How to Optimize a Google Shopping Feed from Search Terms

Welcome to today's video. Today we will look at how to do feed optimization properly. So this video is all about feed optimization mastery for Google Ads and I will show you how to yeah check your GMC, make sure everything runs smoothly so you don't run into any kind of issues on account level or product level so everything stays live, you have no issues. Then of course we will look into how to optimize your feed for improved targeting, better click for rate, better conversion rate, which in turn means higher ROAS and more profits for you. All we want, right, the GMC that stays live and more money in the pocket.

All right, let's go. So like I said already, what we are optimizing for basically we want to avoid any kind of issues in the ChimC. So we check all kind of errors, disapprovals, how to fix them and then basically how to improve the targeting, the relevance of our product listing and also the clarity to have like better click rate and in the end also better conversion rate, right? Because we don't only want to attract the right customer, we also want to make sure the wrong ones doesn't click on our ad and yeah cost us money and then don't convert, right? So relevance and clarity and that's pretty much it.

So a couple of non-negotiables we will have in the video is basically yeah obviously fix the errors first and then we optimize for performance. Then I will show you how to optimize for better keywords in titles and descriptions without keyword stuffing or using random keywords or guessing what works well. And then obviously like never include any fake info or salesy copy in the titles or descriptions that leads to account suspension straight away. And yeah, always keep the feed consistent with what you have on your landing page or in the back end. Also to make sure like everything stays live and you don't run into issues.

So without further ado, let's dive straight into it. Quick workflow overview. I will show you on how to check things. First of all, you always go to Merchant Center and diagnostics and check if you have any issues. Right? We said we want to check that first. Do we have any account issues? Do we have any product issues? Can we fix them? What's up there? Remember, especially with big feeds, there will always be something, right? You always have some minor issues mostly due to syncing errors which gets resolved automatically usually with the next reync. You can also do a manual reync if there are a lot of products disapproved or your best sellers are disapproved for some reason and that's pretty much it right so don't overthink it again there will always be something right like small syncing issues crawler issues where you have like for example I don't know some products disapproved for uh landing page unavailable or like some price mismatch right most of the time like actually nothing is going on and again if like 0.3% of your feed is disapproved today.

Just reync, wait for the next reync and it will be fine, right? So don't over complicate it. Just make sure it's under control. The next step two, you will look at your feed and just a quick health scan. Basically, we will categorize in top sellers. We will look at our wasters, right? That are products that have quite some spend but low performance. Then you can also look at products with some clicks but low conversion rate or like high impressions, low clicks which means low clickthrough rate, right? So we can see all right what do improve there but mostly we want to look at our bests sellers of course and we want to look at the wasters that have high spend but don't perform and then we will always check all right can we improve something with the title can we improve the description can we improve the image quality price competitiveness right so super important for Google we know

pricing is quite important are we competitive in the auction even can we optimize the pricing and push for more volume And then we also will look at yeah product category, product type like does that match is that aligned can improve there and even like other things right can be improved product highlight certain other attributes in the feed that can be improved but those four I would say even like they are the most important by far right and also something I want to add here straight away even with the top sellers don't overthink it because if they're already selling well and performing well for you, you probably shouldn't like tweak too much.

That can backfire, right? So, you want to mostly look at yeah, even those products like the wasters with high budget, low performance, of course, and even products with clicks that already got some clicks but low conversion rate that can lead into a future waster, right? So they have like usually relatively low spend still, but those are kind of like the and even sometimes products with high impressions and low clicks because that's also something like they show in the auction but nobody clicks. You want to think about it and think why, right? Because like they get shown like your ad get shown but nobody clicks. So maybe image quality, right?

or image clarity like it's not fully clear what it is or even a title where it just shows in in the wrong auctions. It's not good for you either. So that that's kind of like the the products you want to focus on. Your top sellers usually need the least attention of you because it's already working quite well. All right. And then like I said, step three, we apply those fixes and optimizations. So we want to check for errors and disapprovals first of course like we said. And then title framework. Can we optimize the title, descriptions, certain variants, categories, images, pricing, and last but not least, you can also like add labels to like put them in certain category for yourself.

