Adwords ( and other PPC ) Master Thread

eliquid

Digital Strategist
Joined
Nov 26, 2014
Messages
1,072
Likes
2,266
Degree
4
Day 13 in the Crash Course only does so much -> https://www.buildersociety.com/threads/day-13-paid-traffic.1302/

So I decided to open up a thread here for anyone and everyone to post their questions and thoughts about Adwords, Bing, Facebook or other PPC traffic sources. Granted, I haven't touched every source and some I haven't bothered with in years.. but lay out your questions here and I will see if I can help.

I want to help people with the PPC issues, but I don't want a lot of armchair quarterbacks coming in and trying to throw their weight around with "well maybe...", "instead you should...", or "here is my thoughts.." and giving out advice that is only somewhat helpful and meaningful.

Unless you have at least 5+ years direct hands-on experience, $1M spend of your own money, or managing a budget of at least $1M a year of other's money for more than 2 years... please don't weigh in with advice that can actually confuse or hurt someone. Last thing I want is someone guiding someone down the wrong path.

I'll start out with something easy and focused..

What's your biggest issues with Adwords?
 
I oversee a small/medium sized adwords campaign that is largely 'brand term' targeted. The campaigns are always ROI positive, really just for exposure on the branded search pages.

I am looking advice on how to begin exploring out into generic terms to gain new sales while maintaining a positive ROI.
 
I oversee a small/medium sized adwords campaign that is largely 'brand term' targeted. The campaigns are always ROI positive, really just for exposure on the branded search pages.

I am looking advice on how to begin exploring out into generic terms to gain new sales while maintaining a positive ROI.

Got the subtle hint OP would prefer no wannabes chiming in but since I´m (barely) qualified according to the set standard, I feel more comfortable contributing instead of just asking. Nuke away if you feel that way, way more interested in your contribution than sharing mine..

@animalstyle

- Product terms
You´re selling sneakers? Someone looking for "Nike SupaDupa XWA 43 in red" is looking to buy. If you can´t convert there´s something wrong with the account architecture (or copy/extensions or whatever in platform end) keeping CPC too high* and/or landing page (QS), UX (do CRO) or product/service (payment options, delivery methods, targeting etc?).

Category terms (sneakers, white sneakers etc) are more difficult or at least have way more variance and/or generally lower margins IME.

*this is obviously not 100% accurate, one or multiple competitors may be buying market share, running sub-optimal campaigns driving the CPC up, or just be way better at business (eg. buying) than your client and have way higher margins allowing higher CPA. Or (in business) cross-sell/upsell/retention again can afford way higher CPA. Research your top 5 competitors at least, and not just what they do in Adwords but on-site etc.

- Competitor terms
In my experience competitor terms are way often super-misunderstood* ("poor coherence" "poor search intent match" "cant compete with the brand, too expensive") but these should (most often) be the third biggest money-makers behind brand and core/product terms if done properly.

Create dedicated landing pages to secure decent QS and to convert (say a comparision page between *your product and *targeted competetitor*) - you have a very shitty product or are poor at copywriting if you can´t find/twist USPs in which your product kicks a specific competitor.

EDIT: Someone is not up to speed on trademark protection, abuse heavily in copy. Like brutally go along the lines of:
*Competitor* *product/category*
*my product* is better than *competitor*
*my product* is better than *competitor* in *USP #1*. Click here


*often leads to limited competition which I like

- DSA
Needs a bit of a budget to burn but works well especially in eComm with xx,xxx products and/or xx markets.

The idea is to find products/categories/anomalies that convert (example a product/brand with very low competition online (and/)or a market with low competition. Once you get pulse create specific campaigns targeting the market/product/brand

- RLSA
For example target existing customers (that have converted) searching higher in the funnel/with lower purchase intent terms.

Feel like I should add like a 57 disclaimers and it´s always case-dependant etc but with a generic question, a broad answer? Surprised someone actually has someone managing a branded campaign but is doing pretty much nothing else. Guess we all have our bubbles?

@eliquid

QUESTION 1:
Local market, competing for the top exact match kw in a top3 vertical/niche (spending close to $100k on this exact match kw alone per month. In relative terms this is THE kw in highly SEM driven industry so if the spend sounds low, it´s likely the market size. In most cases we are not comparable to big markets but this kw/niche I is actually competitive landscape; compared to my experience in global market/industries)

This is a oneword exact match, I´m obviosly doing the long tail (and doing relatively better) but the one-word makes like 80% of the one-word+long-tail combined volume, so can´t just focus in long-tail. Winning this kw is absolutely crucial.

