How Many Clicks Before You Kill a Keyword? (The CPS Rule Amazon Sellers Ignore)

How Many Clicks Before You Kill a Keyword? (The CPS Rule Amazon Sellers Ignore) Speaker 1: What's going on, Badger Nation? Welcome to The PPC Den Podcast, the world's first and longest running show all about how to make your Amazon advertising life a little bit easier and a little bit more profitable. Today, I'm addressing a question that is among the most popular questions I get, which is, if I'm looking at a search term, how many clicks do I give it before I negate it? It's an incredibly important question to know, to think of, to consider. This little chart here is going to make a lot of sense by the end of this episode. Sort of a patience meter to sort of how much you should let something go. Let the games begin. So here's the scenario. You are launching a new campaign. You give it some keywords. Those keywords start generating traffic. They don't have sales yet. How long should you let them go? Some are going to convert really fast. Some are going to convert very slowly. When should you decide to label something as a poor performer? Versus it's too early to tell. That's what this episode addresses. The pain is clear because there's a range of non-converting wasted spend. I've seen everything from 30% non-converting spend on the good side, all the way up to 80% non-converting spend on the bad side. Yes, that means you're spending 10 grand and $8,000 is going towards search terms that generate no sales. So this is huge. This has really been top of mind. Recently, number one, I get this question a lot for the last many years. Number two, we're updating our Bids by Badger algorithm. So our Bids by Badger algorithm, I don't know if you've paid attention to some of the recent episodes, but we optimize like a million plus bids a day. And really like the biggest thing that I see that's still a pain point for a lot of advertising accounts to accomplish their target ACoS on the account level, which in turn they can accomplish a target profitability on the account level, is that Non-converting spend is so present that if you can simply just move your non-converting spend to converting spend, you'll be in a much better spot. If you take $1,000 of stuff that does not convert and you give it $1,000 at a 100% cost, you've just spent the same but generated $1,000 more of sales. The math is very in your favor to really be thinking of non-converting spend. This category of keywords to bid optimize and make decisions on is in the bucket of no conversions yet. And when I say no conversions yet, the time frame can be variable too. Meaning, you can look at a keyword over a year, it did great, over the last 14 days it did not. So understanding how many clicks per sale to go and like what your normal conversion rate is, it's very important. So the metric for you, I love so much. And actually I looked through my own sort of technology and I do not have CPS listed, clicks per sale as a metric listed. So like I want to add this to our tool and basically say like when you're looking at a keyword last 30 days, you can just quickly see what's my normal clicks per sale here. Or like you're looking at the campaign summary, it just has a sort of clicks per sale summary for you. You're looking at some search terms. In the summary row, it's like for this ad group, when I look at these search terms or my account or this product portfolio, whatever, I know the typical clicks per sale. It's beautiful. So the way to get there is just you take 1 divided by your conversion rate. So 1 divided by 10% conversion rate. So that's 1 over 0.1 as a percent would be So if you have a 10% conversion rate, you generally have 10 clicks per sale. The things to know about it is like higher, lower isn't necessarily a good or bad thing. It's more of like a how much data do I have to make a decision? Again, it's kind of like it's not an actual performance metric because you can't take your CPS to the bank. It's really a do I have enough data to make a decision here? So I think when people sort of understand this, I think a lot of people view it the wrong way. And I've been doing a lot of studying and research with what is your actual clicks per sale. Because if you take it at the account level, it includes all different kinds of traffic. It includes actually your converting spend and your non-converting spend. And if like, you know, let's say 50% of your spend doesn't convert, that means you're basically sort of like inflating your clicks per sale of good terms. Anytime we launch something, anything, You can either take a, I want to create more traffic that's similar to what I've already had, which includes wasted spend, or you're saying in the optimistic view, which is I'm launching a campaign and hoping and anticipating and I'm optimistic that what I launch will be new converting spend. So actually you want to compare your new traffic, your anticipated returns based off what you've been doing, like how you generally convert, because you want more of that same kind of traffic. So if your overall account conversion rate is like 10%, that's 10 clicks per sale, and you launch a campaign and say, well, I'll go two times over my normal clicks per sale, that's 20 clicks. But then you take the subsection of your traffic that actually converts And that conversion rate might be 20%. Then you want to use the 20% value because you want more traffic that converts like you're converting traffic. So we want a high intent Clicks Per Sale benchmark. That's what I was just saying. You should not judge unknown terms, including the non-converting traffic that you have. So sort of like trimming out your non-converging spend from your clicks per sale metric gets you closer to creating marketing tests where you launch new keywords and you launch new campaigns That are going to be similar to your already converting stuff. It's actually a really cool