In another blow to advertisers, Google decided in September to limit the amount of query data they would share with advertisers. There are many problems with this change, and much has been written about it, so I don’t want to rehash all the problems like some companies losing 80%+ of their query data.
After being asked many times all the reasons I thought this was happening, I thought I’d through out all my theories at once and get down to what this means in the long term for advertisers.
I Don’t Believe it has Anything to do with Privacy
First off, Google said the change was due to privacy considerations. I think that’s a convenient excuse from Google. I remember back in 2005 having long conversations with Marissa Mayer, Emily White, Cheryl Sandberg, and Larry Page about privacy data. We were managing a lot of accounts (over 40,000) as well as feeding Google a tremendous amount of local business data and wanted to know about privacy.
Google walked us through many filters and layers of privacy protecting data they had at the time and assured us there wasn’t a privacy issue with data. In fact, that’s why you would often have 1%-5% of your search terms marked as ‘other’. The other was a catch-all category of very low impression data or data Google didn’t want to share in the name of privacy. That was 15 years ago, and I don’t think Google suddenly had a privacy problem after continuing to work on this issue for well over a decade.
I recently looked deep into an account where Google is now hiding a significant amount of the search terms (where the account spends over $5 million/month) and when we look back over the past year, there were no queries that were compromising privacy. So Google used one of their convenient labels of privacy (just like they love smart or enhanced as a label that means you have less control) and took away a lot of data.
Ok, so if it’s not about privacy, what is it about?
I don’t believe in all these theories, but these are the ones that spring to mind the fastest.
Google’s Hiding How Bad their Variant Matches Are
I tried to maintain a list of the worst offenders to Google’s variant match. The list quickly grew too long to maintain. I think Google’s machine learning when it comes to data is amazing. It sometimes takes 1-3 years for the machine to get enough data to do a good job (CPA bidding, ROAS bidding), but it gets there.
I can’t say the same for Google’s semantic matching. When it comes to anything language based, their matching is terrible. We haven’t seen variant matching getting any better. Google still has issues putting together RSAs based upon the proper ad formating. Even their DSAs and automatic ads are just pulling headlines off your website and not coming up with anything new.
While this might not be the entire reason for the change, it is a good way for Google to hide how bad their query data has gotten with the variant matching options due to a weakness in their machine learning when it comes to languages.
Admittedly, this does come from someone who also thought Google’s organic results 5 years ago were better than today. On occasion, I’ll track how often I find what I’m looking for when searching, and my percentage of successful searches is about 30% lower than 5 years ago and I find that many more completely irrelevant sites show up in my searches now because Google’s tries to figure out the intent of your term instead of just showing you what you searched for. Since their intent matching can be so bad, you just get a lot of junk search results.
Google Weakens 3rd Party Bid Systems
If you can’t get access to your data, you can’t set good bids. You can’t figure out what queries are spending money to make them keywords. You can’t do full n-gram analysis to make negative keywords. You can’t tag your query data to follow it through to conversions in your CMS system and close your bidding to the conversion loop.
Google just removed some of the most valuable bid data that exists. This means your homegrown bid system that mixes your custom data mixed with query data that was beating Google’s bids for your account is going to run into some issues as it starves for more data.
If no one but Google has the data, it strengthens Google’s argument that their machine learning system for bidding is better than anything else. Of course, it will be better than other systems if no one else has access to the same level of data. We’ve seen many 3rd party systems that use query data really struggle to figure out what to do.
If you are a conspiracy theorist, this is probably the theory you jumped to first. If you can’t see any data, then Google can easily manipulate the auction. The more auction control they have, the more they control the ads and bids, the more money they make.
This is also connected to using Google’s bid tech by weakening other systems. If 3rd party systems aren’t as good, then you move more to Google’s systems. If more people use Google’s bid tech, then Google has more auction control. We could go through a lot here with the more control they have over ad serving (ad rotation), budgets (the massive change to doubling your daily budget), ads (RSAs), and bids (hiding all data for you to do this yourself), the more auction control they have and thus the more CPC manipulation they can perform.