QS 10/10
CTR varies between 10-15% and I don´t think a significant jump can be made here (what would you label a significant jump in this range? If most my variants are 13-14, would it make a differnece if I make it to 17%? I kinda think it wouldnt, especially as IME there´s a point where the increase is made of click-happy users rather than converting users as in shitty traffic?)

Limited control over landing page, but what can be done, is done (ie. can´t count on conv rate increase)

I´m doing the usual:
ad/usp testing
dayparting
geo bid adjustments
demographic adjustments
trying to work the extensions
lots of thing in bidding (there seems to be unusually stron cut-off, at $XX,X I get volume but it´s borderline to expensive; at XX,X-0.01 it´s dead traffic wise and XX,X+anything it just kills the CPA
etc

Problem: CPC is still too high if I want to maximize (or get any) volume.

Guess it´s a broad questions but where would u be looking? Or do you think maybe one of the above is the key I should be focusing on?

QUESTION 2:
In my experience, extensions (especially sitelinks) are a way underrated/neglected factor in Adwords (seen some just huge performance differences that can only be accounted to specific extensions)

- Your opinion on the above (as in significance of extensions)?
- I´m struggling with measuring testing* the extensions, any tips or suggestions etc?

*It doesnt help there´s usually a boatload of bid changes, ad testing etc going on, so how to pinpoint the impact of extensions? Experiments and keep everything else 1:1 for the test period?
 
Last edited:
I oversee a small/medium sized adwords campaign that is largely 'brand term' targeted. The campaigns are always ROI positive, really just for exposure on the branded search pages.

I am looking advice on how to begin exploring out into generic terms to gain new sales while maintaining a positive ROI.

You're gonna have to start out slow and small.

I've seen too many accounts killed over "too much, too soon". Meaning too many changes too fast to where you and the account can't keep up and you lose the ROI.

I'm not sure if you are in shoes like AZSWE said, but i'll use another example:

1. You're bidding on brands right now, but wanting to go more generic. If I was a lead aggregator for Higher Education, I could bid on the brand names of degrees like "Princeton Degrees" or "University of Kentucky Masters In Business Administration Degree" similar to your brand strategy now maybe.

2. Going broad and trying to hit just "MBA Degree", I am going to get slaughtered. You don't want to go this big yet.

3. I could start out a lot smaller and hit up "Womens college MBA degree completion". Not only am I seeking out those that want to go to a womens college, but those only looking for the degree completion option. I could further this by going into "womens catholic college MBA degree completion".

4. When you go to far niche, you cut off available traffic. Hardly anyone is looking for "womens catholic college MBA degree completion", but I use that as an example to help you. You need to find the balance of small profitable and large traffic borderline.

If you are doing shoes, then the brand example would be "Reebok Pumps Size 11 Mens". A more generic term might be "nonlace white tennis shoes for kids" or "best basketball training shoes for strength".



You're going to want to start with 1 or 2 campaigns like this. Highly targeted and focused and just starting out with these until you can make them profitable.

The best to start up a campaign like this would be to make the campaign and then make 3 adgroups. Each adgroup is a match type ( modified broad, phrase, exact ). I'd put your term in the phrase match group and leave the rest alone at this moment and let's pretend you used "training shoes for strength" as the term.

It might take some time, but you will see over weeks that terms come in like "basketball training shoes for strength" and "best basketball training shoes for strength" and even maybe "golf training shoes for strength". You start seeing these come in and you can start negative matching and building out what's coming in from your search query report.

Once you start getting conversions, you put those terms that converted into your Exact Match adgroup and you negatives match them in your Phrase so they only show up in the Exact now. These are your working exact matches now. The Phrase match campaign will no longer show them ( you need to negative match them exactly in the phrase match ) and still "dig for gold" and work for you.

At some point, you do the same with Modified Broad match. A lot of people actually start there and then move their working terms to Phrase and Exact, but I started you at Phrase because it will prob. save you money upfront since you are already ROI positive. At some point, you want to turn this digging of the gold taks to the modified broad adgroup. Depending on your niche and it's size, it might make sense to just start at modified broad, or just phrase. Where you start is not important, the task itself and doing the process is.

You probably already have a good negative list from the Brand stuff that can be recycled for this task. Stuff you know you don't have or carry like maybe golf shoes or an associates degree ( from my Higher Ed example ). Reuse that in your new campaigns. I actually do negatives as Phrase match and pretty much do that only so the wrong stuff never comes back in.

