Hi everyone, my name is Yana Lyashenko and I’m a Google logistician. I’m in the business of delivering targeted business audience with the right parameters and basically, I scale your online sales in different countries around the world using the same AI-ish Google tools.
Today I will be showing a team case study. Our theme is a spare parts store for, let it be, promotional equipment, in order not to specify in more detail, not to burn client data. I will tell, show the difficulties that were in this project. And the general dynamics of what a specialist can do with deliberate thought. That the result you will see is not the merit of some accident, but of deliberate purposeful actions in the account.
Difficulties encountered
So where do we start? We started our work from the month of January. I noticed this on 02.01.01. That is, somewhere from this period we started our work.
How many calls and sales will I get by ordering contextual advertising from you?
Before that entire interval, it was search. The client was basically getting conversions from search. I’m going to select only our advertising campaigns. Here you will immediately see that it took a very long time to develop the account.
The difficulty we encountered is that the fourth analytics is a shameless service for transmitting any normal, adequate data on conversions. I always say to people who come for consultations: “Let’s switch to Google Ads conversion, because it is impossible to work with the fourth analytics. I will even show you the time period in which this fourth analytics worked, what results it produced.
Let’s take a closer look at it now. I’ll take now the segmentation by conversions and you can literally see that from January 29th (here it was tracked in parallel), even from January 28th, we almost completely give up the fourth analytics, we just switch to the purchase tag. Thank you. Oh, my God! Hallelujah! As you can tell.
Here you can even see that in these time intervals we have a very high price per conversion. That is, here for you to see – purchase. Here, please, conversions, purchases. In general, for this period of time, the price per conversion is 85.84 UAH, while the CPA-price that the client gave us was 90 UAH. That is, in principle, we fit in.
Test results with the available budget
In general, what else would you like to note? All of this result is done on a small, relatively, daily advertising budget. This affects the number of tests that can be run in advertising campaigns. Here we have practically everything we need. There are breakdowns in the number of clicks, traffic, there are periods of campaigns that spin off, weekends in which they give less or more conversions.
There are test periods when the price per conversion was raised, and here you can see that they lowered it again to 88 hryvnias. That is, the account, in principle, even within the limits of what we have quite alive, you can see that something is tested, something is done in the context of being better in the end.
Cost-per-click results
Another of such important things that I would like to note is the price per click, which was stabilized. Here you could see that in these time intervals it was really inadequate. Here it was 6 UAH, 9 UAH, for example, here it was also 5.75. Now it is, on average, up to a maximum of 3 UAH somewhere. This was done with the help of optimization of the same feed.
Conversion rate results
Also would like to show the conversion rate. Cost/conv has been selected. Conversion rate is generally quite high over this entire period of time – 2.50%. Why did I want to mention it? For the simple reason that it is high enough and good enough. We consider the lowest Conversion rate in our team to be 1.5%. Why 1.5%, because here CPC on average is somewhere below 5 UAH, so 1.5% is usually our desired Conversion rate when optimizing advertising campaigns.
I remind you that you have two indicators that can affect your profitability or CPA-ka very significantly – it is the price per click and Conversion rate. Why 1.5%? Usually it is advisable not even to rely on my figures, but for each project to calculate your own Conversion rate values, which should be generated by your advertising campaigns. They are calculated on the basis of your margins, the average check that rotates in certain advertising campaigns. The more diverse the check in products – the more for each segment it should be personally calculated. Because in some niches, for example, even a 3-4% Conversion rate is not enough to fit into the KPI. So you probably need to calculate your conversion rates for yourself personally.
Results of the data-driven attribution model
Here, for example, you can see that it was the fourth analytic, GA4. I don’t know if the editor will highlight or not, it was on a data-driven attribution model. Also a shameless conversion, I call it. And the data-driven attribution model, I don’t know how many projects I’ve had, the cheaper the price per conversion, the more likely it is that I’m not going to use that data-driven attribution model.
No matter how much Google promotes it now, I believe that it wastes some money and is not yet useful for advertising accounts. Although in some cases it should be used, but that’s probably a topic for another case study. Not for this one, for sure. That is, even if we assume that I underestimated the price per conversion because I took duplicates of some conversions or something like that – we can look from this period, for example, from January 29. January 29 will be the start date for us and we can see that the price per conversion is 68 UAH, the conversion alone is fixed. And, of course, we have more expensive merchandise.
How many calls and sales will I get by ordering contextual advertising from you?