We do that differently in the agency. But yeah, whatever you want to do there basically. Cool. To keep the video short, it's simple and straight to the point, I will uh skip the diagnostic checklist pretty straightforward and I will dive straight into the attribute mastery. All right. And the most important thing I want to check out with you here is keyword mastery. Right? So that's something I get asked often and to be honest we never had the proper SP on that that really like shows you okay where to get it, how to get it. I always gave a couple of different options but here's the right way in the right order where you can actually get the right keywords without guesswork and with avoiding all the common pitfalls and mistakes.

So again, the goal is include actual search terms people use. Super important, right? Like actual search terms based on data. So without guessing and of course without keyword stuffing, right? That's that's the worst because that leads to turning titles into unreadable garbage. We don't want it either, right? It just doesn't make sense. So actual search terms based on data without guesswork, without keyword stuffing, and without making it look like basically and also remember this is not SEO so we are not optimizing for that it's very similar in a lot of ways but we are optimizing for shopping query alignment right so we run paid ads here it's not sale important so you always want to keep that in mind if you compare that all right first principles first again like I said no guesswork we do not invent keywords super important right that's why we'll go through it here step by step you extract them from real demand based on real data.

A keyword rich title is built from observed search behavior. Again, actual data aligned with how users describe the product where you also keep like the human aspect of it in mind like how do people even look for

it? Why? What's the intent behind it? Right? That sometimes the algorithm doesn't pick up. So, you want to keep that in mind as well. And we will structure it basically into two main categories for the keyword optimization which is primary and secondary modifiers. A primary modifier is the descriptive aspect of what it is. Right? Let's say let's take running shoes as an example. Running shoe would be the primary modifier. That would be the main keyword, the root keyword, right? What is it? Running shoe. You could even say shoe, but that would be almost too broad. So let's say running shoe and then that's the product category the product type right and then you have secondary modifiers which could be gender so we have running shoes for men running shoes for women could be uh the weight quite common for sho running shoes as well right you have lightweight running shoes for example then they can combine it like lightweight running shoes for men right and you can see already where we go with it so the primary modifier or the primary root keyword is like the main keyword, right?

Like what is it actually about? With the secondary modifier, you niche it down a bit more to make it as relevant and as accurate as possible. So the secondary modifier should like make it more detailed, but only with the goal. That's important to like make it more relevant for our actual target audience. Like if it makes sense, if it does make sense and you add random modifiers that are actually not relevant, then it's not good. Then it's like pointless. So it should make the targeting and the relevance of the listing like on point. If it's random stuff, don't add it. Then no need to again no need to keyword stuff in that case, right?

If it's, let's stick to the running shoe example. If it's just an ordinary running shoe, unisex even, right? Like doesn't matter what gender, there's no specific attribute about it, like nothing, then just keep it simple. Then you have running shoe, you have size, color, the brand maybe, you know, that's it. No need for like any weird modifiers. In a lot of cases, you have something that makes sense though. Keyword stuffing, no long keyword list, no spam. And again, don't guess what's relevant. Actually look at the data. Actually make look what makes sense for this type of product you have. Do it properly. It pays off.

All right. Next, most important, where do we actually get the correct keywords from? We said we take real data. So where do we get the data from? Right? This is in the right priority order. The most important first, which is Google Ads, of course. Obviously, because we are optimizing this for our ads, right? So, where do we get the most important data from? The most relevant data, the gold standard, the best signal in Google Ads directly. That would be the search terms report. In that case, let's say we run shopping campaigns, BMAX. We want to look at our search terms, right? And again, we look at different things here.