If you are finding you are still too large in your terms and need to test more, you can always shrink your geolocations. Maybe you run the whole US in the brands, but you find most of your sales come from California. Well, if your new campaign is spending a lot and you dont want to niche down further, just limit the geolocation to California or even just to LA and run it during the hours that are the best hours for the Brand campaign. Chances are, your buyers are in the geolocation and hour range anyways and you can test your new stuff in those restrictions and not waste too much ROI on a new campaign while you test and clean up the new campaign.

Also, does your site have a search bar or search function? You can prob get a TON of data from that if you have a database that recorded the type ins. From that you could break out new campaigns and see how far niche you need to go.
 
Wow.. Perfect timing. How would you go about doing a lead generation campaign for a medical practice? In my example, my client is an endocrine breast surgeon. I also have a brain surgeon client. I'm building their sites for now but I would like to offer lead generation as a value added service.

Do I need to have a medical degree to advertise medical terms on adwords?
 
Oh boy, here I go asking again.

I have a client for whom I run Adwords campaigns. His market is pretty small. The market here in general is microscopic compared to the USA market.

Right now the average QS is around 8 for all the keywords. The bid is around 30% higher than Google recomends for the first page. But the ads are still on average in 4.6 position.

What would be the best way to bring the position up?

The keywords right now are Broad modified.

Should I eliminate the ones that are getting "Low search volume" from ad groups? Could that be bringin the position down?
 
Oh boy, here I go asking again.

I have a client for whom I run Adwords campaigns. His market is pretty small. The market here in general is microscopic compared to the USA market.

Right now the average QS is around 8 for all the keywords. The bid is around 30% higher than Google recomends for the first page. But the ads are still on average in 4.6 position.

What would be the best way to bring the position up?

The keywords right now are Broad modified.

Should I eliminate the ones that are getting "Low search volume" from ad groups? Could that be bringin the position down?

You said QS of 8 is the average for the keywords, but that could be some are 10 and some are 5.

I don't look at the bid that Google gives me, I simply move the keywords up.

There are several reasons why your position might be 4.6, but more than likely you have either bid to low or your QS for some of these is low.

What I would do is increase the bid to get all of them to 2.5 at least on average. You are getting bad data at avg position at 4.6 no matter what regardless.

Increase the bid to 2.5, wait a few days to get clicks in on that position, and then if the QS is low on those, you need to change the ad copy and LP copy for those low QS keywords. You may also need to redo the adgroups/campaigns those low QS keywords are in to make them more tighter

Wow.. Perfect timing. How would you go about doing a lead generation campaign for a medical practice? In my example, my client is an endocrine breast surgeon. I also have a brain surgeon client. I'm building their sites for now but I would like to offer lead generation as a value added service.

Do I need to have a medical degree to advertise medical terms on adwords?

Don't need a medical degree.

To be frank, I'd ask the client what terms people use when they come in and have questions about their service. Sometimes you just wont know and you wont look like a fool for asking either.

I'd simply ask them what other terms are highly relevant to "endocrine breast surgeon". Then Id take those terms and see who ranks well for all those terms now and use a tool like SpyFu or SEMRush to get other terms those sites rank for organically and in PPC. Now you will have a bigger list for you client.

After that, you start setting up PPC ads based on those keywords ( if relevant ).

QUESTION 1:
Local market, competing for the top exact match kw in a top3 vertical/niche (spending close to $100k on this exact match kw alone per month. In relative terms this is THE kw in highly SEM driven industry so if the spend sounds low, it´s likely the market size. In most cases we are not comparable to big markets but this kw/niche I is actually competitive landscape; compared to my experience in global market/industries)

This is a oneword exact match, I´m obviosly doing the long tail (and doing relatively better) but the one-word makes like 80% of the one-word+long-tail combined volume, so can´t just focus in long-tail. Winning this kw is absolutely crucial.

QS 10/10
CTR varies between 10-15% and I don´t think a significant jump can be made here (what would you label a significant jump in this range? If most my variants are 13-14, would it make a differnece if I make it to 17%? I kinda think it wouldnt, especially as IME there´s a point where the increase is made of click-happy users rather than converting users as in shitty traffic?)

Limited control over landing page, but what can be done, is done (ie. can´t count on conv rate increase)

I´m doing the usual:
ad/usp testing
dayparting
geo bid adjustments
demographic adjustments
trying to work the extensions
lots of thing in bidding (there seems to be unusually stron cut-off, at $XX,X I get volume but it´s borderline to expensive; at XX,X-0.01 it´s dead traffic wise and XX,X+anything it just kills the CPA
etc

Problem: CPC is still too high if I want to maximize (or get any) volume.

