Advanced Pay Per Click (PPC) Certification Program

Certification Training
2370 Learners
View Course Now!
42 Chapters +

Managing Search Queries & Negative Keywords Tutorial

4.2 Introduction

Hello, I'm Brad Geddes, the author of Advanced Google AdWords, the founder of AdAlysis, and the PPC Faculty Chair for Market Motive. In this video, we're going to look at how to manage search queries and another keyword type, known as negative keywords.

4.3 Negative Keyword Overview

Now negative keywords are great to use in your account because they stop your ad from showing. This could be times when it's a word that does not describe your products or your services, it could be a word that doesn't convert. So when you want times but you do not want an ad to show you can use negative keywords, now negative keywords have specific benefits in your account. They should help raise your click through rate, because you'll stop your ad from showing on irrelevant queries. They should help your conversion rates because again, queries that don't convert, you can stop your ad from showing on them. They should increase your ad relevancy, and finally, they should lower marketing costs.

4.4 Negative Keyword Match Types

Now, negative keywords have match types, just like positive keywords do, but there's a big difference between the two. With positive words, you can often show for misspellings, plurals, or even similar words. Negative keywords do not show for these other variations. With negative keywords, the search engine looks at your negative keyword and does not match it to any other similarities, but what match types are for, is the ordering of the words to the query. So for instance, if we have a negative query like television, the search engine just says is the word television anywhere in the query, if so, don't show the ad. Now if we have a two word or a three word negative query, like LCD TV, the search engine, when it's broad match negative, just says, are these two words in the query, anywhere in the query regardless of the order? So if we have search queries like LCD Television, LCD vs Plasma TV, LCD TVs, with these negative keywords. Well, on the first query, LCD Television, the word Television in this the search query. It's our negative keyword. Our ad doesn't show. For something like LCD vs Plasma TV, we have a negative keyword LCD TV. Now in that search query, LCD vs Plasma TV, both the word, LCD and TV, are in the search query. They're not in the same order, but it doesn't matter with broad match negatives. It just has to be in the queries. But LCD TVs, it's plural. Our negative is singular LCD TV. We would show, for that search query, assuming we have a word that could be triggered that ad, because LCD TV would not block it, because it does not match to singulars, plurals, so forth. Only that word. Now you can use phrase match negatives. With phrase match, those words must be in the query in the same order for your ad to not show. Now a one word phrase match is the same as a one word broad match, because these words don't match to plural, singular, so forth. So, if we had the word Television in phrase match, it's no different than having the word Television and broad match. It's in our negative keyword list, it's in the query we don't show. However, for a term like LCD TV, in phrase match those words must be in the same ordering. In this second query, LCD vs Plasma TV, both those words are in the query but not in the same order. Therefore, we would show if we used phrase match negative. Now we can also use exact match negative. So, only if the query is that exact negative word does your ad not get displayed. So if we look at those same queries, we actually show for all of them. Only if we added one, LCD TV, which is exactly that negative keyword, does our ad not get displayed. So with negative keywords and match types, the only thing that matters is the ordering of the words in the query. If you don't want to show for a plural and a singular, or you don't want to show for a certain misspelling, you must add those as negatives.

4.5 Negative Keyword and Account Structure

Now, when looking at our account structure and where we can use negatives, there is a campaign negative list you can use and apply it to one or more campaigns. And that negative list is used in every campaign where it's applied. Now, you can also use negatives at the campaign level. You use Negatives at the Campaign level they affect the Ad Groups, and all the Ad Groups within that Campaign. You can then use Negatives at the Ad Group level as well. They only affect the keywords in that ad group. So, if we add it to this blue ad group and negative keyword at the ad group level, it wouldn't affect our yellow one, and vice versa. But, if we made it at the campaign level, it would affect all the ad groups in that campaign.