We want to see what converts best, what's trending, what has the most potential, what gets a lot of clicks but doesn't convert well, right? So, you want to look that not only for the ad optimization but also for the feed optimization. Super important. Where do we get it from? Like I said, search them sweepboard basically for your shopping campaigns. BMAX perfectly fine. And again, we look at what converts, what gets a lot of clicks, what gets a lot of clicks and sells well, what gets a lot of clicks and doesn't sell well. You want to look at it from different angles, right? To get a clearer idea of it, to see patterns and to just see what's going on in general.

Yeah. Search terms that clearly describe the product, you will get some good ideas from that. And sometimes it's

really, really surprising. I had it before. or I had products that sold super well, had a certain title. It was a winter jacket. I remember it quite vividly. Then we looked at the search term and the search terms that were converting were something completely different than what we actually had in in the title. That was quite interesting. Was still converting well. But the thing is, you should also optimize the title and your listing for exactly that in that case then, right? And I said before, usually we don't really touch best sellers if it's already working well. In that case, we actually optimize the listing for that specific search term.

We could even get more out of it. I think it was around like 20 30% more output if within one week. And I'm saying output. This is quite random actually. It was like a mix of sales volume and rowers. We kept around the same rowers in the end because we were bushing. But we had like around 20 to 30% more sales volume with around the more or less the same rowers, right? Doesn't matter. Short [snorts] story, but in the end, what you really want to have is like you want to make sure that's that's matching and that makes sense. And even for other products, you will get good insights on all right, what converts, what doesn't converts, and yeah, just what people are looking for as well, how they are searching for.

So what to look for again product type terms, right? So that's the primary modifier. Then you have the secondary modifier which is intend modifiers like we said before. Could be gender, could be um the material, could be the USB of the product or like the function, could be size, could be color, depends on the product as well, right? But the style descriptors could be also secondary modifier basically the same like size and gender. Of course, the main thing if it converts it belongs in a title. So you want to check that out primarily, right? Then also in the Merchant Center, you get some good search queries basically similar to Google Ads.

Basically, you see what people are looking for, right? But not necessarily clicking is quite interesting as well. Basically similar to Google Ads, but you can check the query data there as well. Helps you to understand primarily how Google serves your feed, right? when your product show why they show what people are looking for again and you want to see if your primary keyword if that's matching already if that makes sense another powerful research option is Google autocomplete right uh that's a very old trick but still works quite well so you type in your core product name your primary keyword your primary uh modifier like we said basically and you see what's coming up right out the complete suggestions like here for example silk pillowcase and it gives you a good idea of like okay what gets actually searched for.

All right. And then two more. Then we have obviously like check out competitor shopping listings. You can search for your own products. You can search for similar products. You can search for competitors and you will see what's coming up. Right? I would recommend not looking for your own products too much. Like that's something that's just doesn't really make uh too much sense. And then you start over thinking, oh, why do I not show it first place or second place? Like if you look for your own product yourself and you do it multiple times, you know, you get weird results. So don't really do it with your product too often.

Just look at your competitors primarily similar products. How do they do it? Right? Just try to learn from them like do they have a different primary keyword for the same kind of products, right? If so, why? Especially if you start to see patterns. You always want to look for patterns in that case, right? Look at the top ranking listing. Look at what they do differently, right? try to be quite objective like is it the product image is that different is the price maybe something that could lead them to have more success right try to look at it from different angles basically and again mainly look at the

different angles what they do how they do it what they don't do as well right and then learn from them but obviously don't copy oneonone especially if it maybe doesn't make sense for you try to do it even better don't copy like weird promo language especially if you don't have then you get a misrep and avoid yeah lies or overexaggerations even if it works well for them I wouldn't do it can always lead to an account suspension right so we don't want that basically look for patterns learn what other people are using it will help you to get a good feeling for it as well and then last but not least we have certain keyword tools and that's usually always something we think about first right and something I also like always mention as well like keyword planner Google trends certain co tools but really you want to look at the data that you already have, right?

