As an Unbounce blog reader, you already know conversion rate optimization is a great way to drive more results from the money you’re already spending to acquire traffic to your site. So today I’d like to focus on another way to make more money from your marketing efforts; namely, ad optimization.
After a decade of building AdWords at Google, I cofounded Optmyzr because I found it was too time consuming to run best practice optimizations. And — considering that last year over $35 billion was spent advertising on Google (more than twice what was spent on Facebook) — there is tremendous value in making AdWords perform even a little bit better.
While there are many ways to improve your ad performance, one of the most pertinent is to improve your Quality Score (QS), particularly the subcomponent, landing page quality (LPQ).
But before I get into how to improve your Quality Score via landing pages, here’s some behind-the-scenes context based on my time at Google.
What is Google Quality Score?
One of the teams I worked on for seven years while at Google was the Quality Score team so I’ve written extensively on the topic. For a primer, here are some of my favorites:
- Quality Score Explained by a Former Googler
- Quality Score: All You Need to Know
- Calculating Quality Scores (Video)
Ultimately, Quality Score is Google’s way of using collective wisdom of many searchers to measure the relevance of a keyword. In short, it’s a measure of how good of a job you’re doing providing people with search queries with strong answers.
This is important because if you are the best answer to someone’s query with your ad and resulting landing page, Google gives you your clicks for less money because you’re helping provide a better user experience, which makes Google more money.
Quality Score impacts ad rank and costs
It’s fairly well known that Quality Score is one of the factors that determines Ad Rank. The other factors are the bid and the impact of ad extensions.
When you improve any of these three components, your Ad Rank increases, which leads to one of two possible outcomes:
- You win a better ad position (but pay the same or more for improved placement on the page), or
- The position of your ad remains the same, but you get a discount and pay less for any clicks you get.
In short: You can get a discount on AdWords by making your ads (and corresponding landing pages) more relevant!
This is because the CPC you pay is only part of what determines how Google makes more money.
Ad Rank now vs. then
To deepen your understanding of AdWords, I find it useful to take a quick trip down memory lane about the ad auction. Ad Rank used to be calculated like this:
Notice that Ad Rank is actually Cost Per Mille (CPM or cost per thousand impressions) in this equation! Over time, Google started to use a predicted rather than historical CTR, so we changed how we explained the formula and introduced the concept of Quality Score so we could stop talking about pCTR (predicted CTR).
Then we refined the algorithm to deal with some weird edge cases and rather than just multiplying the factors, the formula became more advanced. We communicated it as follows:
With this historical context, we see that the ad auction is basically a CPM auction where the way the CPM is calculated has evolved over time.
So even though advertisers place CPC bids, Google is awarding the best ad positions to advertisers who deliver the best CPM for Google (aka: those who provide more relevant ads that get the most clicks).
How Quality Score can reduce CPCs
As an example to show how this discounting of CPCs really works, let’s say there are two advertisers, Julie and Robert and they both bid a max CPC of $1. However they have different Quality Scores, a 5 and a 10.
|Julie||$1||5||5||$1 (or the auction minimum)|
As you can see, Robert (who’s winning the auction) actually pays a lower CPC than Julie, because his Quality Score is better.
Effective CPC for Robert = 5 / 10 + $0.01 = $0.51
This classic example oversimplifies the calculation of the CPC, however. For the example to be correct, we’d have to say that Robert’s CTR is twice that of Julie, and not just that his Quality Score number is twice that of Julie’s.
That’s because the visible QS number (between 1 and 10) is based on a non-linear assignment of a visible number to an underlying prediction of relevance.
To calculate the rank and CPC, Google uses a real-time prediction of CTR based on as many as hundreds of factors.
Real time Quality Score vs visible Quality Score
The Quality Score number between 1 and 10 that you see next to each keyword is simply an indicator and can be used to help prioritize what to optimize. After optimizations, it can be used to determine if improving relevance was achieved.
However, this number is NOT what is used in the Ad Rank formula.
Some indicators have more precision than others and there are also indicators that are linear and some that are not (the visible QS indicator is not necessarily linear).
The speedometer in a car and the signal strength bars on your cell phone are both linear, but the former is more precise than the latter. As you drive faster, your speedometer goes up precisely to tell you that you’ve increased speed. The bars on your cell phone, however, may take a while to go from one bar to two bars even though the signal strength has been gradually increasing for a while. Visible Quality Score is more like the phone’s signal strength indicator, except that there are 10 levels.