4.6 Campaign Level Negatives

Now so let's say then that we've been running our ads for a bit, and we start to get some information about our cost for acquisition. And we have a target CPA of $25. And we look at our search queries. And we have some search queries that are below our target CPA. Great, those are wonderful words. We have some that are significantly higher than that. So we have a bunch of options here. So, for the keywords that are just slightly above our target cost per acquisition, such as blue tennis shoes and women's tennis shoes, they're not that far off. We could just adjust the bids downward to hit the target CPA. Now we have two keywords that are significantly above our target CPA, trail running shoes and black trail running shoes. So, we have a couple options of what to do with these keywords. We could lower the bids dramatically, and hope to hit that target cost per acquisition. If you're a larger account with a healthy span, that's probably going to be your option. Now, let's say though that we don't have a huge budget and we really only want to spend budget on words that lead to good conversions for us. And we don't want to have to do all this bid manipulation where an ad may never really show for a keyword, in order to get to our target CPA, because there are other words we could use. Then, we could take those two words and make them negatives, so we don't show for these words significantly above our target. So, now when we go to add these, there's two options how we can make them negatives. So let's just make the assumption we don't want to show for these words. So we want to make sure they're negatived out. So, our first option is to add all non-converting keywords as negative. So we add the term trail running shoe, and we add the term black trail running shoe. If we added these as exact match negatives, we're not going to show for those exact queries. It's not going to match to anything else. Before we do this, we probably want to take a look to see if a term like, green trail running shoes, is actually converting. Because if we add trail running shoes as a phrase or broad match negative, then if a word like green trail running shoes was converting, we would actually stop ourselves from showing because we've made trail running shoes a negative keyword and any query with those three words in it we're not going to show for. So, that's consideration one is when you add these negatives, are you going to suddenly stop yet other words from converting? If so, then you probably want to use an exact match negative and not a phrase or broad match negative for a term like trail running shoe. Now, if we notice, both these negatives have the word trail in them. So, what happened here is we add, you know, trail running shoes and black trail running shoes as negative keywords, and suddenly, we start showing for something like trail running or trail shoes. And those words might not convert either. So if we look at our whole account. And we look for the queries with the word trial in them, and say, oh wow! When the query contains the word trail, our conversion rates aren't very good, or we don't have any conversions. We just want to stop all queries with the word trail from showing from our account. In that case, we could add just the one negative trail, and we would stop both those queries and all the others. And so, this is where, when you add negatives, you're going to stop your ad from showing on potentially more than just that word, unless you use exact match negatives. So, in some cases, that's good. If we look at our account and say, oh wow, the word trail, we don't convert for, we should just make this a negative. And use the word trail, you stop it from your whole campaign or your whole account, depending on the level you use it. Or, If you realize, oh wow, green trail running shoes converts, we would not want to use this word, because we'd stop our ad from showing in that query, as well. So when you add negatives, you'll be aware of the implications of not showing for yet more queries. Negatives are important, you will have some. But just make sure as you add them, you didn't stop other good queries from showing.

4.7 Before Adding Negatives

Now it can be also useful to just take a step back before you add negatives and say, is the problem a query and landing page combination? Or a query and ad combination, and it's converting somewhere else. So for instance, if we have this search term Trail Running Shoe and we realize that it's actually going to two different landing pages. And one landing page has a $25 CPA, which is our target, and one page is the $76 CPA. In this case, it's not the matter of stopping your ad from showing for Trail Running Shoe across your whole account. It’s from stopping it for that landing page, whether it's adding it to just that ad group or removing that landing page being associated to that query. It's also possible there's an ad problem. So if we look at the query and say, oh, when it's shown with Ad 1, we have a $19 CPA. When it's shown with Ad 2, we have $109 CPA. So in that case, it's a matter of removing Ad 2 from showing from that particular search query. And so we'll look at a higher level of your account if you have a query showing for multiple ad groups. because that's sometimes the issue is you're using some modified broad match or some phrase match and suddenly a query is showing from multiple ad groups. And so in cases like that, that's when ad group level negatives are really useful. So campaign level negatives, great for stopping an ad from showing from every ad group in the campaign.

4.8 Ad Group Level Negatives

But sometimes you want it shown from one ad group but not from another one. That's when ad group level negatives are good. Now, ad group level negatives only stop ads from showing in that specific ad group. So, let's say we're a hosting company, and we have one ad group that's about hosting. Online hosting, hosting type of keywords. We have a different ad group that's website hosting. Let's assume the website hosting keyword is another ad group. Another ad group has VPS hosting and virtual private server hosting in it. And yet another ad group is about cheap hosting and discount hosting. Now that's a decent segmentation for a hosting company. But let's say we have a search query. Like cheap VPS website hosting. And let's assume the keywords in these ad groups, are the ad group name modified broad match. In this case, every ad group could be triggered. Hosting ad group, if hosting in the query. We have a website hosting ad group, website hosting's in the query. We have a VPS Hosting Ad Group. VPS Hosting is in the query. And we have a Cheap Hosting ad group. Cheap's in the query. And so, suddenly you don't control which ad is going to be displayed with this query because all four can match it. So when you have these types of situations, ad group level negatives are great to use because we could take our VPS hosting ad group and say well mine is cheap because we've got a cheap VPS hosting ad group. We could take our cheap hosting ad group and say mine is VPS so if the query contains VPS, don't trigger this ad group, we could have yet another ad group, web sit hosting, minus cheap and minus VPS and find our hosting ad group and then more of these. So by layering these negative what we can do now is force the search engine to show the most specific ad for the query and you can take back your ad serving control. Now not everyone has this problem.