This is a great way for the fox to rule the henhouse.
Less Customer Support for Google
Google spends a lot of time and money on customer support. Think about this, if you have no data and you can’t change any settings, what does that do to Google’s customer support demands? They decrease. The move to smart shopping, smart bidding, less query data, variant matches, etc means more of the system is automated and you can’t do anything about it. If you can’t do anything about it, then you don’t need to talk to Google about your options. They might save a lot of money with their changes in the past few years by lessening customer support needs.
Make Microsoft Ads Weaker
How does Google removing data make Microsoft weaker you ask? Let’s look at how many advertisers manage Microsoft.
- They make a Google Ads account
- They manage their Google Ads account creating more keywords from query data
- They sync their Google Ads account with Microsoft
Microsoft has less query data than Google since it has less search usage. This means you’re less likely to spend a lot of time in Microsoft Ads and rely on the sync between the two systems. If Google stops giving you data to add keywords, negative keywords, and manage your account properly, then there’s less data to share and sync with Microsoft, essentially weakening your Microsoft account.
This theory definitely leans more into the conspiracy theory category and probably wasn’t a big reason this change was made, but it is an interesting way of looking at the world.
Google Just Continues to Weaken Large Company Advertising
When you look at the major changes the past few years, match types, query data, etc, they hurt the large sophisticated advertise the most.
We still see people struggling to overcome Google’s variant matching issues. Yes, it’s been a couple of years since the changes to was made. However, when you spend millions a month and have to completely redo your structure due to how poor some of the matching is, you don’t have time to dig into missing queries yet. By the time they can dig into missing queries, they won’t have any data left since Google is now hiding the data.
It’s strange, but it seems Google has been on a mission to help small companies and hurt large ones in the last few years. We’ve seen finance, insurance, and travel companies (some of the highest spending industries) really struggle with Google Ads more than any other industries (outside of those facing regulation changes) the past few years.
It had been about a year since Google made a change that hurt the large advertisers, this was just their annual way of messing with their best clients.
So, What’s Really the Reason?
No one outside of Google really knows, but if I had to bet, I’d put my first bet on auction control and my second bet on Google hiding weakness in their semantic machine learning engine.
Regardless of the actual reason, it will eventually lead to more people using automated bidding. This also means you have to use optimize ad serving, which can be terrible, so Google has more auction control. With more ability to control the bid and ad serving process, Google can make more money from each auction.
Controlling the auction isn’t always a bad thing. If Google believes that only 3 advertisers could get a conversion from a specific search, then with more auction control, they can make sure those are the top 3 advertisers and then just put the other advertisers in the auction to keep the minimum bids up. This could be seen as a long term efficiency in the auction process to maximize the total possible amount of conversions advertisers can pull from any auction. So while this is a good conspiracy theory, there’s also some good business sense behind this theory.
Of course, this also assumes that Google knows everything about the advertiser’s and their goals. We know that’s not always the case now, but it’s how Google wants the world to work.
For now, the only thing you can do is make sure you have maxed out your spend on Microsoft Ads or other places that will give you the data. Then you can take that data to use to optimize your Google Ads account.
Years ago, we use to rely on our Google Ads data to optimize other advertising channels. Increasingly, we’re using data from Microsoft, Facebook, and other platforms to optimize Google Ads.
As Google continues to hide data and force your account to show irrelevant impressions, we expect this trend to continue. As you’re using other channels, think about how you can use that data to ensure you are properly adding new keywords, audiences, and negative keywords to your Google Ads data.
Google will continue to expand your impressions and give you less data. That’s been something they have been focused on in the past few years. Ensuring your money is well spent is your job. Make sure as you examine each channel individually that you are thinking about how the data from that channel can be used to effectively manage and optimize your other advertising channels.
As data is taken away, this will be a key for smart advertisers – ensuring that the data you do have, you are using to its fullest potential across all your advertising channels.