Inside Google Ads podcast: Episode 106 - Performance Drops
Is your Google Ads performance really dropping or are you just panicking?
Because we all know that feeling. You log into Google Ads, you see those red arrows pointing down next to your conversions or ROAS and your stomach drops.
How do you know if you actually need to overhaul your campaign or if you just need to close your laptop and go for a walk?
I'm going to show you how to tell the difference between a real problem and a normal performance fluctuation, and exactly what metrics you should be looking at to get the best results in Google Ads.
I'm your host, Jyll Saskin Gales. I spent six years working for big brands at Google, and now I work for you.
This is Inside Google Ads, Episode 106: Performance Drops.
Our first question today comes from DrumSolo 1094 on YouTube, and they ask a three-part question:
Which metrics should beginners focus on?
Which metrics should you focus on with regards to the campaign objective?
And when should you evaluate your new campaign's performance? Monthly, weekly, daily?
I love this question because it touches on the habits we form as advertisers.
Let's start with the timing because that's actually the place most beginners go wrong and even a lot of advanced practitioners go wrong. If there's one thing you're going to learn from this episode, let it be this: Do not evaluate your campaign performance daily.
I know it's tempting. I know you want to see what's happening, but looking at your ads every day is like weighing yourself every hour when you're trying to lose weight. It's just noise. Daily fluctuations are normal and if you react to them, you will likely over-optimize and actually hurt your results.
As a general rule of thumb, if you are an agency managing ads for a client, feel free to look at the ads every day, but you shouldn't be optimizing more than a weekly basis unless you're working with a very large budget.
If you're a business owner and your account is in a pretty good groove, you can probably go down to just checking on your campaigns weekly and making changes monthly.
If you've just launched a brand new campaign, feel free to check on it daily just for the first couple of days to make sure it's serving and nothing is broken. But you should not be making optimization changes that quickly because you're not gonna have to actually learn anything and make informed decisions.
Now for metrics, I like to divide these into two buckets: reach and efficiency. Reach metrics are absolute numbers that tell you how much you got for impressions, clicks, conversions, cost, et cetera. Efficiency metrics are ratios, click-through rate, cost per click, conversion rate, ROAS. You calculate them by dividing one reach metric by another.
Which ones you focus on depend on your campaign objective, but I recommend picking one reach metric and one efficiency metric as your key performance indicators. Because your goal is likely conversions, sales, leads, you will likely choose either conversions or conversion value as your reach metric, and then CPA or ROAS as your efficiency metric.
If you are a beginner, don't get distracted by vanity metrics like impressions or CPCs. Have a look at them, use them as a potential optimization lever, but at the end of the day, focus on the metrics that actually pay your bills.
Our second question comes from Awesh Raza on LinkedIn, and they say, if a Demand Gen or Display campaign has been performing well historically, but performance drops in the last three to five days, what's the best approach? Should we duplicate it and create a new campaign with the same audience? Or should we add new custom segments directly into the existing campaign?
Awesh, my first answer to you is do nothing. You mentioned performance dropped in the last three to five days on a Demand Gen or Display campaign. That is not a trend. That's a fluctuation. Google Ads naturally ebbs and flows. If you make drastic changes because of one bad Tuesday and Wednesday, you're creating a problem, not fixing one.
With what I'll call the non-search-based campaigns, such as Display, Demand Gen, and Video, they work on a different time horizon than search or shopping. They can be volatile. They need time.
I have a Google Ads coaching client in a similar position. We've met quite a few times now, but he spends quite a bit of money on ads, and so he gets antsy. For example, before one call, he had changed his target CPA every single day for the last few days to try to optimize. And I told him what I'm telling you now: Pick your hands up off the keyboard, look at them, then put them under your butt and sit on them.
You're welcome.
Now, let's say the drop continues for two to three weeks and you decide that you do actually need to optimize and make a change. Should you launch a new campaign?
Almost certainly not. If you launch a new Display, Demand Gen or video campaign with the same audience, same creative, you're just splitting your data. You're taking all that historical learning from your current campaign and throwing it away to start from scratch in a new one. I know the previous campaign wasn't working well in the moment, but in the new campaign, nothing's changed. Nothing about the audience, your conversion tracking, or your creative is magically changing overnight. You're just throwing out your money for a new Demand Gen campaign to learn the exact same thing as the old one. You need to figure out the root cause of the performance drop and then fix it from there.
