A lot of the reason why I think analytics is scary and people don’t check their stats is because there is a lot of lingo. I thought I’d dive in with some explanations of typical analytics jargon, because sometimes it can feel downright confusing, even to me.
For the purposes of today’s discussion, I’m mostly referring to Google Analytics, but in fact these terms are used wildly across a variety of analytics platforms.
If you’re not sure what analytics means, that’s ok – let’s start right there. Web analytics are simply tools that you use to measure how much traffic your website recieves, where it comes from, and what those users are doing while they’re on your site. This data is important because it can help you know what percentage of your web users are actually buyers, and help you gauge the effectiveness of your various online marketing – you don’t want to pay for online advertising that doesn’t send paying customers, right?
Ok, now let’s dive right into the murk. Here are some terms worth paying attention to, and what they mean.
In alphabetical order…
Ouch, alphabetical order means we start out on a hard one. But A/B testing is simply testing two different scenarios to determine which one has better conversions. Only two versions are tested in an A/B test; when multiple variables are changed, this is a multivariate test (see below).
How You’d Use This: Let’s say you have just 1 page where people buy something. You’re curious if people click a green BUY button more often than a orange BUY button. Google Website Optimizer offers you the ability to do this, but it’s definitely for the intermediate/expert folks.
Bounce Rate is the percentage of people who came to your website and left without visiting another page. It’s a number many people obsess over without reason, because there are lots of times when someone is a bounce and it’s fine – say, they land on your blog post that sends them over to another site for a special deal. Ok, they only visited one page, but it’s fine.
How You’d Use This: You want to get this number as low as possible. Just based on how the web works, it will never be zero, so stop trying. You’re doing great if you are below 50%. Smart people look at the bounce rate of important, individual pages, not an entire website.
A conversion is where a website visitor completes a specific action. Conversions don’t always have to be a sale; a conversion could simply be a signup to a newsletter form, or to leave their phone number for a callback. The percentage conversion or conversion rate is the percentage of all people who saw the offer and took it (so 100 people saw it, 5 phone calls, that’s 5% conversion.)
How You’d Use This: You need to know what your conversions are (a sale, signup, etc.) Then you have to compare those to your overall visits to get percentages. In Google, this can be tracked under the “goals” functionality.
A conversion funnel is the defined path that visitors should take to reach a final conversion. For example, if you have 3 screens of data to be entered before someone buys your product/service, each page or screen is a step in the funnel. The reason they call it a funnel is because as you move through each step, you’ll lose a few people, so you want as few steps as possible.
How You’d Use This: In Google, this is also under the “goals” functionality. If you have multiple steps in the checkout process, I suggest you include all of them in setting up a goal.
A cookie is a tiny text file placed on your visitor’s computer while browsing a website. Cookies contain information to track user behavior and returning visitors.
How You’d Use This: You don’t – Google and other software products handle this for you.
Direct traffic is a visitor to your website that didn’t come from anywhere specific – it wasn’t a referral, they had your web link and they clicked it, say from their browser favorites. Of course, there are many ways a visitor can get to your site without a referral – for example, your business card is a great referral source to your website, but since we’re jumping from offline to online, that isn’t trackable.
How You’d Use This: High levels of direct traffic are good because they could be repeat customers, but ask yourself where they might be coming from?
An entry page is a page that is the first page a visitor sees when reaching your site. Entry pages aren’t always your homepage; let’s say you have an advertisement that sends someone to a particular sales page for a package. The sales page is the entry page for all those ad clicks.
How You’d Use This: This is useful by seeing what you most popular content is. When a new customer comes to that page, what kind of experience are they getting?
An exit page is the last page a visitor sees on your site. They’re on the site and they click that “x” button to close the browser, or they click on a link that takes them to another site.
How You’d Use This: People tend to overengineer this one – the fact is that people have to leave your website at some point, right? So just pay attention to pages that rank high on the exit list. The top page should be the “thanks for your order” page, right? But are lots of people looking at a sales page and then leaving? What could be wrong?
A goal is another word for a conversion. Google calls it a goal because it isn’t always a sale. See conversions, above.
A heat map is a tool that shows levels of activity on a web page in different colors,with reds and yellows showing the most activity and blues and violets the least. The result is a map which makes it easy to see what areas on the page are the most active.
How You’d Use This: This is great for the visual learners who want to know what’s going on with a particular page – heat maps are only good for pages with lots of traffic and “next step” options, like a homepage. Google doesn’t technically offer a heat map tool, but you get something very close with their “In Page Analytics” option. (I’ve often found this buggy, your mileage may vary.)