concept. And I've been studying this a lot. And basically, like the statistics behind this, and because I've been working on the BidByBadger algorithm since like 2017, like a lot of these terms I've learned and had experience with and like learned about this stuff, it's actually pretty interesting. So the first thing is that like a click either converts or it doesn't. And generally, what happens is that the data that you're getting should be consistent for some time. Meaning, if you have a statistically significant clicks per sale, a statistically significant conversion rate, what you would expect is that this is probably going to be similar in the next chunk of time. And you can use this information to find typical behavior for what your clicks per sales looks like. So, we basically want to understand that even of our converting data, even of the stuff that does convert, you don't want to be too heavily weighed at the 1% conversion rate or the 90% conversion rate. You want sort of typical behavior, not loud behavior. There's an assumption that most of your keywords are typical keywords, that most keywords that do convert follow sort of a range of performance. And you can sort of understand, well, if I have 100 keywords and I have one with a 1% conversion rate, but all the others are 10% conversion rate, That more than likely I should like weight my typical, median, normal clicks per sale at what is actually happening to the keywords that do convert. So really what you, so like in practical terms, what this means is like you might have branded terms that convert a 50%, competitor terms that convert a 2%, And generic terms I convert at 10%. So you have a clicks per sale of a huge range. You have a clicks per sale of two, like branded terms on average, every two clicks you get a sale. Competitor terms, you have to wait 50 clicks to get a sale. And if you sort of mix all these things together, you would end up with like, again, the same thing that we just talked about, which is if I get my normal clicks per sale based off My non-converting spend and my converting spend. So if I only look at my converting spend and then within my converting spend, I also have this big range. This also matters a lot too. You wouldn't want to kill off a bottom of funnel branded term by using the same measuring stick you would for a competitor term, right? Like these kinds of things happen. So if you were to sort of qualify And three simple buckets, Brand Terms, Generic Terms, Competitor Terms. Those are generally the three buckets. Each one of these could be parsed out a little bit more. But if you were just to break these out, Branded Terms, High Conversion Rate, Low Clicks Per Sale, you want to make quicker decisions because the conversion rate is so high. You don't need to wait so long to make a decision. And today we're going to talk about generic terms, competitor terms. So as you are looking at these things and as you're making decisions, we'll want to keep this in mind. Like what is our normal clicks per sale for branded terms or generic terms or competitor terms? And then you can simply begin to measure your search terms relative to your clicks per sale benchmark per category. So for example, you could then say, if I have a keyword, Search term. If I have a search term, anywhere from 0% to 75% clicks per sale for that category. So like a branded term or a competitor term. Let's take a competitor term of 50 clicks per sale, right? That's a high clicks per sale for a competitor term. If I have 75% of 50, it's not enough data. So what I would try to get it to do is I would try to get it into anywhere from 75% to 150% clicks per sale. Once it hits over a certain threshold of 1.5 clicks per sale of converting traffic, Per that bucket, you can then make a decision. And with all the data analysis that we have, you can make a really, really good, strong decision that, hey, 150% of my normal converting clicks per sale of my normal sort of bucket that it's in, I have extremely high confidence that this isn't going to be a dud keyword. So like this little exercise, and it doesn't take long, right? Run a search and report. Clicks Per Sale. That's a number that I think everyone on this show knows how to get. Clicks Per Sale for your branded terms. Run a filter, type in your branded term. What is the normal clicks per sale? You have your answer. And then you would do another one. Search term does not contain your brand name. You have that as well. And then competitor terms, you'd get the same thing. So once something hits like 1.5, your normal clicks per sale, it's going to be really simple to know what the answer here is, where you can then negate. You can pause. It's relevant, but maybe not right now. You can bring it in later once your conversion rate improves. All those things. So again, keeping this really simple. I think that having no game plan is bad. Having a game plan of just like your account level, non-converting spend and converting spend and just say, Hey, 10 clicks per sale. I'm going to start judging things. The issue with that is there are converting terms that convert like lower than your account average. So like you don't want to trim out those like 1, 2, 3, 4, 5% things. Um, too fast, but again, still better than other. We're trying to improve, not be perfect here. And it just adds a layer of confidence. So being intentional about what the search term is and how many clicks per sale does it normally get is important. So again, once something really hits like 1.5, your normal clicks per sale, you're going to be in a good spot to actually make a decision. Unknown Speaker: I've launched campaigns and picked keywords. I've got my bids, set placements too. No bad mistakes, I've made a few. I've had my share of problems. We are the PPCs, my friends And we'll keep on the blazing Who are the BBC dead? We're talking about Amazon. No time for medicals, cause we've fixed the game.