Articles claiming you can reduce your cost by 50% by doubling your QS number are oversimplifying, but they still make a valid point: that better Quality Score will lead to lower costs (assuming no jump in position).
The only way to reduce your CPC by half is by doubling your predicted CTR.
What factors go into Quality Score?
As an AdWords marketer, you want to get yourself a better Quality Score and a lower cost per click. There are several factors that go into Quality Score so let’s take a look at what those are:
- Ad relevance
- Expected CTR
- Landing page experience
Ad relevance is an indicator of how well your ad text matches your keywords.
It’s usually a good idea to include the core concept of the keyword in your ad text, and to also include some compelling unique value propositions that will make your ad stand out from competitors. Over the years I’ve personally noticed that even minor changes in word choice can have drastic impacts on how well the user understands the ad.
For example, when eBay changed their ads to say, “buy it on eBay” instead of, “find it on eBay.” Because the word “buy” implies ecommerce, this got a far better CTR than the word ‘find’ which didn’t directly suggest one could buy the desired item.
Expected CTR is an indicator of how likely your ad is to be clicked.
At the most basic level, this requires choosing good, relevant keywords, and grouping them in logical ad groups so that you can write compelling ads that get users to click.
Landing page relevance
Landing page relevance indicates how well your landing page meets the needs of users, and there are many ways you can go about improving this.
As an example, Joe Khoei from PPC agency SalesX (where I serve on the board) says that using Dynamic Text Replacement on Unbounce landing pages for the Children’s Learning Adventure helped his client increase conversion rates (calls and form fills) from 1.4% to 3.3% over 8 months.
Generally, using personalization features like DTR will correlate to better Landing Page Quality (LPQ) because users are getting what they want and that is what Google wants too: happy users who continue to engage with ads.
There’s an interesting tug of war between motives of landing page optimization; an optimization for conversion rate could hurt Quality Score, and Quality Score optimization could decrease conversion rates.
The trick is to find that right balance, and ideally aim to optimize where both QS and CR improve. Fortunately, they’re not mutually exclusive outcomes.
How to better track your Quality Score improvements
Once you make optimizations to your ads or landing pages with the goal of improving Quality Score, you’ll need to track if the changes are working. In this stage it’s critical you don’t lose sight of the bigger picture, which is to grow your business.
No executive ever said their goal for the company was to improve their QS next quarter. So remember, it’s a useful gauge to see if you could lower costs, but it’s not a business KPI.
Up until fairly recently, the easiest way to track changes in QS required using an AdWords Script, or signing up for a tool like Optmyzr, but now the data is also available directly in AdWords.
Additionally, tools like Optmyzr still make it easier to get historical QS data.
To get at the daily variations in AdWords, you have to do a few things:
- Use the old interface. The new one doesn’t include subcomponents yet.
- Look at the data with the segmentation for “day” turned on. Simply looking at the QS between two compared date ranges doesn’t get you the daily data because both fields will show the ending value.
To see how your QS has evolved based on changes to your landing pages, turn on the landing page component in the AdWords interface and then download the data with a daily segment.
You’ll end up with columns like this:
After exporting this data, you can use a simple spreadsheet formula that compares the current value to the historical one, e.g. =IF(A2=B2,1,0), so that the field will contain a 1 if the value has remained the same.
From there, filter these out and you’ll see only instances where the LPQ has changed. This will help you see where optimizations to your landing pages are positively affecting LPQ as a subfactor in Quality Score.
The purpose and importance of landing page quality
Landing page was the last component to be added to Quality Score and I was still on the team when we made this change. We had come to realize that it was too easy for advertisers to game the system by writing must-click ads, but then lead the user to a not so spectacular landing page, and in some cases even to a scammy site.
We had to start looking at what happened after the click, so we used both manual processes with the policy team and automated ones through the QS indicator to find sites that weren’t delivering a great experience.
What matters for landing page quality?
Here are some of the things Google cares about for landing pages:
- The page needs to deliver what the ad promises
- The user’s privacy and personal information must be protected
- The page should be transparent about its purpose
- The landing page should let users freely navigate the web
- The page should load quickly
Deliver what the ad promises
The first part should be the easiest to abide by. After all, if you want to drive conversions, you should be doing CRO and you should take users to landing pages that are relevant to what you offer in your ads.
Protect the user’s privacy
Privacy gets a little trickier.
What Google wants here is that you don’t share user’s information with third parties without their permission. So if you’re collecting leads and reselling these, you could be running into LPQ issues.
Be transparent about the page’s purpose
This also rolls into transparency.