If you do have new audiences, like new custom segments you want to test, add them into your existing campaign. Work within the structure you have. And please wait more than three to five days before you do anything and at least one to two weeks before deciding there is an actual drop that needs optimization.
However, if you're feeling uneasy, you're not sure if you can wait, or you have waited and now you don't know how to actually fix the performance drop, that's exactly why I offer Google Ads Coaching. Let's optimize your campaigns together and get your Google Ads working the way you want them to.
You can book a call with me at jyll.ca, that's J-Y-L-L dot C-A, or follow the link in the episode description.
Our final question comes from Hassan Jahangir on LinkedIn, and they say, I'm currently running a seasonal campaign in the UK for a portable barbecue grill. The performance was strong until the last 14 days, but with the summer season ending, results have started to decline. Could you please guide me on what strategies I can implement to maintain or improve performance during this period?
Hassan, I have some tough love for you, too. You are selling barbecue grills in the UK and summer is ending. The decline in performance is not a campaign failure, it's a market reality. You can't force people to buy barbecues when it's raining and cold and cookout season is over.
When we see a performance drop due to seasonality, it usually shows up first as a drop in conversion rate and then changes in Impression Share and impressions. People might still click on your ad, but they're less likely to buy, so conversion rate drop. And then fewer people are likely searching, so you'll see really big changes in impression share.
So what can you do? Let's start tactically what to do right now, and then we'll zoom out to a bigger picture strategy.
First, you can adjust your creative and lean into the seasonality. Is there an end of summer sale? Can you pivot the messaging to be about getting ready for next year?
For example, I'm recording this episode in January and I just saw an ad recently for a window cleaning company and it’s a video ad that specifically says, “We know you don't need window cleaning right now, but we are fully booked up for the summer by May 1st. So book early to ensure you have your spot this summer.”
So this is exactly what they were doing, adjusting creative to lean into the seasonality.
Next, adjust your targets. If fewer people are converting, your CPA will likely go up and ROAS will likely go down. So if you want to maintain volume, you have to loosen your targets, raise your CPA or decrease ROAS. If you're not willing to do that, that means you're just not willing to advertise right now.
Which brings us to the third thing you need to do, accept the drop. You can't make sales on searches that literally don't exist anymore. So if you can't be profitable at that lower conversion rate, then you might want to lower the budget or pause the campaign until next spring. Don't try to optimize your way out of a seasonal trend by changing keywords or bid strategies wildly because you're fighting human behavior and you will lose.
As you prepare for next year, I recommend you discuss with your client two possible strategies for the off season.
First, have a frank conversation with your client about seasonal demand and decide together that you're going to lower the budget in the off season, how much will you lower it by, what will the new metrics for success be, et cetera.
Or two, have a frank conversation with your client about seasonal demand and decide to shift budget from Search into Demand Gen, Display or Video to build awareness during the off season.
Note that both of these strategies involve having a frank conversation with your client, which is the hardest part. And if you are the business owner managing these ads yourself, then you must be honest with yourself about seasonal realities in the market. You can decide if you want to maintain consistent budget throughout the year, but with different objectives, conversion versus awareness, or if you only want to focus on sales, in which case you need to lower your budget so you don't overspend in the off season and blame Google Ads when the problem was really your strategy.
To wrap up, I know performance drops are scary. When it's your own money, when it's your client's money, when it's your boss's money, it's scary. But often, performance drops are just noise. If it's been less than a week, wait. If it's a non-search campaign, wait longer. If it's seasonal, adapt your expectations to reality.
And remember, look at your data over a 30 day period before making big decisions, potentially longer if you have conversion lag.
And if you still aren't quite sure or you need a second set of expert eyes on your account, you can book a Google Ads coaching call with me. With more than 700 coaching calls under my belt in the last few years, I have seen every type of account and solved the most complex problems you can imagine and ones you can't even imagine.
You can book a call with me at jyll.ca, that's J-Y-L-L dot C-A, or follow the link in the episode description.
Today's Insider Challenge is this. You are auditing a Google Ads account and you see that for a specific campaign, the CPA has doubled in the last 7 days compared to the previous 30 days. It looks like performance has fallen off the cliff. However, when you look at the Attribution Report, you see it takes 12 days on average for users to convert. Does this change your diagnosis of the performance drop? Why or why not?
The beauty of the Insider Challenge is there's no right or wrong answer, just an opportunity to stretch your brain on real life Google Ads problem solving.
Last Episode's Challenge, Episode 105, was this. You have a client who is a chain restaurant. You launch a PMax campaign with an audience signal around business professionals, people who like to dine out, and a customer list of past diners. You also add search themes for relevant terms like “date night restaurant,” “pizza restaurant near me,” et cetera. Three weeks later, you check the audience insights tab and you see that the top converting segment is cooking enthusiasts. And when you check the Search Terms Report, you see a lot of brand traffic that you didn't exclude. The campaign is spending its budget, achieving its goals as set by its bid strategy and conversion tracking, but the client is complaining that they aren't seeing an increase in diners. What do you do? What's your next move?
I asked JyllBot how it would solve this. JyllBot is an AI chatbot that I created exclusively for my Inside Google Ads course members. It's been trained on all my podcast episodes, newsletters, blog posts, articles, but also trained on all the lessons in my courses, my bestselling book about audiences, all my LinkedIn posts and comments, information that's not publicly available from me. So I'm going to share with you what JyllBot said because I agree with most of it. And then I have an additional tweak at the end.
JyllBot says that this is a classic case of cannibalization and a misalignment between the platform metrics with Google Ads reports and the business metrics with the client sees. The PMax campaign is taking credit for people who are already going to dine there. That's the brand traffic. And that's why the ad account looks great, you know, high ROAS or good CPA, but the restaurant isn't busier.
So here's what JyllBot suggests:
1. Diagnosis. Understand why cooking enthusiasts are the top segment. It's likely not because PMax found a hidden pocket of chefs who love chain restaurants.
Human Jyll context to add here. It's probably just because a lot of people are categorized as cooking enthusiasts and this restaurant, it's a chain restaurant that would be appealing to lots of different kinds of people.
2. JyllBot recommends adding brand exclusions, and I agree. Either add the brand exclusion in PMax or add negative keywords around your brand.
To be clear, this is not a blanket recommendation that I have. I don't always think brand exclusions are the answer. I actually have no problem with brand in PMax. But for this restaurant specifically where there's a lot of brand demand, we don't need to be advertising on that right now. That's not the objective that the business owner is looking for to increase diners. So let's exclude that, but that does mean that it's going to look like performance is dropping in PMax because of course that brand traffic was very highly converting.
3. JyllBot recommends auditing the conversion action. And I agree, the client is saying, you know, we're not seeing more diners. So what are we counting as a conversion? Is it someone just viewing the menu? Is it getting directions? Do we have some way to actually measure if they make a reservation? Of course, that's going to impact results here.
4. Now here's where JyllBot and I disagree. JyllBot's last recommendation was to use the new customer acquisition goal. And honestly, I think that's overkill here.
Instead, my fourth step here is to bring in the theme of today's episode. It's a new campaign. It's Performance Max, and restaurant dining is not a performance marketing play. It's not like - right now, in real life, we need new snow removal tools at my house. There's been a ton of snow in Toronto. So I have been aggressively Googling, watching videos, and reading reviews for two days straight. And I just dropped four figures on heated mats, shovels, and plows, because there's a lot of snow out there, it's an immediate need, I bought a ton of stuff.
Dining out at a restaurant isn't quite like that. Yes, some people may search “where to dine tonight”, but others may just be browsing in their area, planning a work dinner, visiting a new city. So in just three weeks of advertising, even if you had already implemented all the previous suggestions made here, I don't think you would notice an increase in diners yet. There's some misaligned expectations here.
So we need to do some client communication. I need to share with the client that it's only been three weeks. Here's some optimizations I made to improve things, but it may take a few months for you to see a noticeable difference in more diners. And if that's not a time horizon that works for you, we may want to reevaluate the strategy.
What about you? Would you do something the same, something different?
I'm Jyll Saskin Gales, and I'll see you next time Inside Google Ads.