And what other people are doing as well because that's closest to the source and backed by actual data, especially your ad account data. But still super interesting sometimes. I love Google Trends, I love keyword planner, certain co tools to give you like an idea of like search volume, what people are looking for, how they're looking for things, but again, not even close to your actual own data. And almost nobody does that. So that's like the real hack, right? the real secrets on how to optimize it properly. It's quite some work to be honest. That's why you cannot do it for every product properly, but it pays off right especially for your top sellers and for your wasters.

You want to look at that stuff. And something that's super important here for me to mention is just because a certain search term or keyword has a lot of search volume doesn't mean it's a great search term, right? Like that's a that's You don't want to look at something that has like the highest search volume. doesn't mean it will get you the most amount of sales. Sometimes it can be a very niche type of product, a very niche type of keyword that brings you the most sales with a good drawers as well, right? Because also especially you have to consider search terms with high search volume usually also have a lot of competition.

And again, you also want to go for relevance, right? It has to be relevant to your product listing. If you go super broad, you usually target everybody and then also like the quality of the traffic is lower. So you want to be relevant, right? Remember the running shoe example for example, you don't want to have just a random running shoe. You want to be like all right, which gender, what material or like what attribute for example. And even if the search volume is done lower, you are like way more relevant to your actual target audience. And in turn, you have a better converting product here.

Super important, right? Forget the volume. Like that's something I don't even care about or not that much at least. Of course, there needs to be some search volume. If nobody's looking for what you're selling, you will have a hard time selling it. But I don't care if like 500,000 people are looking for it every month or if it's like 50,000 or 100,000. As long as people are looking for it, I can make sales, right? So that's super important. Like I cannot stress it enough because I know everybody's looking for this like high search volume keywords but I couldn't give a how many people are looking for it.

So super important right? All right next how to decide the primary keyword. I mean basically we went over it already right first of all data. You want to look at the data but also think from the human side of things again like how would someone describe it right? It's usually like very easy to describe. Nothing fancy about the primary root keyword like all right buffer jacket, running shoes like we said, silk pillowcase, standing desk. It's usually like two, three words max, right? The primary keyword is like very very simple, very straightforward. A lot of people try to over complicate it, but it's honestly very very straightforward.

And then the secondary keyword, like the modifiers, they can be a bit more complex sometimes, and that's where we really have to do the research. And super important, I cannot stress that enough. Let's go over it. Modifiers are added only if they change buying intent,

if it makes it different, right? For the running shoe, for example, like if it's a running shoe for men or women, that would change buying intent, right? Or even like the different attributes. Like there's so many different running shoes for different type of cases, which obviously changes the buying intent, right? If I'm looking for that one specific running shoe, that type of attribute, I wouldn't buy the other, right? I couldn't care less about the other one then. That's super important. Edit if it's relevant for the yeah buying process. If you change the buying intent, if it's a commonly searched keyword and if it's accurate, do your listing as well.

You cannot just add random stuff because you think it sells well. Cannot even offer that that thing, right? Whatever it is. So, comma modifier categories. Again, it's usually it depends on the niche as well. It's like gender, size, material, certain use cases. You USBs and yeah, it depends on the niche as well. You know, can be can be a taste for supplements for example the taste super important especially for for protein powder for example, right? You have like certain taste that's more uh more popular than others like chocolate, vanilla, whatever. So, it depends on the niche as well. But again, don't over complicate it. Just straight to the point.

And I like this rule as well. If removing the modifier would attract the wrong buyer, include it. Right? Like that's a good modifier. Like adds to the relevance and to the clarity of your listing. All right. And then in terms of title structure, I don't want to go too much over it. It also depends sometimes on on the niche and like what's what's best. But very simple. Primary keyword, key modifiers first, and then you have like variant titles, the brand at the end. That's pretty much it. You can like play around with it a little bit. have the brand first. For example, I would have the brand first if it's like a very popular brand, right?

