PPC Basics: Part 8. Evaluating Data
Even the best-laid PPC plans can fail if you’re not measuring and evaluating data. In this article, we’ll talk about ways to sift through the multitude of available PPC data to get to the golden nuggets.
Articles in This Series
- Part 1. How Paid Search Fits into Your Marketing Mix
- Part 2. Keyword Research
- Part 3. Account and Campaign Settings
- Part 4. Keyword Match Types
- Part 5. Ad Copy Development and Testing
- Part 6. Bid Management
- Part 7. Quality Score
- Part 8. Evaluating Data
- Part 9. Dayparting
- Part 10. Geo-Targeting
Establish Key Metrics
First, you need to establish key metrics. Ideally, you’ll take this step before you launch your PPC campaign. What are you trying to accomplish with PPC? Are you trying to sell more products, generate leads, create awareness, or a combination of all these?
A surprising number of advertisers forget this step, and that makes data evaluation nearly impossible. If you don’t know what your goal is, how will you know if you’ve achieved it?
Take the time to decide what your key metrics are, and make sure you have a way to measure them. You’ll get a lot of that data right in the PPC engines. All major PPC platforms will provide the following information:
- Impressions. How often your ad is shown. Each appearance of your ad equals one impression.
- Impression share. The percentage of impressions you received divided by the estimated number of impressions you were eligible to receive.
- Clicks. The number of times a user interacts with your ad by clicking on it.
- Cost. Total cost, calculated by the number of clicks times the cost per click.
- Cost per click. How much you actually paid for each click on your ad.
- Click-through rate. Expressed as a percentage, this metric is the number of clicks divided by the number of impressions.
- Average ad position. A statistic that describes how your ad typically ranks against other ads. This rank determines in which order ads appear on the page. Position one is the first ad on the page.
If your goal is awareness, the data above is really all you’ll need. Awareness is all about getting your brand and/or message in front of as many people as possible, as often as possible. If you’re trying to build awareness, you’ll be most interested in impressions: the more the better.
Impression share will likely also be of interest to awareness advertisers. Impression share tells you how much of your target market you’re reaching, and how much you’re not. Awareness campaigns should seek to maximize their impression share for best reach.
You may also be interested in click-through rate (CTR), because a higher CTR indicates a higher level of involvement and engagement with your brand and message.
Awareness is good for some advertisers, but many will be more interested in driving sales or leads. In this case, you’ll need a couple of additional metrics:
- Conversions. When a person who clicked your ad completes a valuable action on your website, such as buying something or requesting more information.
- Conversion rate. The percentage of clicks that resulted in a conversion.
- Cost per conversion. Total conversions divided by total cost.
- ROAS. Return on ad spend. Also called ROI (return on investment), this metric measures how profitable your advertising is. ROAS is calculated thusly: (Revenue – Cost) / Cost = ROAS
In order to measure conversions, you’ll need to have conversion tracking enabled on your website. While precise methods of conversion tracking vary, it usually involves placing a snippet of web code on the page that indicates that the desired action has taken place. For example, in the case of an online purchase, the conversion code would be placed on the order confirmation or “Thank You” page.
There are a few different methods of tracking conversions:
- Web analytics are the measurement, collection, analysis and reporting of internet data for purposes of understanding and optimizing web usage. Put simply, web analytics is software that tells you how visitors interact with your website as a whole. Most analytics packages include conversion tracking tags.
- Channel conversion tracking. The major PPC engines offer a free conversion tracking script that can be used to track conversions within each channel. For example, if you’re using Google AdWords, you can install the AdWords conversion tracking script, which will then provide conversion tracking for AdWords traffic only. It’s important to remember that these free scripts do not track all site traffic — they only reflect conversions from their own PPC engine.
- Proprietary tracking systems. Many large companies and marketing agencies have developed their own conversion tracking systems. Usually these systems work the same way as web analytics, but the code is proprietary. The advantage of building your own code is that you can customize the system to meet your needs. However, this requires coding know-how, and usually an in-house development staff.
Any of the above methods will help you determine which campaigns, ad groups, and keywords are actually achieving your goals.
Now let’s talk about evaluating results. This can be done a few different ways.
Eyeballing simply means looking at the numbers and making decisions without further analysis. While I don’t recommend evaluating results solely by eyeballing, it can be a quick way to weed out low-performing ads or keywords.
So what’s a bad ad or keyword? A good rule of thumb is any ad or keyword with at least 100 clicks and 0 conversions. Simple math tells us that this ad or keyword has a conversion rate of less than 1%. For nearly every advertiser, that’s a poor conversion rate. You’ll want to either edit or pause these ads or keywords right away so you don’t continue to lose money on them.
Sorting and Filtering
All the major PPC engines allow advertisers to sort and filter by various metrics. For instance, you may want to sort by cost per conversion, like this:
By sorting this way, the campaigns, ad groups, or keywords with the worst cost per conversion will appear at the top so you can look at them in more detail.
One caveat here: This method will push data with 0 conversions to the bottom of the list, where it might be missed. It’s a good idea to set up a filter for this data so you can review it separately.
When it comes to evaluating high-volume campaigns, ad groups, or especially ad copy tests, it’s a good idea to use statistical analysis. A lot of people are frightened of statistics, recalling scary college courses or hours poring over a calculator.
There’s no need to be afraid, though. Analysis tools are prevalent these days, and many of them are free. My favorite tool for simple, one-metric analysis is SplitTester. If you have two ads you’re testing, and you want to know which ad has the higher click-through rate, you simply plug in the number of clicks and the CTR, and the tool will tell you which ad is winning, at what confidence level.
Another tool I like is the PPC-specific Ad Testing Tool by WebShare. This tool will tell you which ad has better overall results, taking into account both CTR and conversion rate. All you need to do is enter a few pieces of data for each ad variation.
Most good PPC managers will use a combination of eyeballing, sorting and filtering, and statistical analysis to evaluate their PPC data. With a little practice, you’ll quickly become a pro yourself.