Hi everyone. My name is Jana and I am a Google Logistician. I am in the business of delivering targeted business audience with the right parameters. I scale online sales of businesses around the world using the very AI tools of Google Ads.
Today’s video is going to be a breakdown of almost from August month, from the end of August month of 2024 to basically March 10, (I took the eighth) the team’s case study. The niche we have is, let’s call it health products. I won’t specify which ones in terms of customer privacy. There are a lot of health products, beauty products as well. And this is the market we have in the United States, that is make America greate again, the United States of America, this is the market.
I’ll say right away that I’m not sure that this is the perfect bone of contention. Let’s make sure that I haven’t lied to you anywhere. Here, for example, I will redo the conversions so that you can be sure that the perches are really correct, Cost of conversion is also displayed adequately.
How many calls and sales will I get by ordering contextual advertising from you?
Honest case study shows it the way it is. Tells the difficulties we have encountered. You can’t really make sales in this niche just like that. I can honestly say that you can’t get sales just like that here, you can’t. That is to say, you can’t just go in and get something, your piece of the pie. The problem is not only with competitors who may be analog to you. Especially the competitors that we felt, we felt this array, this pressure from the markets are competitors like Amazon, who just stupidly took out our rendition. Not our branded one, but the traffic that we had selling during the Black Friday period. I’m going to honestly show you everything as it is. That we did not take off on Black Friday, although we also had discounts, sales, and there is an audience that is ready to buy from us.
I want to emphasize that there are niches that are really difficult and that you can procrastinate. I will show this period of time, when you are fiddling around, treading on one place and you can’t do it right, adequately to the growth of this number of conversions.
What was done by the team and the period of support?
I’m proud of my team that we still don’t give up, that we come up with tests all the time, we’re constantly working on the structure, we’re constantly monitoring the dynamics of the auction statistics, we look at where we are, where we are adequately, what kind of traffic we’re getting, whether it really converts us or not, what kind of conversion rate we have in broad traffic, in narrower branded model traffic.
I’m pleased that we developed a strategy to bypass the squeezing of traffic by the big advertisers and that really after starting to work on that strategy, we went into really big growth. And I want to show that this growth result is not due to, for example, Black Friday or the December 24-25 sale. Yes, there is February 14, but excuse me, the week of February 14.
Here you go, yes, you see this week. On the 10th we had a peak that even the week of March 3rd through March 9th we have 103 sales. And as far as I know, March 9th, March 8th is not a super cool holiday that everybody celebrates in the States. I mean, I want to show that this growth is not the result of us just being lucky, that we just launched, we just got into some auction on time, but it’s the result of the team really working. That it’s trying new things, testing new things. Yes, not super into budgets she can make these tests big, because still you want to keep the price per conversion low enough.
What should the price per conversion be?
I’ll honestly say that the price per conversion should be a little lower. Honestly, objectively I will say that. I’m not going to scorch profitability and other things here, because I realized that there are situations where even on the basis of profitability I can monitor the statistics of auctions, even understand who works where, so I won’t show it. All the more so because the key KPI is the cost of the conversion. It should be a little lower. And I’m not going to say halfway down here it’s not $7. We’re very close to that, but still a little lower. I think it will be much better then. In terms of profitability KPIs, we are within it. I would like to do more, of course. But due to all these tests that we are doing and we want to catch up, overtake this period of stagnation that we had and against the background of these auction periods, for example, until February 14, we would also like to catch up.
I’m showing you honestly how it is with the conversion price, that it’s not super cool right away. We don’t get 1,000 and 1 sales for $1. When we started work, we have a start date of August 25.
Well, give or take a few days to set up. This is roughly the period when we started to get involved. I mean, we started with a number of sales somewhere in the neighborhood of 20, let’s take 30, right? 25, like 30 sales, that’s where we started. Well, the CPA-ka here was in the neighborhood of 20$, really expensive. We tried all sorts of methods to bring it down. What have we gotten to? We got to 49 or so. On average, we were able to get somewhere around 40-45 sales plus or minus steadily.
Working on lowering the price per conversion
Then we started trying to lower that price per conversion. There were even situations where we got it down to a pretty good number there. Even on Black Friday period we were getting 11$ price per conversion.
But when we were working here, we realized that the variant of the way we scale, it is not suitable. And the key starting point for this was the Black Friday period, when we purely physically could not under our limitations, which we went along with the test campaigns, to make a shot in the pieces of these conversions.
How many calls and sales will I get by ordering contextual advertising from you?
Why was it not possible to increase the number of sales?
Why did that happen? Because, in principle, I can honestly and objectively say that a major advertiser crushed us. This big advertiser, it is not a secret, it is in many niches – Amazon. He saw at some point that this is a very incremental story. Not only that, this account has a history for the 23rd year when they were growing, what was the price per conversion, other things, but I’m not going to show all these situations, I’m showing our work. In those moments Amazon was still just probing this niche and now it’s just tinkering at the max on it and it’s really eating up a chunk of sales.
