraging users to upload their own customer data to help the campaigns understand your audience better (and possibly exclude those same users from campaigns which aim to acquire brand new customers). E-commerce stores naturally collect a lot of this data. But it’s important in B2B campaigns as well…this means hooking up your CRM with Google Ads in order to pass important conversion data through.

Here is how we have begun working in this new data-driven approach to campaign management:

The power of creative assets

Pretty much all of our ecommerce clients are now using PMax. Lately, we’ve been setting it up right away (even with brand new accounts). In many ways, it’s replaced Standard Shopping for us. 

We only do this if we have really good images and media we can use for the client. Without great creative assets to pull from, PMax campaigns rarely work super well. But not all clients have great creative assets…it’s expensive to produce commercials and do product shoots and edit together a variety of videos.

In the near future, for those less-creatively-endowed accounts, we’ll be able to generate these assets within Google Ads.

Would you like a luxury female for a voiceover?

These additions to generated creative assets allow us to move our focus to strategies and tactics that really move the needle even more. Instead of spending most of our time worrying about writing 8 unique headlines as Google used to want us to do for an RSA, we’ll just be able to use Conversational Experience to plug in the client’s URL and some general descriptions, and hit “go”. Some quick editing and style curation, and we’ll be good to go.

Using Product Studio, we’ll be able to take lackluster client product shots and generate cool backgrounds to use in their campaigns:

Single-handedly putting all those Fiverr Photoshop experts out of business.

And with Demand Gen campaigns, we can even easily deploy branding/awareness campaigns that focus on visuals in a social media-esque format:

I thought we were supposed to be banning TikTok, not cloning it.

Automated product feeds

One of the most time-consuming portions of our work with ecommerce brands is product feed management. It’s a beast, to say the least.

Everything you thought you knew about Merchant Center is now out the window. Merchant Center Next is coming as soon as next year, and will scrape data directly from your website instead of depending upon a finicky, fragile, and overly complex data feed that never seems to work just how you want it.

Now do we have to use a dozen different feed management tools?

A common thread in these tools? It’s less about the individual control of ads. It’s more about letting the data do it’s thing, and using this data to serve the best creative possible.

The problem? It doesn’t always work. Despite automation apologists claiming it’s a magic bullet, it is not. Anyone who says it is a magic bullet does not have breadth of experience in various niches or geographies.

But it may get there one day.

How I stopped worrying and learned to love automation

That’s an exaggeration. I’m still skeptical of many recent ML advancements and I do still worry that there is a lot of unknowns out there.

But at the end of the day, our job as PPC specialists is to make our clients more money.

And if they make more money embracing the recent changes, well, that’s what we have to do.

I’m deciding to let go.

Stop worrying.

And hope it’s not a bomb.

Leave a Reply

Your email address will not be published. Required fields are marked *