Interpreting Website Analytics
Understanding how to interpret basic website analytics will help you understand who your users are and what parts of your website are succeeding or failing.
In this article, we will explain common website usage metrics and how to interpret them. This is a skill that is highly in demand, whether you are trying to get a job as a product manager, technical marketer, or if you simply just want to figure out how to drive more traffic to your own personal site!
Metrics to track
Which metrics are the most important website to track? Here are some of the most common benchmarks that can give you a good overview of how users are interacting with your site.
Throughout this article, we will refer to a fictional pizza e-commerce site, Kiki’s Pizza Delivery Service.
Here are some key terms to know to interpret website traffic:
- Users: Total number of users that visit your site.
- Sessions: A session is a group of user interactions that take place within a given time frame, often ended after 30 minutes of inactivity or when the user logs out. The total number of sessions will be greater than the number of unique visitors because a user can have multiple sessions on the same day. The total number of sessions will be less than the number of page views because different page views will be grouped together into one session unless the user logs out or is inactive for 30 minutes.
- Bounce rate: The proportion of users that leave your site after viewing just one page. In the example above, we see a bounce rate of 33.14%.
You typically want your bounce rate to be as low as possible because you want to keep your user’s attention on your site and get your users to make transactions.
- Most online businesses want to lower their bounce rate, but higher bounce rates are normal for some kinds of sites. For example, news or blog sites that produce individual articles know that many users will just read one (if that!) and will likely expect higher bounce rates than other kinds of sites.
- Users are also more prone to bouncing when visiting sites on mobile devices. Users typically do not browse a site for content on mobile, they are likely to stick mainly to apps on their phone, and only look at sites for short periods of time.
- Session duration: How long is the average site visit? For a site that is content driven, getting longer site visits might be crucial to your strategy. For an e-commerce site, if your customers are making the transactions that you want, you might not care if they are taking longer or shorter amounts of time to do this.
Website traffic will typically be displayed over time, so you can look for patterns in traffic usage, and also see if different marketing and social campaigns impacted overall traffic.
Some amount of seasonality is expected in many industries. For example, there will be more visits to e-commerce sites in the period between Thanksgiving and Christmas. It’s more important to compare performance against your projections than to take every peak or valley of your traffic graphs too seriously.
How is Kiki’s pizza delivery service doing? It looks like traffic and sessions are down over the last week, and the bounce rate is up! That’s not good news. However, this might make sense as this period lines up with Thanksgiving. You would probably expect fewer people to order pizza when they are preparing Thanksgiving dinner or eating leftovers. You might want to check out a longer time period in the analytics dashboard before drawing any conclusions.
Understanding Traffic By Devices
This figure is fairly self-explanatory: it breaks down the number of sessions by the type of device. The “desktop” category here is somewhat misleading since it covers both laptop and desktop users. Desktops and laptops have very similar capabilities and displays, so this isn’t something that we often want to separate with analytics.
Make sure to note both the percent and the trend: in this example, mobile is a small contributor to the overall traffic, but it’s rapidly growing!
For most sites on the internet, the division between desktop and mobile visits is about 50/50. However, the average session time is much longer for users accessing sites through a desktop.
The ideal balance between desktop and mobile is different in every site and industry. Sites related to entertainment, social networking, and food and drink are more likely to be mobile-driven, and sites related to education and finance are more likely to be desktop driven.
Understanding Traffic Channels
Search: how many users came to your website through search engines? Google is by far the most common search engine, with Microsoft’s Bing a distant second, and all others (Yahoo, Ask) delivering much less traffic. You can improve your website’s ranking in search engine results in a number of different ways, such as writing descriptive headlines, using common keywords, and creating content that will drive other sites to link to it.
Referral: how many users came in from links on other websites? Getting mentioned on other websites shows credibility, drives additional traffic to your site, and will also generally improve your site’s position in search results.
Direct: how many users type your domain name directly into their browser? Direct traffic indicates loyal, repeat visitors.
Social: how many users come to your site through social networks, like Facebook or Twitter? Good performance from social traffic sources indicates that your content is timely and sharable.
Other: Analytics tools don’t always know the exact origin of traffic, especially if users block cookies. Visitors that end up in “other” probably come from advertisements that haven’t been properly tracked.
Creating diverse traffic sources and growth in all possible channels is typically the best long-term strategy, though it’s reasonable to set other priorities in the short term. If your website is overly dependent on one channel, it can be vulnerable to external changes. For example, many publishing platforms like BuzzFeed and Upworthy that relied on viral sharing via Facebook have needed to drastically adjust their strategy as the Facebook newsfeed algorithm continues to change.
How is Kiki’s Pizza Delivery Service doing with respect to traffic sources? It seems that the majority of the traffic comes through search, which indicates that the page is well optimized for search engines. The amount of traffic through social referrals is low, so it might be worth investing in shareable content (like blog posts), building up a social media following, or offering social media referral discounts for people that share coupons.
Transactions and Conversion
Conversion is a term that refers to how frequently your users are engaging in the actions that drive the desired business outcomes. The conversion rate is the number of people who make a transaction divided by the number of total visitors.
What you use to define a transaction is intentionally left vague. A transaction on an e-commerce site often means purchasing a product. But for some websites, a conversion might mean subscribing to an email list, downloading an application, or enrolling in a free trial for a monthly recurring membership.
In the example above, you’ll see that things like “Saving Payment Information” and “Setting Up Pizza Profile” are considered transactions, not just discrete purchases like “Ordering A Pizza”. Presumably, Kiki’s Pizza Delivery Service investigated these actions, and found that they led people to make more purchases down the line, so they are useful to track as transactions even if they don’t immediately create profit.
Conversion rates by traffic source
It’s often useful to break down conversion rates across different traffic sources. Your largest traffic source isn’t necessarily your most valuable traffic source. This varies from site to site, but direct traffic often leads to higher conversion rates than search and social traffic, since it usually represents customers that are already very familiar with your site.
We’ve covered a wide variety of analytics:
- Total Traffic
- Users vs. Sessions
- Traffic by Device
- Transactions and Conversions
None of these metrics tell the whole story of who uses your website and in what ways, but as you combine them and track them over time, a much fuller picture emerges that can help drive improvements to your website and guide marketing.
Website analytics can also be used to inform more in-depth user research, such as focus groups and usability testing. Because user research can quickly become time consuming and expensive, you want to do everything you can with website analytics to make sure that you identify the right users and user personas before doing more in-depth research.