For example, I just see here Nike running shoes. You always want to have Nike first, right? If it's like a brand with that gravity just makes sense. So, I will skip that. It's like pretty straightforward. What I want to show you is the keyword density rules. That's important. Of course, only show each keywords once. Doesn't make sense to have it twice. No repetition. And also don't stack synonyms because Google understands that. All right, if you talk about checks, you talk about checks, right? You don't need like different synonyms for checks. I don't even know here for example putting code there, right? If it's not a code, it's not a code.

Don't try to like add similar type of search terms just because it's close to check, right? So yeah, that would be that would be an example of that like putting checket and code in the same category like is it a checket or is it a code like stick to one thing, right? Things like that that also like dilute your traffic quality and then don't be surprised if it doesn't work, right? And I saw that type that type of stores actually where people like really thought they optimized the keywords nicely because they included a lot of stuff and they they had all these search search terms but in the end it was just like a bunch of right?

So we cannot do anything with that. All right, how to actually implement this? I mean also pretty straightforward, right? I will not like tell you how exactly to do that. It depends on what workflow you prefer. But you know what's important? You know how to do it in the end. Ah there's just one thing I want to mention it here because I see it here right now. That's quite important. Never hardcode keywords that already exist as attributes. Right? So that's something what what that means is basically you have an attribute for brand. You have attributes for size, color, gender, even material. Right? So you don't want to add that in your original product title because you can add those things with your feed optimization as an attribute.

Right? So you basically take your keyword, your main keyword which would be the primary keyword and one of this modifier for example, but then you can add size and color as the variant title. Then you can add gender, then you can add the material. Then you can add the brand at the end without including that in the initial title, right? You can

just stack those attributes. And yeah, that's just a cleaner and more rulebased version on how to do that. And that's also how we do it in the agency, right? doesn't make sense to like stack all that stuff in the initial title would be quite stupid to be honest as well. Doesn't help you with the performance. All right. And last but not least, let's look at the most common mistakes that I see with the feed optimization. Again, don't guess keywords. Look at the actual data. What works? Don't use SEO keywords. Again, there's a difference between SEO and shopping. like don't just randomly use SEO keywords. Actually look at keywords with buying intent.

Right? So there's a difference. There's absolutely difference and you have to keep that in mind. Don't add random product attributes that don't apply for your product without any proof, right? That stuff can lead to suspensions. Don't mix multiple product types or multiple keywords. Don't keyword stack randomly. Make sure it makes sense. And again, don't over optimize it just to have a long keyword or a long title in that sense. I mean, and then it also again doesn't really make sense, right? It's just like synonym stacking and it will hurt you more than it benefit you honestly. Final rule, a keyword rich title is not the most keywords possible, right?

Like we just said, it is the clearest reflection of real buyer language. So that's what we focus on the buying intent, not search volume, not sale, not guesswork, real buyer language, right? Like we want to sell something. Keep that in mind. The clearest reflection of real buyer language placed in the right order. Isn't that beautiful? That's basically everything it is, right? All right. And that's pretty much it about how to optimize for keywords this time. like actually super super detailed. I hope it helps. I will not go tier one, tier two, tier three like you can check it out. I will attach the notion to the bottom of the video.

All right, two more things. We keep it short and quick here. Feed rules, what to do, what not to do. So very short, just a quick overview. Like I said, optimize top spenders first. So top spenders in terms of high performers and wasters. 8020 rule, right? Especially if you have big feeds, keep it simple. Keep your taxonomy consistent, your naming consistent. Have certain rules, certain patterns that apply that helps with optimizing big feeds. Use supplemental feeds honestly for surgical overrides if it's like title and stuff like that. So we always have like our main basic feed that's connected with a feed to like the Google channel app or simp.