4.9 Pivot Tables

Something we will talk about throughout these videos are pivot tables. What pivot tables do, is let you aggregate data or pivot data based upon some key. So, what we could do is take our query data, put it in a pivot table and then say, for my search query, which is the row label, the search query. How many ad groups has this query shown from? And if you've got a query that's being shown from 13 different ad groups, or eight different ad groups, then you might not be in control of your ad serving, because multiple ad groups have shown for that particular search query. So then you probably want to use some ad group level negatives, so that you're control of ad serving, of which keyword will be triggered by a search query, what ad is going to be used, and what landing page will be used. Now, it's very easy at an account level, which could be 20 campaigns, could be one, to run this kind of data and say how many ad groups per query. But there are times, ad group could be more than one and it's not a problem. This is when you might have geographically separated campaigns. You might have a campaign to Los Angeles, a different campaign shown to San Francisco. Both these campaigns could have the exact same keywords in them. They are shown to different regions. So just because a query at an account level has come from two ad groups, when you dig into it, it's actually shown once from our LA campaign, once from our San Francisco campaign, they don't compete so it's okay. So if you use the same keywords in multiple campaigns because they're not competing against each other, then you need to add a campaign level segmentation to this kind of data. If your campaigns don't compete, then just saying, here's my search queries. How many ad groups have they displayed from? If more than one, I need to use ad group level negatives. If no, then I don't need to.

4.10 Cost Data

If you use modified broad match at all, and especially broad match, you're going to have some ad groups that are more than one, some search queries that are shown for more than one ad group. And if it's been one impression or one click came from a second ad group and it cost you 18 cents, that's not nearly a big of a deal as if 50 clicks came from the wrong ad group, and it cost you 200 dollars. So, when you go through this kind of exercise, look at your cost ad. It can be useful, because if you've got one ad group that $200 has served incorrectly, a different one, it's 20 cents has been served incorrectly, you want to spend your time on the higher priority, higher spend ones. And you might not spend an hour diagnosing 18 cents of incorrect spend. Usually not worth your time, better things to do with it. So, campaign level negatives, are for stopping a query from showing for the whole campaign. And potentially multiple queries showing for the whole campaign. Ad group level negatives are to ensure that this ad group doesn't show for a query, but a different one probably does. And then, when you start to get data, you can look at your query data by ad group, and say do we have certain ad groups that are multiple ad groups serving the same query. And they're spending a lot of money. If so, then we need ad group level negatives. If no, we don't need ad group level negatives, just campaign level negatives.

4.11 Match Type Organization

So, something that is important to consider then, is how are you doing your match types? So, if you have a campaign. You've got a broad match ad group, and a phrase match ad group and exact match ad group. Now, if you assume you've got the same keyword in all three ad groups, just different match types. Well, what you need to do in your broad match ad group is add all those same keywords as [INAUDIBLE] phrase match. And your phrase match tag at all those in keywords exact match and make sure those keywords are in your exact match ad group. This sort of setup looks good when you create it. But you will add keywords over time, you remove keywords over time and in veritably, this setup starts to disintegrate, because you added some new broad match keywords and you forgot to add the modest phrase matches. And there's not an easy way to catching up on this organization. So, this kind of organization is rarely recommended, unless you have a pretty small account. If you want to segment out your match types by different ad groups it's better to use different campaigns. You could have a campaign that's all exact, a different campaign that's all phrase match uses the minus exact match keywords. Another campaign that's all broad matched uses the minus phrase match keywords. That's one way to do it. Another way is using a Discovery and Managed Campaign campaigns. So, for instance you start a new account and you don't have any data yet, so you're not sure what's going to be your best word, what's going to be your worst word, you don't have this information yet. So, you got a healthy budget, maybe not exceptionally large, but not tiny. So, you start with all modified broad match words. And then, what you do is you look at the search query data. And you say, okay, does the query convert? If it is converting, then you add this word to your Manage Campaign, and in a phrase or exact match type. And then, you make it a negative keyword in your Discovery Campaign. And so, if you do this over time, what you find is that your discovery campaign is for finding good queries. Your managed campaign is only for queries that are converting. Then, if you look at a query in your discovery campaign and it's just not converting, you make it a negative key word. So, this is a pretty common way for mid sized accounts to be managed. Now for really, really big ones, it doubles your campaigns, so you won't see it as often. But if you're spending $10,000 to $100,000 a month, it's a good organizational structure. If you've got 300 campaigns and you're spending $1M a month, it's not as good as just starting by match type spy campaign and we will do a whole other section on really in-depth campaign structure and account structure which matters. But as you walk through key words some high level thoughts about management because match type are important too, to consider for your ad surfing. Now, it's okay to have a campaign with your ad groups and use all of the match types inside of one ad group. This is really common if you're doing cost per action bidding, especially, if you use say, an automated tool from a search engine. The only thing to remember if you have all the match types within a single ad group is that your exact match should be the highest bid, then your phrase, then your broad, because exact match, you know what someone's looking for. This should be a better conversion rate than your phrase match. Your phrase match should be a better conversion rate than your broad match, because you don't know as much about the user query with broad match, because they could have your words in different orders, different queries could go with this. So, it's okay to mix some of the ad group as long as you're just controlling the bits.