Every request to the server is recorded as a hit; this isn’t a very useful measurement as even search engine spiders are included, as are a simple request for an image.
How You’d Use This: You don’t. If you see anything about number of hits, ignore it.
A keyword referral is when someone typed in a phrase into a search engine, and clicked on a result that sent them to your website. Google sometimes blocks the information and you’ll see “(not provided)” here, and if you advertise with Google Adwords sometimes this will say “(not set)”.
How You’d Use This: If you’re doing any search engine optimization work, I sure hope the words you are working on show up here. If not, you got problems. (Also: there will always be weird things here. I can’t explain it.)
A multivariate test is different than an A/B test (see above) in that you’re changing multiple options on a page at one time, trying to determine the best conversion outcomes. An A/B test is changing one thing at a time, whereas multivariate test involves changing multiple things at a time and using sophisticated analytics to track.
How You’d Use This: Google Website Optimizer offers you the ability to do this, but it isn’t the best tool. I wouldn’t get into multivariate tests until you’ve mastered A/B testing.
This figure tells you, on average, how many pages a single website visitor views on your site during a single session. For example, if I looked at your homepage, a blog post, and then two of your package pages, that would be 4 pages for my visit.
How You’d Use This: The higher the number, the more “sticky” your website is. Most sites are in between 1 and 4. I wouldn’t worry too much about this one.
A pageview is a single person (“visitor”) viewing a single page on your website. So if you had 100 visitors and each looked at 4 pages, that’s 400 pageviews. This is the most common “traffic” measurement used currently.
How You’d Use This: Just like I mention below on visitors, this figure is a common way to measure traffic, though I always say it isn’t about how much traffic you have but what that traffic does. For a business owner, this is only meaningful in giving you trend information (example: “wow, we get twice as much traffic during the holidays, with another peak in April. We should rotate our deals and ads before those periods.”)
This is the percentage of the total visits that came from new visitors. A new visitor is simply someone without your site’s cookie present in their browser. As you can imagine, then, this figure has room for error.
How You’d Use This: Use this in conjunction with your marketing efforts; a high number of new visits is fine if you’ve just done a lot of SEO work, paid for some advertising, or other outreach efforts. It isn’t the percentage that’s important here, it’s why.
The referrer is the website URL that sent a user to a page on your website. Facebook, Twitter, Website XYZ, Newspaper column ABC are all types of referrers. Traffic that has no referrer information (they came from your business card, or their browser bookmarks) is called direct traffic.
How You’d Use This: This is important to see where your website traffic is coming from, and if it matches up with where you are spending your time/money. If you just paid big bucks for a Twitter ad and you aren’t getting any traffic, that’s a problem. You want to watch this for unexpected traffic, like a media/blog mention you weren’t expecting.
Analytics can result in a glut of data, for sure. Segmentation is breaking that data down into smaller piles so you can spot trends. For example, you could segment based on country or mobile phone users versus standard desktop users and maybe you’d see that 1) you have lots of mobile traffic and can justify site upgrades, or 2) nobody on a phone buys anything, what does that mean. Or you can segment off traffic from particular region to see the buying behavior of types of customers.
How You’d Use This: This depends on your business but take the opportunity to break down your traffic into meaningful groups if you can. Some ideas: Direct traffic vs referral traffic; New traffic versus repeat visitors; US/Canada visitors versus UK/Europe visitors; Mobile traffic versus desktop computer traffic…
Time on Page (or alternatively, time on site) is an average that tells you how long a visitor spends on a page or the site overall. Most sites have a widely variable figure for each page – homepages have short on page time as you’ll probably enticing them to move on – while other pages like where you’re making an offer will have longer times. So, careful with this one.
How You’d Use This: Overall, time on site is useful as a trend, but to make this data meaningful, you must drill down into specific pages and review the data. For example, if someone only spends 30 seconds on your about page, but 5 minutes on your sales packages? That’s not a problem. So it requires some thought to actually make this data useful.
An individual visiting a web site (excludes search engine spiders). Sometimes a media buyer or ad agency will call these” uniques” – short for unique visitors.
How You’d Use This: Just like I said above on pageviews, this figure is a common way to measure traffic, though I always say it isn’t about how much traffic you have but what that traffic does. I’d get more into the details, like “we have the same number of visitors each month of the year, but our Google traffic spends a lot more money than our paid traffic.”
Helpful? If you need to talk more about your stats, check out my top-to-bottom website review.