And then I like to use a sub feed where optimize the main feed, right? without touching the main feed too much. It's a preference thing. You can optimize your main feed to a certain degree. I like to keep the main feed untouched and then I have a sub feed where I can add all that stuff. It makes me more flexible, right? But in the end, it doesn't matter. Just giving you that two options here. Basically, yeah, keep track of your changes, especially if you test and you want to test the differences and how it works out for you. And like we said before already, standardize your naming conventions, right?

have a certain rule on how you work with your feed. It's very important, especially with big feeds. Again, don't just rewrite the entire catalog at once. Randomly focus on your bigger spenders first. Don't push promo language. It leads to disapprovals. Don't change product type and category mapping every week. But generally speaking, don't make too much changes, right? Like that's that's again where you think oh you're doing something you're optimizing it then you make so many little changes that it actually like never can get out of the

learning phase right so that's super important as well yes you want to optimize yes you want to improve it but don't make like mini changes all the time because then you just keep yourself stuck in the learning phase death by overoptimization right it's a real thing and I see it very often with beginners who think they always have to touch something all the time and Then don't use misleading attributes or like certain search terms to chase traffic, right? To chase more volume. Again, relevance, quality over quantity and search volume. Super important. And then testing framework. We briefly touched it already. So you don't want to overoptimize.

How do you test properly? Right? And keep track of the improvements, the potential improvements as well. So you only test one major thing per SKU because if you do a lot of changes at the same time, you don't even really know okay what actually improve the listing, right? Unless your listing is completely and you know that like if everything is bad of course like please for the sake of your performance just improve everything then. But if it's just about testing little tweaks just one thing at once otherwise it's not a good test anymore. And what does good look like? Basically, right? What are we looking for?

It's improved clickthrough rate, improved conversion rate even sometimes or at least we don't want to collapse our conversion rate or roars. Then query quality improvement, especially if you see a lot of unrelevant search terms that don't really make sense for you. And again, just like the main thing, right? no CPC spikes if potential CPC decrease because of a better quality listing, stable conversion rate, stable rewards, and always remember to test it for an extended period of time depending on how much traffic you get. But yeah, usually at least one to two weeks if you don't have a lot of spend yet. Also, don't test things, right?

If you have a budget of like 40 50 a day and you get a couple of clicks per SKU, like you cannot test anything yet. You need a lot of budget, a lot of data to also test this properly. Otherwise, the testing period would be like too long, then you have to extend the testing period. Doesn't make sense. Also important, if clickthrough rate is up but conversion rate is down, you probably made it a bit too broad. That's also where the keyword stuffing comes into it, right? Where you have like great, you have all these like high search volume keywords, right? Brilliant. Now you get a lot of clicks don't really make sense.

You have poor quality of traffic, high bounce rate, which leads to higher CPC. You made it too broad, too misleading, and nobody buys. Then you come to me and you tell me, "All right, Chris, I have this like optimized website. Everything makes sense. It's a proof winner. I have a great product, great pricing, great website, and still nobody buys, right?" And that's something you want to look into it then because then you have high click rate. You probably get a lot of unrelevant search terms and your conversion rate is down. Nobody buys in the end, right? You don't want to be attractive to everybody, right?

You want to only find your ideal target audience, the hot market, the ready to buy audience that is willing to buy your product because like you don't want to like and that's I think that's the main thing that a lot of people get wrong with keyword optimization, right? That's exactly why they're looking for, oh, what has the biggest search volume? What has the biggest volume, the biggest volume, the biggest volume? Again, what about the volume, right? You only need like a very small group of ready buyers that want to buy what you have to offer and that gives you the results that you need, right?

You don't want to like get a lot of clicks from everyone from people that are not willing to buy anyway in the first place. So, what's the point, right? You want to have sales. You want to have a good rewards. So, that's what we're looking for. And yeah, that's it for today, ladies and gentlemen. I hope that helped. I think it's a super valuable resource and I hope it helps you to yeah crush it and get some good results. So that's it. See you in the next one. [music]