4.12 Examining Query Data

Now just remember, broad match does not convert higher than exact match for the same word. What happens if you see stats like this that our broad match word has a better conversion rate and a lower cost per conversion than our exact match, the issue here is. You know what someone typed in the search engine with exact much, it was that word or a singular plural misspelling of it. Your broad match work could be matching for a lot of different varieties. So if you don't have words in the campaign that aren't controlled in the ad survey, what's really happening is we can have some other words here. Coffee mug blue, coffee mug red, coffee mug, that aren't in the account. They have good conversion rates, but they don't match to the exact match keyword because exact match of course is a very closely related. So the stats get aggregated to the broad match word and makes broad match look better than exact match. But if we added in this case blue coffee mug and red coffee mug as exact match words in that account, then those words would get the statistics. You could see the difference in those conversion rates and what you want to bid for it and then broad match would go back to what it is as finding these other queries not your best one. So whenever broad looks better than the exact metric standpoint, you definitely need to look at your search query data, what the user actually typed on the engine, because you have keywords you want to add to your account. So always go to the queries, look at the queries, look at their conversion rates, look at their cost per conversions. If they don't convert at all, make them negatives. If they are converting, then make them actual keywords.

4.13 Aggregate Match Type Data

Now another way just to sort of view your account, is to get an idea of how all your different queries are doing your matched types. Is you could make again a pivot table and aggregate it at the match type level. You'll see here's our conversion for broad match. And then if you're using true broad match, you may see something known as session based matches. This happens when someone does a search, they see your ad, and they don't leave the search results. They search again for something that could be unrelated to your ad completely. But because they're in the same session, an engine may show an ad back to that user based upon their first query. And so, those are session-based matches. And you can't remove those. If you use broad match, you may see session-based matches. Then you can see your exact matched data. And then close variants. So close variants are the misspellings, the singulars and plurals you can show from. Your phrase match and your phrase close variants. And what you usually see is your lowest cost per acquisitions are exact match. Your second lowest phrase match. Your third lowest is modify broad match and then broad match. And that kind of gives you a high level picture of how each match type is doing for you. And again, if you see broad doing really well, you have query data you need to go look at, and add keywords.

4.14 Search Query Flow Chart

So from a workflow standpoint, this is really how to work with it, is look at the query data, what the user actually typed into the engine. And if you're using the conversion tracking systems the engines provide, you can see your cost per conversion, your conversion rates, how often a user is actually taking an action, by each of these search queries. And then you ask yourself a simple question, is it converting? Well if it is converting, then you need to say, is there a better Ad Group for this? If no, then just make that query a keyword in that existing Ad Group. If yes, there is a better Ad Group or you could have a better ad for it, then create a new Ad Group or add one. Now, this assumes you're adding the keyword to that campaign. If you use a discovery and manage campaign, then you look at your query data and say, it is converting, let's make an exact match on our other campaign. But then look at your query data and say, oh, it's not converting. So the first question is should it convert? Is this a word that really is descriptive of your business? And if you think, you know what, this word should convert but it's not. So what you want to do, is not just remove it because it should convert. You need to put it in the correct Ad Group, create an Ad Group, control the ad in landing page. Make a note now to yourself, just to remind you to go look at this data after a week or a month, depending on how fast you aggregate data. Say, okay, now is it converting? because if it is now converting, great, you have a new source. Now if you look at the word and say should it convert? No, right? This is not related to what we do, it's ancillary information. Then that's when you make it a negative keyword. So this is a good workflow to use as you look at query data.

