COMMENTS
Love the article and the metrics are right on - especially the bounce rate. Part of the battle is knowing what to look for, the other part is knowing what to do once you've discovered these idiosyncrasies. One of the neat things that we do with our clients is help them to make the changes that need to be done after evaluating the analytic information. As a
Hubspot Certified Partner our goal is to help to make websites much more efficient, which is what the Hubspot software is terrific for!
I'm still very much an apprentice internet marketer; articles like this one are incredibly helpful. I especially appreciated the information about comparing bounce rates, and analyzing visitors to my site. Thanks!
Regarding "a page with a high bounce rate is performing poorly"... For argument's sake, let's assume the following: 1) the page in question is a blog page, 2) search keywords that brought the visitor to my page are a little outside the target being aimed at (e.g. the visitor is not targeted traffic). Given that, is there any best-practice to get un-targeted traffic to somehow react, engage, or otherwise do something besides hit the back button?
Curt, that's a great question. If you have a blog article that covers a topic that's slightly outside the realm of the rest of the content on your site, it might be harder to convince that visitor to click around some more.
Ideally, you'd want each piece of content to have a logical connection to the rest of the content on your site, but sometimes you might need to nudge the visitor to make the connection.
At HubSpot, we add calls-to-action to the bottom of every blog post, and try to match the content of the post with the content on the landing page as best we can, to encourage first-time visitors who found us on search to continue clicking around. Targeted CTAs on blog posts is a great tactic that I'd recommend for your blog.
Good general advice, but one thing that could be stressed a bit more is that each of these statistics should be correlated to the "obvious key metrics" of leads and customers - which, ultimately, is your goal.
For example, a page with a higher bounce rate may actually generate more sales or leads than a page with a lower one - sure, more people are bouncing out, but more of the ones who stay become buyers.
In that sense, the page itself is likely effective, but the methods used to drive traffic to that page are drawing in a lot of people who aren't likely to buy. Eliminating or retooling such a page likely does more harm than good.
Knowing the basics is important, but asking "why?" (and sometimes, asking it three or four times) is also critical to not being misled by your own numbers.
@Jim, that's an interesting thought. So, you are saying that the bounce rate may be acceptable, and that the page may still be effective. If the search is purely organic, then wouldn't that imply that you could be using poor choices of keywords? Are you suggesting traffic sent to the page by other methods, like PPC?
I've always measured against unique visitors. Its great to finally have an industry leader confirm unique instead of all visitors.
Cheers
-Christina
Hartley, I agree that these are all very important metrics. We closely monitor each as well. With respect to conversion rate, marketers may also want to consider integrating call tracking to capture conversions that take place offline (i.e. phone calls). The combination of online and offline conversion data will help to establish a more accurate conversion rate, which marketers can then use to evolve campaigns and adjust budget allocations.
Can't agree with you on the bounce rate.
I guess I'd start this way: are you really getting enough traffic to your landing pages to have a hope of finding a statistically significant bounce rate? For the vast majority of B2B, the landing pages get so few page views that the bounce rate is just noise.
The second question I'd have is that bounce rates say as much about your link traffic
source as they do about your landing page. Say you get a really useless link from Miley Cyrus's homepage, and that creates a really high bounce rate on a page. Then you optimize that page, and you find that pink and rainbows and sparkles increases your conversion rate. Your bounce rate goes down. But your business is terrible - these 11-year-olds aren't buying whatever you're selling.
What you should be doing is closed-loop marketing. You can bounce 99.99% of your visitor traffic from a landing page, but still have the highest ROI. Track which sources are driving the most leads and customers.
I'm a big fan of optimizing landing pages, but I don't use bounce rate to do so.
@Jody: that's more or less what I'm saying - that bounce rate (or any secondary metric) alone shouldn't lead to an automatic reaction because the problem may not be what it immediately seems.
Poor keyword choice, a badly targeted marketing campaign, and various other phenomena can cause a page to get a high number of visitors who have a low level of interest - but at the same time, the page may still be entirely effective in communicating the value of a product to those visitors who are "right" for it. In which case, the action to take is to revisit the traffic sources to appeal to the right visitors rather than re-touch the page to attempt to hold the attention of the non-buyer (and possible decrease the value for better qualified buyers).
It's still a red flag, to be sure, and I'd suggest the same is true for each of these metrics: they should raise concern, but merit a bit closer inspection before taking action to make sure you're "fixing" the right thing.
Excellent post and very timely for me. I have a new site and am just learning to use analytics to my advantage. Thanks for the great advice on using metrics -- it will be put to good use!