And we realized that on this format of work that we built, this structure that we have, it is not one, not two, not even five campaigns, it was much more, we cannot grow in any way. I mean, we’re treading water. Not only that, we slept in on Black Friday and that possibility could still be around this December 25. Well, this is the Christmas sales period, when discounts are also given and there is also a great demand for products. And we realize that we can also fail with another campaign.
A new version of the strategy has been developed
We have developed a new variant of the strategy. Of course, based on some strategy variants that have already been tested. But remember, as I say on my YouTube channel, each ad account has its own individual shade, like a fingerprint. Why?
Because specifically in this account, we scale through AI-driven Performance Max campaigns, (Pmaxes). And unfortunately, or fortunately, it operates on your own analytics, including your customers’ pre-purchase behavior, already collected analytics. And it gives that uniqueness to your account. And, of course, understanding this whole story, the hypothesis, we started to implement additional tests. These additional tests allowed us to reach 67 sales already in the period of these sales. That’s since December 16th. Which December is that? It’s 16th through December 23 somewhere, or December 22. Then 23-25 = 74. There you go, February 14th. That’s the junction of February 10th and February 17th. Look at this, 111-93.
Yes, there’s a very high price per conversion here. It was lower here, you see, 15.00. And so the period from March 3 through March 8, when I’m recording this, March 9 or 10, 103 sales at $13.00.
Yes, there can be some vicissitudes. We still do a lot of test variants to get a feel for where we can go deeper into some auctions and pull out these cheap sales.
Why is this niche difficult?
Why is this niche difficult? Because there are competitors, an auction and you have to fight for your place in the sun. The main difficulty is that there are actually too many offers. And that offer is made by these Amazon-like price aggregators, if you can call it that. When you go to Amazon, it’s just a big search engine itself. There’s also a sampling of these products. And there are hundreds more stores that offer something analog similar. Moreover, to achieve results in “Health” or “Beauty” you can, in principle, use some poultices, lotions or folk methods, right? That is, there are 1000 and 1 options to satisfy the needs of the end user, the client. Right? And that also makes the supply greater than the demand for that commodity. And you also have to create demand for this product. So I am pleased that we have found a way around this situation. We did not repeat in December the situation that we had with the Black Friday period, that we did not get enough sales.
Planned work for the future
I think that we are fixing this result now, we will continue to grow systematically and we will make even greater results for the client. Because I believe that the fact that the team knows how to correctly use the reports that Google gives, to correctly understand where we are in the auction statistics, in which auctions, in which categories of goods, in which categories of requests, who is our main competitor, in which cases the same Amazon, for example, becomes more active or some other competitor becomes more aggressive? It’s understanding where you are in the auction, how to properly operate the product, how we can bypass the big advertiser, because in fact we have done a little bit of it. I’m not going to say it’s in a week, let’s be objective.
Team result in the period of the month
For example, if I take the month period, I want to brag about the team result. After all, the month isn’t over yet. I’m shooting this March 10 through March 8. See, why March 8th? Because conversions are pulling up late.
Look, just for the actual 8 days of March there are already 129 sales. Using November as an example, we had 137 for the whole month. That is, we made November during the Black Friday period. And the cost per conversion is only a couple bucks more. Why is it more expensive? Because we do test campaigns. Not all of those tests actually end up being successful. Because AI-powered performance marketing and performance work is about trying to add new similar traffic, getting into the analog auction of a competitor that’s close to you, trying some narrower ways of targeting. It’s a whole bunch of different hypotheses when you’re building them. And it’s refining the feeds, optimizing the feeds. There is actually a lot of work being done here. It seems like it’s nothing and it’s a credit just to get something done, but to actually make 129 in 8 days is a cool result I think. Considering we have a more failed November. Not only that, for the entire month of December we have 241 sales, for $15.83, and here we have 13 for only 8 – 129 sales.
Conclusions
I think that this is a very cool case, which shows that sometimes the result is achieved not by chance, the period of discounts, sales, seasonality, but due to the fact that the team sits down, comes up with variants, how to bypass, how to find variants, how it is better there, how to penetrate some auction. In fact, in most cases, when analyzing in the right way, we operate with reports on auction statistics. Each of you has them in the context of each test campaign.
This is, for example, the case I wanted to boast about. Why don’t I tell you all the details of how it was targeted? I believe that this is the intellectual property of the team, including my development of the strategy, options for bypassing those very large advertisers. I shouldn’t share it, right? Especially since my client’s outcome depends on it. We try to keep each client’s information confidential as much as possible. So here’s a case like this, yes, a full-fledged bragging, but a bragging justified in terms of showing real results, that the price per conversion is not one dollar, something like that.
If I take Monthly, you can see that the direct on price per conversion is still a little bit higher in these intervals than it was here. Right? You can see that. You can see everything objectively. But we achieved growth in the price per conversion by not x2 price per conversion, but we went a little bit differently, by a different mechanics, by a different algorithm. Just to remind you, this is the States market. The niche is super difficult, it’s hard to promote. I won’t say it’s a super-light story. It’s actually very hard to figure out, to get around, to figure out what to stuff the AI thing with to make it produce more, to make that result cheaper. It’s really a piece of work. It’s basically a piece of work to analyze, to systematize, to test these hypotheses.