Guess it´s a broad questions but where would u be looking? Or do you think maybe one of the above is the key I should be focusing on?

I had to think about this one for a while. My best answer is, if you are already doing all this and have a QS 10 and cant change the page much, then you have done all you can.

You can do more, but with what is left.. you won't gain more. The only thing I can think of is Call Only ads, but you might not be able to use those in your setup, so I would say at some point you have hit the point of max return for input you have put in.

At this point you can't really bid more unless you get click happy traffic that wont make you money, and you can't bid less or else you get less volume.

I'd be worried if I were you though in the above case... what happens when a new competitor comes in and outbids you for a couple months.. do you just die out because if they outbid you, you drop in position and lose volume.. but if you bid up you lose ROI?

Based on this, I think this is something you wont get much out of trying more of X, whatever X will be. The only other thing I could think of is Call Only, but that's dependent on how you're setup.

QUESTION 2:
In my experience, extensions (especially sitelinks) are a way underrated/neglected factor in Adwords (seen some just huge performance differences that can only be accounted to specific extensions)

- Your opinion on the above (as in significance of extensions)?
- I´m struggling with measuring testing* the extensions, any tips or suggestions etc?

*It doesnt help there´s usually a boatload of bid changes, ad testing etc going on, so how to pinpoint the impact of extensions? Experiments and keep everything else 1:1 for the test period?

I like extensions. Mostly to push competitors down the page so that my ad takes up more real estate.

I also like them to try to display more call to actions to the potential client.

Extensions are very significant and help a ton in different ways like I listed above, so I would for sure always use them if I were you.

As far as testing for impact of extensions versus without, I wouldn't have much advice there. It's like saying.. how much impact does me having clean fingernails impact my chances of landing a hot date on Saturday night at the bar? We all know it does, the ladies are looking for clean men so do we really need to test the impact? Just do it.

The reason I say this is because the extensions push your competitors down the page and your ad is huge and the customers see it and focus on it when you have them. That can't be denied. More eyeballs ALWAYS = good.

2nd, you get to put in more calls to action. You get to add more value. You get to standout. That will also ALWAYS = good.

So if you want to know impact only, id say don't bother.

If you want to know impact of specific extensions, like how your call outs do.. that is different. Just make 6 and run them for a month and then look at how performance wise at the end of the month and cut the losers. Do this for the others.

In my mind though, there is no need to see if adding extensions of any kind will impact me. I know they will just by human nature and how people interact with serps.
 
QUESTION 1:
Local market, competing for the top exact match kw in a top3 vertical/niche (spending close to $100k on this exact match kw alone per month. In relative terms this is THE kw in highly SEM driven industry so if the spend sounds low, it´s likely the market size. In most cases we are not comparable to big markets but this kw/niche I is actually competitive landscape; compared to my experience in global market/industries)

This is a oneword exact match, I´m obviosly doing the long tail (and doing relatively better) but the one-word makes like 80% of the one-word+long-tail combined volume, so can´t just focus in long-tail. Winning this kw is absolutely crucial.

QS 10/10
CTR varies between 10-15% and I don´t think a significant jump can be made here (what would you label a significant jump in this range? If most my variants are 13-14, would it make a differnece if I make it to 17%? I kinda think it wouldnt, especially as IME there´s a point where the increase is made of click-happy users rather than converting users as in shitty traffic?)

Limited control over landing page, but what can be done, is done (ie. can´t count on conv rate increase)

I´m doing the usual:
ad/usp testing
dayparting
geo bid adjustments
demographic adjustments
trying to work the extensions
lots of thing in bidding (there seems to be unusually stron cut-off, at $XX,X I get volume but it´s borderline to expensive; at XX,X-0.01 it´s dead traffic wise and XX,X+anything it just kills the CPA
etc

Problem: CPC is still too high if I want to maximize (or get any) volume.

Guess it´s a broad questions but where would u be looking? Or do you think maybe one of the above is the key I should be focusing on?

Have you tested Bing?
 
Hi eliquid,

What do you think of Semrush and Ahrefs, do you think these guys are a game changer when it comes to keywords and Adwords? Are these two the leaders when it comes to that?