4.15 Other Notes on Negative Keywords

Now just some other notes on negatives. Campaign negative lists are really useful. You can make a campaign list or multiple campaign lists. Put words in it. Apply that list to one or more campaigns so if you find yourself adding a negative key to campaign one and something campaign two shows the same query. Make it a negative in campaign two. Campaign three show for negative lists because they're shared between campaigns or to stop this from happening anywhere. When you add negative keywords, some negatives will be obvious. Like we just don't do this, it's a negative. Other times, if you're not sure, make sure you have data. I see companies all the time, worried about negative keywords, and they first add words like free. Well we're an e-commerce company, we sell stuff, free is never a good word. But if a user searches for your product plus free shipping. And you offer free shipping, that is a good word. Free does convert sometimes. So if you are not sure if a word is going to convert or not make sure you use data before you just add negatives. In search engines, you often can't see things like interactions metric by query. You can see them by keywords if you connect say, Google Analytics or an analytics package to your systems. So for instance, if someone looked at you and said, well this word has a 100 clicks and no conversions. I'd be tempted to add a negative keyword. But that isn't much data wise. Now someone looked at you and said, well this word has 100 clicks, and it has a 97% bounce rate, meaning someone got to your website, and immediately left. Okay, those visitors really don't care about your offer. That should be a negative keyword, or at least a different landing page. Some adjustment to that query needs to be done. But if you looked at it and said okay this word has got 100 clicks and a 10% bounce rate and an eight minute timeline site visit, that's great, that's wonderful metrics. So you probably want to make it a keyword and just monitor it. So if you are using say Google or Bing and you're tracking your information in Google Analytics for a different analytics system. Take a look at your query data in the analytics system because you can often get better insights into some of those words to see if they really are great words that just need to be nurtured some or if they should be negatives.

4.16 Recap

So to recap, first remember your negative keyword goals. Negative keywords should help increase click-through rates. They should help increase conversion rates. They should lower marketing costs. If they're not helping you want to make them negatives. But always be careful because click-through rates and conversion rates are ratios. So you want to watch total conversions. So let's see you add a negative, and your click-through rate goes up, your conversion rate goes up, but your conversions per day goes from 100 to ten. You added something that you probably shouldn't have because you just lost most of your conversions. So in that case, as you add negatives watch your bench marks. And they're also great for controlling ad serving. Multiple ad groups could show for a query. Those are good ad group level negatives. So if you ad keywords and these things don't happen, click-through rates don't go up or conversion rates don't go up, total conversion drops, you want to revisit those negatives. So, when we think in negatives use your search query data. What the user actually typed into the engine to find non-converters spending money. Now first consider if a word should be shown from a different ad group. If it should be in a different ad group, then add that word as appropriate. Now if you look at the word, and you say I never want to show for this query again. That should be a global negative. That's great for campaign negative lists or campaign level data. If you want the engine to show from the correct ad group when multiple ad groups in the same campaign can show you for a query. That's what ad level negatives are for. Now if you want to make sure that the correct campaign is being displayed or you don't want a keyword to show from a specific campaign then you want to do campaign negative. Then finally, when you add negative keywords, make sure that you either a, use a correct match type for that negative, or they use the root word. That trail running shoe. If green and black and blue trail running shoes don't convert, then you might just add minus trail. So if you don't want multiple campaigns to show for a query, that's a campaign negative list. If you don't want a specific campaign to show for a query, that's a campaign negative. If you don't want a specific ad group to show for a query that's an ad group negative. And then as you add the words, add negatives, always consider should it be in a specific match type or can it just be broad match? Which is the most common negative used or should we use that query or is there a root word? That describes multiple crates you want to shuffle. So I'm going to add that instead. So negatives deserve a lot of attention. Positive keywords make your ad show. Negative keywords stop your ads from showing. If you use any kind of Broad Match or Modified Broad Match, which are great to use, to reach a large universal queries. There will be queries that don't convert. It's going to happen. So it's not a bad thing. It's just a matter of understanding what these words are so that you know it's negative to add so that you can lower your marketing costs and make sure that you're using negatives and managing queries correctly to hit your marketing goals.

  • Disclaimer
  • PMP, PMI, PMBOK, CAPM, PgMP, PfMP, ACP, PBA, RMP, SP, and OPM3 are registered marks of the Project Management Institute, Inc.

Request more information

For individuals
For business
Name*
Email*
Phone Number*
Your Message (Optional)
We are looking into your query.
Our consultants will get in touch with you soon.

A Simplilearn representative will get back to you in one business day.

First Name*
Last Name*
Email*
Phone Number*
Company*
Job Title*