If you’re a middleman, you need to be upfront about that. Affiliates who create thin landing pages and doorway pages are likely to run into LPQ issues because they usually add very little value and yet force the user to go through an extra step to get to where they wanted to go in the first place.
Allow users to freely navigate
Providing options is another tricky one for advertisers who deploy landing pages for their ads rather than taking the user to the most relevant page on their site. At issue here is that the user can do only do two things on most of these pages: instantly convert, or go away. Even a user who is interested in what you offer may not reach the comfort level needed to submit their info right away and if they can’t find more info by browsing the main site, their only real option is to go away and that is seen as a bad thing by Google.
While it can be a best practice to have a classic landing page perform only one goal and not include navigation or distractions, what I’m advocating for here is that you consider the allocation of your paid spend to a combo of the right, most relevant pages (whichever they may be).
That is – consider for each keyword or ad group which page on your site (or which landing page) may be best to serve up for a given situation. You may be pointing paid traffic to a landing page that is too high commitment for someone at the top of the funnel and this could hurt your Quality Score if someone’s only option is to convert or bounce. It’s a matter of their perceived readiness.
To ensure you’re not penalized this way, get strategic and point your paid traffic to highly relevant pages, either on your site, or build more valuable, relevant landing pages for each stage of the buying process to access via your ads. By having especially relevant landing pages that present the right offer at the right time, you should be able to avoid the issue of options as visitors will be served up the right option for them on their journey.
Again, if you have a great offer, and a decent site, the idea is that users should be able to get all their questions answered before being asked to turn over their details.
Landing page speed
And finally, the speed of your landing page is very important as a factor Google considers.
Just know Google is not actually that stringent and so long as your landing page is not an outlier in terms of slowness, you’ll be fine.
But you should still care tremendously about load times because as a 2017 study by Akamai found, a 100-millisecond delay in website load time can hurt conversion rates by 7 percent and a two-second delay in web page load time increase bounce rates by 103 percent.
Google’s Group Product Manager for AdWords, Jon Diorio, recently shared a stat from SOASTA that a 1 second delay in landing page load times can decrease retail conversion rates by 20%.
Pay close attention to signals about whether landing page visitors are satisfied
When I was at Google and I gave presentations at industry events about Quality Score, someone would always ask how we measured landing page quality. While I couldn’t answer that question directly then, and still can’t today, I recommend that you pay close attention to signals like bounce rates and time on site.
Google Analytics is a great way to track these signals which are fundamentally a measure of how satisfied users are with your landing pages.
If a user sees an ugly page, a page that takes too long to load, or one that seems off topic for what the ad promised, they will use the back button and try their luck at the next site.
Whether you’re using Unbounce or not, you are hopefully already paying close attention to these things. After all, this is part of CRO. If a user doesn’t stay long enough to consider your offer, they surely won’t have time to convert and your cost for the click to get them to your page will have been wasted.
See the original post:
There is no doubt about it, split testing paid traffic can be confusing. Even if you know what numbers to review, it’s hard to know what they mean or how they can guide your strategy.
If that’s you, I’m going to clear that up for you in this article. I’ll show you how to use this data to get better results from your paid traffic without paying too much.
We’re going to cover 3 important metrics that you need to split test. All 3 are critical to getting paid traffic right:
- CTR (clickthrough rate)
- CPC (cost per click)
- EPC (earnings per click)
Once you know how these metrics affect your campaigns and how to change them for the better, it won’t be long before you’re getting great results.
3 Metrics You Must Know to Optimize Your Paid Traffic
Click To Tweet
CTR: The Basics of CTR
CTR stands for clickthrough rate. It is often used in the world of paid traffic and can determine a number of things.
1. CTR is commonly used to determine how relevant your ad is in relation to the people you have targeted.
If your ad has a great market-to-message match, the CTR should generally take care of itself.
If people come across your ad whilst on Facebook or entering a Google search, they will probably click on it if it is relevant to them. However, if your ad is not relevant, no one will want to click on it. That is why CTR can be a useful metric when measuring the effectiveness of an ad.
2. CTR can also determine how much your ad is going to cost.
For some ad platforms like Facebook, a low CTR can lead to a higher cost per click (CPC). This is often because platforms like Facebook want to keep things as relevant as possible and so will penalize those who have chosen to display irrelevant ads. Google tends to this as part of their ‘Quality Score.’
How CTR is Calculated
With paid traffic, CTR is the result of the following equation:
people who saw your ad/people who clicked your ad
This number is then displayed as a percentage.
