How to manage and keep track of your ad budget with Remi Turcotte
In this podcast episode, we catch up with Remi Turcotte, Founder of Turko Marketing, to learn about managing an ad budget.
Here's what you'll learn
Subscribe to the Marketing Analytics Show newsletter
Be the first to know when a new episode drops.
Subscribe to the Marketing Analytics Show podcast
Learn everything you need to know about marketing analytics.
Hello, Remi, and welcome to the show.
Thank you, Anna. Thank you for inviting me.
It’s awesome to have you today, we have a super, super interesting topic for our paid performance marketers. And it is ad performance reporting for paid media accounts. So my first question to you, Remi, is first of all, why should marketers improve their ad performance reporting for paid media accounts? And what kind of tools can they use to do it effectively?
So like, first and foremost, when you manage your multiple accounts, it’s very difficult to keep track of how well media budgets are being spent. So what I mean by that is that the KPIs that the platforms are going to give us by default are really, really standard metrics. But when it comes to how much I spent on high-performing ads versus low-performing ads, it’s kind of hard to get that information right off the bat. So, as I said, the process is painful and honestly, it’s very important to do this exercise though, because it’s really a very important part of the success of a marketing campaign.
What we do to extract this information is that you can do it manually and do some exports to all the platforms and start to pivot all the information. And it’s very, very hard, and if you have multiple accounts, it’s going to be close to impossible to do that manually. So what we do, we use a Google Sheet and Supermetrics. And honestly, it’s very simple, the approach. We don’t need to install a special program or learn a new UI. It’s really, really like just flat exporting the ads from all of our accounts with Supermetrics, Google Sheet connector. And with Google Sheets, we have some simple formulas to gather the information.
Awesome, and I’m super happy to hear that Supermetrics works nicely for you. And now let’s move into the specifics a little bit. My next question here would be what kind of metrics would you include in your ad performance reports? And also maybe you could clarify why you would include these specific metrics.
Sure thing. Basically, we’re going to report on Google ads, Facebook ads, and we also report a little bit on Bing. And what we are going to check for measuring how well the budget is being spent from an ad performance standpoint is that we’re going to look essentially at all the quality-related metrics for all of the platforms. And we do that because it’s going to help us later kind of pinpoint which ads are underperforming and which ads are good. So let’s say, for example, the metrics that we’re going to collect on Google ads, for example, is going to be they serve historical quality score. That’s like the main score that we want to look at. And then we also have the possibility on Google Ads to see, for example, the predicted CTR and the landing page quality score. So that’s very interesting also to, later on, understand where is the problem, more or less from a quantitative standpoint.
More or less the same metrics that we’re going to look at on Facebook and Bing from a quality standpoint. And I also like to add to this all the other metrics, such as the cost impression, the clicks, the CTR, the CPC, and the CPM. So I like to add these metrics, especially because I’m going to need later to order the ads by which one has the most cost to the least so that I can prioritize my actions. And from a dimension standpoint, I’m going to extract the following dimension. So I’m going to extract the account name. I’m going to extract the campaign name. I’m going to extract the ad group name. And if it’s Google, I’m going to extract also the keyword and the match type. If it’s for Facebook, I’m going to stop at the ad group name. But if it’s Google or Bing, I’m going to go all the way down to the match type.
All right. Yeah, that sounds awesome. And also sounds like there are a lot of metrics and dimensions you have to keep track of. So now let’s talk about examples a little bit so our listeners could understand how to pull it all together and analyze it. So if we take Google ads as an example, can you please explain these indicators and how they work a little bit further on, let’s say we have an underperforming account. So how would you look at these data and what kind of conclusions would you make that would help you optimize your campaigns further?
So let’s say that I start from a blank Google Sheet with Supermetrics connected. I can, I’m going to create a blank tab and I can name it something like campaigns ad performance for all accounts, for example. And I’m going to create a Supermetrics query. And in that Supermetrics query, I’m going to add, as I’ve said, I’m going to add for the dimensions, I’m going to add the account name, the campaign name, the ad group name, the keyword, and the match type. As for the metrics, I’m going to pull the historical quality score, the historical predicted CTR, the historical landing page quality score, the historical creative quality score. So everything quality score related, so there are four metrics just for quality score information. And the standard performance data that I like to pull, I’m going to pull the following metrics. The costs, the impressions, the clicks, the CTR, the CPC, and the CPM. You could also pull the conversions if you want but really the point of the exercise at this point is to be able to see them from an ad quality score standpoint.
And basically, if you have that number, the quality score, then you don’t necessarily need all the other numbers. Essentially, I like to have them because I can compare, let’s say one ad to the other, which one has, let’s say, the highest CPC or lowest CTR. But essentially I could also boil it down to just having the metric cost. And like I’ve said previously, to order it by order of number. And basically, once I pull this information and I do this for all of the accounts, so you might want to, for example, you might want to have a lot of rows in your Supermetrics query. You want to limit the number of rows based on the number of ads that you have, for example. I don’t know, it could be for managing a lot of accounts. It can go from, I don’t know, let’s say 10,000 maximum rows to, I think you can go up to one million or something like that. Yeah, one million with the Supermetrics extract. So it’s kind of efficient.
But what I do is to ensure that I don’t have ads that never run. So I filter, I create a filter with the impression so everything that had at least one impression, I want to see it in my database query. And in terms of the dates that I’ve got to select, I’m going to select, well, it’s up to you. You could do by quarter, you could do by month, you could do a year to date. It really depends on you. I like to have it, let’s say last month, and do that check maybe on a monthly basis. And basically what it’s going to happen is that I’m going to have a whole data set of all the ads with all their quality scores and how much they cost. I’m going to create the descending order of the cost. Then I’m going to have them… Then I know which one I want to look at them when I look at them in order of priority.
And what I do is that I do a filter on the historical quality score. And I, let’s say as a first exercise, I really want to have only those that have, let’s say a score that is lowered, and let’s say five or four, for example. And then it’s going to give me a nice list of ads that are underperforming in my opinion and, from a Google standpoint. And then I can see, I have all the information at my disposal to know, for example, which campaign is the problematic one, which ad group, which keywords. So that comes back to why I like to extract it all the way down to keyword and even the match type.
And in the end, if you’re, you just do your job as a marketer, and you just try to make assumptions based on the information that you have about what is the problem. So, for example, a campaign for account XYZ that has a quality score of three out of 10 for a certain keyword. And then I can, when I look at that account and when I look at the ad that’s being pulled for this keyword, it’s quick for me to realize that, let’s say the keyword doesn’t match, for example, with the search intent. So often it’s going to be as common sense, basically. That’s pretty much how you work through it.
Awesome. Thank you so much for such a detailed explanation. It does sound like a very, very comprehensive report. And, so, let’s imagine the continuation of this scenario. So a marketer has to put together a report, they’ve identified low-performing or poor performing campaigns with low-quality scores, and they have needed certain assumptions to optimize them. Now, what would happen next? What kind of methods would you check to see whether your optimizations have resulted in improvement in performance and for how long should you track that?
Well, it’s a good question because being able to work on this, on the quality score, it is very doable with pulling all this information from just one query, for example. But the thing is that if you have a… If you’re working for a client or you’re working for your upper management and you need to report the progress of that, because otherwise who is going to know that you’re doing great work, right? So it’s very important to refer to the progress. And basically, how can you report the progress on the quality score improvements that you have done when you’re working on several accounts at the same time? So it’s kind of hard to do such things.
So one approach that I believe is efficient is to look at the amounts spent, and then what I do is that I separate it into three buckets. I separate it into over-performing ads, so these are the ads that have a score of 10, nine, eight. Then I have the second bucket, which is the average ads, so the ones that are performing between seven, six, and five. And then I have the low-performing ads, the ones that have four, three, two, one, and even zero as a quality score.
So I do a little bit of Google Sheet calculation to give me these numbers. And what I do is that I do a weighted average based on the cost. So that’s really like, I can see, for example, I can see what’s the percentage of ad spend that has been done on low performing ads, on average performing ads, and on over-performing ads. And I believe that that number which you can turn into a percentage, you can easily benchmark and report on that number. So that’s really an interesting way in my opinion to work on optimization and also report on the improvement.
Okay, that sounds excellent. Thank you so much. It was a really, really well-done analysis. I loved it. And now let’s talk about the typical mistakes a little bit because I’m pretty sure that somebody will try to pull together a schedule report and then analyze it and make assumptions. They will run into some kind of mistake. So what have you noticed from your experience? What would be a typical mistake a marketer could make while analyzing their historical quality score data?
Well, one obvious mistake is that, well, first of all, one mistake that marketers make is not to analyze the quality score. So I believe that in the era where everything is automated, people believe, a lot of marketers will trust the platforms and they will just go with the platforms’ recommendations. But you need to do that extra step, that manual action to actually be able to analyze the quality score. So really, I would say the first mistake is to not analyze that because it’s a very, very important piece of information for optimizing your account performances.
But when it comes to… Also, I would add that when it comes to analyzing the accounts, I would say another mistake is to forget to use your common sense to conclude or make an assumption of what you’re seeing. For example, if an ad has a zero quality score quantitatively speaking, you could say, “Oh, this is a very bad ad.” But when you look at the ad, sometimes you’ll see, oh, it’s a call ad and the algorithm or the API doesn’t understand the call portion of the engagement and stuff like that. So you always need to start with the quantitative analysis but also add your marketer expertise and your common sense to everything that you’re looking at.
Awesome. Thank you so much for sharing. These are some really great tips. And now if the audience would love to learn more about your work, can they find you?
So you can follow me or my agency on LinkedIn. So you can search for either my name Remi Turcotte or Turko Marketing. Or you can go on our website, www.turko, that’s T-U-R-K-O.com.
Fantastic. Thank you so much for coming to the show today, Remi.
Well, thank you for having me.
Try Supermetrics for free
Get full access to Supermetrics with a 14-day free trial.
No credit card required.