Thank you
 
How do you approach a client who is looking to use PPC only for awareness? They're currently happy just knowing that people visit the site. Do you aim for cheap CPCs and traffic volume? Or do you try to get them to shift their focus on converting that traffic into email signups, lead captures, etc.
 
Hi eliquid,

What do you think of Semrush and Ahrefs, do you think these guys are a game changer when it comes to keywords and Adwords? Are these two the leaders when it comes to that?

Thank you

Game changers? No.

Maybe back in the day I would consider that, like pre-2008. Now they are more like copycats looking to rip features off other tools to keep market share.

If you want some basic info, then yes look into those tools as a starting point. However, there is no reason to use them after that point.

However, each campaign and site is different. Each flow and conversion path is different. If you stick to this you will see where tools that spit out data about "ads still running" is flawed to be used as data for your campaign.

How do you know if the client you are watching in SEMRush isn't the one @ryandiscord is talking about in this thread?

How do you know they even have a good conversion path and if they don't, the data you are looking at is flawed?

All questions you should ask yourself.

If you used a keyword tool though ( like Google's or a free one, or one from SERPWoo ) you would probably get the same data and a better beginning set to test real metrics with that are going to be based off your site and conversion path.

I'd still look at the data from SEMRush, Ahrefs, SpyFu, etc because that's how real marketers roll. We look at everything available to us and make up our own minds. Just don't get dead set on ONLY using those. Find out why it might be flawed and cover your ass for that just in case.

How do you approach a client who is looking to use PPC only for awareness? They're currently happy just knowing that people visit the site. Do you aim for cheap CPCs and traffic volume? Or do you try to get them to shift their focus on converting that traffic into email signups, lead captures, etc.

At some point it is extremely hard to change someone's mind. Knowing that, give them what they want right now, but build a few other campaigns in the background that are paused and ready to go based on conversions and signups.

Sometime in the future they will realize they want to focus more on doing stuff with this traffic and you will have the campaigns ready to go and possibly some data behind them too to show at the right time.

I've had a couple clients like this and I generally build it out in the background and keep paused. I work them over the next 4-8 weeks on going down this road with little open loops on each convo that finally get them to bite and at least give me a testing budget for it.
 
I got a question for everyone because I want to see how you will react.

How do you know when it's time to cut a keyword? What do you look at?
 
Knowing that, give them what they want right now, but build a few other campaigns in the background that are paused and ready to go based on conversions and signups.

Thanks! Would you build that out in the account or Adwords Editor?

I got a question for everyone because I want to see how you will react.

How do you know when it's time to cut a keyword? What do you look at?

It depends on the match type. I've used broad match for research, when I collect enough search term data, I'll add new phrase match keywords and kill the broad. If the broad term converts, I'll keep running it and extracting search terms. Once that fades out because I've moved the converting search terms out, then I consider pausing it.

Exact match terms, I'll run as long as there is a positive ROI. If the keyword doesn't perform and I've tested enough ad text and landing page options then I'll kill it.

I look at my return on a campaign level so I can afford to spend on collecting data from broad and fine tuning exact match. I've actually found some hidden gems by unpausing keywords that were too hastily paused from previous campaign managers.
 
I got a question for everyone because I want to see how you will react.

How do you know when it's time to cut a keyword? What do you look at?

What I have here is a super small search market, so I don't have the time to collect statistically significant data! (As retarded as it may sound)

At some point for Adwords I have to do an educated guess.

I'm mostly looking at ROI. I get ~100 clicks on the keyword, if it is converting then I work with it, working with ad text. If it is not converting at all 95% cases I will cut it.
 
Monthly shouldn't be an issue. Unless maybe your high traffic day is at the end of a month.
Keep in mind, you won’t be charged more than your monthly charging limit: the average number of days in a month (30.4) multiplied by your average daily budget.
Even if you do overspend:
Sometimes we deliver over your monthly budget. In those cases, we'll credit the overdelivery cost back. If you'd like to see whether we've given you overdelivery credits
I have had it happen but it's rare. Maybe it will happen more often now.
 
@ryandiscord

Yeah the bigger issue will be if you run out earlier in the month now, like say the first 2 or 3 weeks and left with 0 budget in the month
 
Interesting point. The budget is still daily though so wouldn't that just become overspend?
On days with lots of high quality traffic, your costs could be up to 2 times your daily budget. This spending is balanced by days when your spend is below your daily budget.
Maybe they account for a campaign historically being under its daily budget to determine how much they could go over in a day? Seems like a bit of guesswork that they're willing to credit you for if they get it wrong.
 
Back