So imagine you were using Facebook advertising and 100,000 people saw your ad. Then imagine that 1,000 of those people decided to click your ad. That means that you had a CTR of 1%. Now, that might sound low, but you’d be surprised to find out it is actually a good number to aim for when it comes to Facebook Advertising.
In any case, you need to remember something important. Each form of paid traffic is going to have its own ideal CTR. For search campaigns on Google Adwords, the CTR generally tends to be much higher than 1%. You should therefore avoid using one ideal CTR number across all platforms.
How Does CTR Apply to Split Testing?
As mentioned earlier, you want an ad with a high CTR. So if you have a number of similar ads, you might want to improve the winners by improving their CTR.
A high CTR has the potential to reduce your cost per lead. That’s because you are now receiving more relevant clicks to your ad, and hence, more of these people will enter your ‘lead flow.’ This of course assumes that they find your landing page relevant. If all is well, this should eventually lead to more revenue for you.
How Can You Improve Your CTR With Split Testing?
To improve your CTR, you’ll want to create several versions of the same ad. You’ll then want to implement some differences amongst the ads being displayed.
Experiment With Targeting
If you are using Facebook or Twitter, you might want to improve the targeting.
Is your audience too broad or too narrow? If there are too many people seeing your ad, naturally you are going to have a low CTR. That is because you’re trying to reach too many people. If your ad is too narrow in its reach, it might not be reaching enough people to produce a good CTR.
On Google Adwords, you may want to make your keywords more specific. Experiment with varying specificity levels when using targeting/keywords to improve your CTR.
Vary the Copy
If you have a low CTR, you may also want to improve the copy in your ad. To keep things simple, I always try to follow the concept of entering the conversation that is going on the mind of my prospects. When using Facebook Ads, I have found that:
- the headline is good for calling out your ideal client (an example might be: ‘Are You Looking to Get Married?’)
- The image should catch their attention
- the text should let them know what to do next
Change the Ad Format
If you find that your ads are producing a low CTR, you can change the ad format. This generally involves creating a totally new ad that works in a different manner.
If you are using Facebook Ads, you might find that newsfeed ads are able to provide you with higher CTR levels. If you are using Google Adwords, you might want to experiment with the display network.
Keep in mind, however, that targeting is still a big focus. Your CTR will not rise in a productive manner if you’re just changing the style of ad and ignoring the importance of targeting.
Go After Different Groups of People
This is slightly different from the point above. Here we want to go after completely different groups of people. This is often a good idea when using Facebook Ads.
If you have a winning ad, you could try using that ad with different targeting preferences. So for instance, if you were selling a weight loss product, you might decide to target people who were engaged and about to get married, instead of people who like fitness-related pages.
As alluded to earlier, you could also work on your copy to max out results. You could take things further by mentioning in your ad the fact that these people might want to lose weight in time for their wedding. Your copy could be something along the lines of, ‘Want to Look Good in that Wedding Dress?’ This of course is just an example, but hopefully you get what I mean.
Try Different Devices
You could also try targeting mobile devices. Some people find that their ads do much better on mobile devices. This all depends on the niche you’re going after. It also depends on what your giveaway is going to be, assuming you have one. If your giveaway is a video, it might be hard to view on a mobile device.
Replace Your Ads
With ad platforms like Twitter and Facebook you will need to replace your ads on a regular basis. That is because your CTR will fall even if you had a great ad with great targeting. This tends to happen because the same pool of people are seeing are your ad and so will become satiated with it.
After that, something known as banner blindness will begin to take hold. On Facebook, there is something known as ‘frequency.’ This lets you know how many times people have seen your ad. As this number rises, you’ll tend to find your CTR fall.
Facebooks ad tip: as frequency rises, CTR falls. Read here for the fix.
Click To Tweet
Don’t Be Careless
You should not try to raise your CTR for the sake of it. Just because people click on your ad does not mean that they are going to take an action on your landing page.
If you have a crazy image on your Facebook Ad just to get clicks, you may be successful in getting a lot of clicks. Yet there is a big chance that a lot of these people will not find your landing page relevant and so will just leave.
This is a phenomenon known as curiosity clicks and is to be avoided. You can minimize this by pre-qualifying people in your ad copy.
The Golden Rule?
I have found that in order to improve the CTR of an ad, the golden rule is this:
As long as you are targeting the right people, curiosity clicks should be kept to a minimum and CTR should rise. On Google Adwords your copy should be used to screen out time wasters.
The Importance of CPC
CPC refers to ‘cost per click.’ It lets you know how much you are paying every time someone is clicking your ad. When using paid traffic, there is the option to use CPC or CPM (cost per impression).
