Sign up for our daily recaps of the ever-changing search marketing landscape.
Usage Data May Be Your Site’s Most Valuable Asset
Some will proclaim that the most valuable asset for their web site is their: technology, merchandise, content, service, (fill in the blank), because without it, there is no there there. But the answer may actually be usage metrics. Without metrics, your stuff is there, but you are in the dark.
Analyzing usage metrics is like turning on the lights. They let you see if you are doing the right things. They help you understand user behavior, which allows you to do more of what users want and less of what they don’t. This drives increased usage and revenue, and decreases wasted expenditures. Like the old adage in television advertising that says ‘you know half your money is wasted, you just don’t know which half,’ usage metrics can be the difference between guessing and knowing.
The advent of the web enabled a quantum leap forward in our ability to understand user demand and behavior. Users tell us every day what they like and don’t like on our sites. All we have to do is listen. The more we listen, the better equipped we are to deliver what our users want.
I call it democracy. A web page is like a ballot. Users click on what they like, they don’t click on what they don’t like. And if they don’t see what they want, the search box is a write in vote—”I want this.” Think about the power of these simple actions.
During my tenure at MSN, the team was able to understand what users were interested in at different times of day, different days of the week, and each day of the year. How? Because users told us. We responded by highlighting specific content and services when users were looking for them, and they rewarded us with increased usage.
At Microsoft Office Online, the team was able to identify highly rated content that was rarely read, and poorly rated content that was often read. We responded by promoting the highly rated content and fixing the poorly rated content to achieve higher user engagement and satisfaction.
So which metrics should we pay attention to?
Page views? Early on, web sites put counters on their pages to record page views, but a high number could be the result of one industrious user hitting refresh a lot.
Unique users? In response, sites began measuring unique users, but perhaps each unique user hit the homepage and left.
Minutes per session? As a measure of engagement, sites began reporting minutes per session, or how long the average user spent on the site during a visit. But high minutes on a chat site might have high costs, whereas low minutes on an e-commerce site might have high profits.
And so on.
No single metric defines the whole picture. It takes a constellation of metrics, and which constellation depends on the goals of the site and the sophistication of the system and analysts.
Page views may still identify popular areas of a site. Unique users are still a standard benchmark of reach, or popularity. And minutes per session and sessions per month are good measures of engagement.
Additional metrics include:
- Click-through rates on specific links and as a function of location on a page
- Search queries on your site to identify seasonal trends and new topics
- Entry and exit pages, referring sites, and search-engine keywords driving traffic to your site
- Click-stream analysis to follow user pathways, identify abandonment spots, and identify related areas within your site
- Multivariate testing to optimize combinations of factors on your site
And there are many, many others. Toss in financial analysis and you can pivot these in terms of revenue. Compare your data with third-party providers, and they will differ, because what seems like a simple counting exercise is significantly more complex.
If you’re new to this, take heart—it is OK to begin with simple metrics. For each question answered, a new question will take its place and spur you on to understand that next level of detail. I did not say metrics were easy; I said they were among your most valuable assets.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.