General analysis of available advertising campaigns
But let’s take a look at the overall analytics picture. We don’t have a lot of ad campaigns here, as you can see. The main pulling force is Pmax and the usual trade advertising campaign. There are also three test campaigns launched here, but they have not received any impressions. Why so few advertising campaigns? Because, unfortunately, the daily advertising budget is a little low, and we either have to sacrifice these things of conversions, to go into the scaling of different segments of some campaigns. Or the entire budget to direct to these advertising campaigns, which was done.
What do we have? We have a regular commodity, we have Pmax. They’re launched at the same time. Completely these are all intelligent strategies. This is the answer for those who say that you don’t need to automate everything, that you need necessarily to have some on manual bidding companies – no. All automated advertising campaigns cooperate normally, everything can be adequate. An ordinary product, the purpose of its launch was to test a certain value of profitability – whether it pulls or does not pull.
In principle, its overall result is a ten, it is at least twice as expensive as the desired price for conversion. But it gives a very high value of the average check in certain periods of time, so for now it is left as a variant of testing on profitability.
The other campaign is running in test mode without limitation pretty well for now, but we need to scale this account as well. Since this is the end of the maintenance period, we need to discuss with the client more about scaling in pieces, because as soon as we get the “Budget Restricted” status, he can give us a bad name in the context that he will start to cut our reach a little bit even at a very cheap price per conversion.
It’s an ordinary product, it’s not super top in terms of traffic, but you can see that in terms of conversions and clicks it’s one of the more or less stable ones. Why does this campaign have a rather expensive price per conversion? Because there’s a significant portion of products that are spinning off and have a certain cost part, depending on what we’re going to be looking at.
Yeah, we’ve got some of the bones here. If you filter through it, you can see that there’s something here that’s already been excluded from this campaign altogether. For example, eliminated. There’s out of stock. There’s excluded. Something is left out based on certain criteria. That is, it is one of these most predictable ad campaign optimization systems.
Why is it excluded here? To not already shine the goods show – take Shopping products – Item ID and take a filter, for example, less than 60 UAH. Right? They’re not unscrewed to 90. And, for example, more than 5 hryvnias. We will see that some of these goods that spent less than 60 UAH, here 57 UAH and less. They made up the main share – 7054 hryvnias. Let’s also remove conversions – the filter “Less than one conversion”. This is just, by the way, the period when whole numbers were purchased.
Here’s 6,000 hryvnia. If we look at the commodity – you can see that the price per click everywhere is super cheap and there are enough clicks. But the cost amount is far from 90 hryvnias. This is the main cost aspect that carries this sense that the price per conversion is much more expensive than what can be done.
I mean, if I take away, for example, the filters on this selection now and change it the other way around. I’ll make the conversions “Greater than zero” to see from one conversion. And I’ll take “high to low.” You see, there’s 5 yr, 26, 11, 80.
There will be something expensive in the price per conversion depending on the type of tests that were done. For example, this product well right on it well it crosseil converts well. 411 UAH here, for example. But mostly there are goods for 80 hryvnia. 70 kop, but there are two conversions – there is a question here, here it was from one click. 74, 5 UAH, for example. That is, it’s all rather smeared and not all are straight conversions. They’re all inexpensive in fact. If you take the average 30 UAH very primitive. But the high price for conversion of 198 UAH is due to the fact that there is a large piece of goods that scrolled, advertised, but did not bring conversion. That’s why they form such a high price for conversion.
Conclusions Conclusions
In general, the average picture so far is like this. During this period 112 UAH. Well, and Conversion rate was even higher. Yeah, I agree here. There’s this common coworker here, she’s ruining the picture. God, look at that beautiful number 666, conspirator. This All Items campaign is over-performing. 68 hryvnia, everything is awesome, cool. The main share of conversions is carried, that is, it can be scaled up further.
This is an example of a case study that I can show you so that you can see what results can be achieved. The bulk of the campaigns here were not ours, not client campaigns. That’s why they’re here. I’m going to filter it now, I’ll just show you ours. And the Conversion rate of 3.12% is pretty good, good.
Most likely the problem is all items – and the fact that the profitability is weak. Yes, 900% for this niche is a very low percentage. In order to feel where we are in the auction statistics in reality or not in reality, what price per click is suitable for us, what rates and so on. If you see 1.85 UAH, and here it is 2.15. So, by the way, this is the answer to the question of whether it is necessary to reduce the price per click very much? As you can see, even with a more expensive price for clicks – you can win, unambiguously.
That’s why cases like this with demonstrations of what’s good, what’s bad and what really…. If everything here was pure randomness in performance – most likely even this campaign would be super cheap and with a very high Conversion rate. Here you have another case study of a team due to proper placement, planning of actions to be performed in an advertising account.