CPC tends to be the best way to go about things because it can provide you with some solid numbers. These numbers relate to how many people are clicking your ad and can eventually let you know how much each lead is costing you. With CPM, you will be charged in relation to how many times your ad is shown.
How CPC Works
Whether you are using Facebook Ads, Google Adwords or Twitter Ads, CPC can be a very dynamic number. Most ad platforms use a bidding mechanism. If there are a lot of people trying to target the same group of people, the average CPC will rise.
If you bid too low, your ad might not be displayed. If you bid too high, your ad will get a lot of exposure, yet you will also end up spending a lot of money. In fact, with a bid that is too high, you will generally tend to spend way more than you need to.
A CPC figure can change a lot, as your competitors are always starting and stopping their own ads. This means that they may or may not be bidding for ad space that you’d rather have, therefore causing price fluctuations.
Split Testing to Find a Good CPC
I have often found that if you try to be too cheap with paid traffic methods, you won’t get the best results. Of course don’t be frivolous with your spend, but keep in mind that sometimes paying what the platform suggests can bring the best results.
Nevertheless, you can do some split testing on your campaigns to improve your CPC. Create two versions of the same ad and then experiment. Here are a few ideas.
Who Has Your Competitor Forgotten About?
For a start, you can use the same ad and target different groups of people. If you target people that your competitors have forgotten about, you might be able to make some quick wins. The low competition means bidding costs are low. However, keep in mind that in some cases, people are not targeting these people for a reason.
You can also experiment with how specific you are being with your targeting. This tends to vary depending on the niche. However, by making your ad targeting/keyword more narrow or broad, your CPC could fall whilst providing you with the equal results.
Remember to Look at the Bigger Picture
Keep in mind that, to obtain a great CPC, it helps to look at the bigger picture. Remember ad platforms like Facebook Ads and Google Adwords tend to raise your CPC if your CTR is low. If you go do things that make your ad irrelevant, it could cost you either way.
In order to keep things affordable, you’ll want to do all that you can in order to keep your CTR high. That will stop things from spiraling out of control.
If you want to experiment with bidding amounts, there are a number of strategies that work. Many people find that it can be a good idea to let the ad platform pick the best bid for them. This gives them a chance to outbid everyone else by a small amount and give their ad the most exposure.
You could also try using the lowest possible suggested bid and then slowly reducing your bid amount. Some people try to do this until their results are no good. It has the potential, however, to produce lower quality leads.
EPC: The Most Important Metric?
EPC means ‘earnings per click.’ It refers to how much money you make per person who clicks on your ad. You can calculate EPC with this formula:
the money you make/the number of clicks generated
To be honest, I think that this is the most important number of them all. If your EPC is higher than your CPC, you are making money.
Ad tip: If your EPC is higher than your CPC, you are making money. Here’s why
Click To Tweet
To increase your EPC, you need to make sure that everything else has been split tested so it performs better. You cannot really split test your EPC directly. If you wanted to, the best EPC split test would be to try another traffic source.
It is important I mention EPC because it tends to be the reason you are doing all of your split testing in the first place. After all, who cares if you have a high CTR if you’re not increasing your bottom line? Always split test and then refer to the change in results to see how they have affected your EPC.
If you have a great EPC, you can afford to spend more on your CPC and expand your efforts. It is easy to get bogged down in the details when split testing paid traffic. By focusing on EPC you shouldn’t have that problem.
A Small Caveat: You Must Spend
When you want to split test, it is important that you are working with a good amount of reliable data. If your sample size is too small, you might not be able to make decisions that are effective, as you are working with incomplete information.
It is therefore important for you to spend around $50–$100 before you make some major changes. You should try to do this over the course of a day. When you have done this, you’ll then know whether or not your results are repeatable.
Time for Better Results?
To get the best results from paid traffic, you will need to split test. In my opinion, the metrics mentioned here are the most important when it comes to split testing. The most important number of all is EPC.
The points mentioned here assume that everything else already been optimized to its highest level. Delving deep into one aspect of your funnel can only bring you big benefits if everything is else is working well.
All of this testing is often done in the name of making money. If your tests produce a higher income, you know you’re making some great wins. However, if they just produce some vanity metrics, it might be a better idea to focus on other parts of your funnel. Using the information here you should be able to test your way to success.
In any case, try some of the tips mentioned and see how they work out for you. If you have some tips that have worked for you, feel free to mention them in the comments below.
The post 3 Metrics You Must Know to Optimize Your Paid Traffic appeared first on The Daily Egg.
See the original article here: