Can you believe we started this little blog only 3 months ago? We’ve learned so much, and met so many awesome people, in that short amount of time. It’s hard to believe that we ever lived our lives differently.
We’ve had the great fortune, also, to grow in that time and break some of our early goals: having over 500, 1,000, 2,000 views in a day, having a week average over 1,000 views daily, doing our first product giveaway, etc. We’ve learned a lot – but not nearly everything, probably like <0.1% of everything – about what works for us and what we can do to help people find our site. Also, we WANT MORE.
Google Analytics is absolutely indispensable in this. It is an incredibly powerful tool that can give you so much information about what is working, and what isn’t working, to grow your blog.
However, like any powerful tool, it all depends on how you use it. We’ve only begun to learn the utility of it, as two people who love data and quantitative analysis. But we thought it would be fun to write up a little about the very basic stuff that we’ve learned so far, and how we use it to inform our growth.
Of course, the way to use stats will change dramatically depending on where you are in your blog’s growth. This is written from our perspective as a very new, very small blog, hoping to grow and find new readers.
How Does Your Blog Grow?
Views are great. One of our favorite things to do is watch the pageview count go up every day. They are the whole point, after all. But how do you get more of them?
Our “theory of growth” is that we’ll get more views if we:
- get more people in the door to see our site
- provide good content so they stick around and read many pages while they’re here; and thus
- convince them to come back frequently to read more.
That may sound pretty obvious, but it helps us stay on track through the ups and downs. Pageviews are fickle. We may get excited when we see a lot of new people visiting the site (#1), but if they’re only reading one page while there here (#2), it is unlikely that they will think to come back again (#3). So all that follows are the stats that we use to see how we are doing on those three steps.
Referrals and Quality of Views
To get people in the door, we put our content out there on other sites – on linky parties, on pinterest, on other blogs who are kind enough to feature our content, on craftgawker and dwelling gawker, Apartment Therapy, Design*Sponge, etc. Pretty much ANY way that we can get people in the door is a good thing.
Those people who come from one of these other sites are referred views. We see these views by going to Acquisition > All Traffic > Referrals.
There, we can see the sites that send views our way and some statistics about them. Of course, one of the things we can look at is number of views they send, but it goes much deeper than that.
All referrals are not created equal. We primarily pay attention to the “bounce rate” column.
As you can see, in this particular snapshot (which is from a random period in March) we had 485 views from Pinterest. Awesome! However, nearly 86% of them were “bounce” views, which means that they left the site from that same page they entered on. We don’t know if they thought we were awful and left right away, or read the whole thing and loved it, but they didn’t visit any more of our pages. Partly because of that, the average pages per session for visitors who came from Pinterest was only 1.18. Bummer!
By contrast, several lines below you can see that we got many fewer views from two “linkies” that we submitted our content to. These are sort of time-intensive to find, track, and submit to, but we’ve found that the people who come from these other sites are more interested in “getting to know us” than those from many other places. For these two, about 1/3 to 1/2 of the people who were referred to our site decided to read some of our other stuff. They saw more pages on average, and stayed on the site for longer. Maybe a few of them will come back again to see what new stuff we’ve been up to!
(You might wonder if we care about whether these views are New Users, and if we mainly want people to find our site for the first time. We don’t (yet). While we definitely want new people to come to the site, we know from our own experience that sometimes you have to “stumble” across a blog a few times for different reasons before you realize that this is a place you want to proactively return to.)
Going Deeper – Quality of Non-Bounce Views
So now we can get a sense of which sites refer “better” pageviews. But not even all non-bounce views are equal!
So far, we’ve just been looking at all views as a whole. To dig a little deeper, we add a “cut” of the data, by selecting “Add Segment” along the top of the page, and selecting “Non-Bounce Sessions” from the list.
Now all the stats that we look at will include a subset just for those who have decided to click around and see more than one page. Here is a good example from a period of time in February.
In this period of time we had our content featured on Craft Gossip and on dwellinggawker and got about 100 referred views from each. We could begin to see that the dwellinggawker views were a little better than the Craft Gossip ones by looking at the bounce rate, but still, 75% and 87% are not so far off as to be compelling. Where we really see the difference is in what sort of non-bounce views we got.
Those who came from Craft Gossip who decided to click around more after reading their original page STILL only visited an average of two pages and stayed on the site for about 4 minutes. By contrast, those from dwellinggawker who decided to click around more saw nearly 4 pages and stayed on the site for over 9 minutes. Seeing those numbers, we get hopeful we may have found a couple of new readers!
However, what this still doesn’t tell us is why those views are better. For some sites, it makes sense that there is a high bounce rate and “worse quality” non-bounce sessions. With Pinterest, for example, people who are browsing around there probably think they have a pretty good thing going on with Pinterest, and are not really interested in getting distracted by another website. We’d have to really “wow” them to get them to switch modes and check out our site for any period of time. On other sites – like blogs, the gawkers, etc. – people might be more amenable to checking out another blog.
So while it is possible that people from Craft Gossip are just more interested in hanging around Craft Gossip than people on dwellinggawker are interested in hanging around dwellinggawker, it is also quite possible that the post we “offered” them when they came to our site just wasn’t as good.
How Engaging is My Content?
To check that out further, we go to look at our statistics by post and page, so we can see what people like and what they don’t like. We look at this by going to Behavior > Site Content > All Pages. (Some of these other sub-pages have some good content, also, but we like “All Pages” as a solid summary view.)
There is a lot to learn on this page from a relatively small number of measures. We like to remove any other “cuts” of the data that I’ve been using (like the Non-Bounce Sessions segment I added in the section above) to make things a little simpler.
In this screenshot, showing a period of time in February, I’ve sorted by “entrances” because I am trying to see which posts are good for people to come into, and which are not as good.
Average Time on Page is an interesting measure, because it lets us know whether those who visit that page find it worth reading, or whether they leave quickly. Even though it is not quite apples-to-apples, we think it is pretty comparable, because while our posts do vary some in length they all have a fair amount of content.
Looking at these stats, I can tell that the 174 visitors who came over to check out my post about picture frames (marked with the frowny-face arrow) were NOT that impressed, since they spent less than 2 minutes on the post, and over 80% of them left after that. Obviously something about this post or project was not that engaging, and this isn’t the sort of content we should produce!
More standard is my post on fixing up some old drawers with paint and wood polish (the red box at the top) and Sage’s post on her $2 art and DIY mat cutting (red box at the bottom). The Exit % is only okay for both of these, but it looks like people spent time really reading these posts! We probably provided something “valuable” to them if they read the whole thing, so this is a good sort of content for us to provide.
However, reading time doesn’t always correlate with engagement. Visitors only spent about 2.5 minutes reading about our dual makeovers of thrift store lamps, but they must have liked what they saw! Less than 30% of them left after seeing that post. This makes us think that this was an engaging concept that we should try to do more of.
Most of this, and most of what we do in Google Analytics, has to do with judging our progress on step #2 of our “theory of growth.” However, over the long-term we track our progress not only by growth in page views (which are fickle and fluctuate a lot) but in growth in return visitors (step #3).
To look at this, we go back to the Audience Overview page (Audience > Overview) and add Returning Users as its own segment (like we did with Non-Bounce Sessions above). We like to look at the views over a longer period of time and group them by week to reduce the randomness and fluctuations.
The WHOLE POINT is for that orange line to grow!
We are obviously not even scratching the surface so far of what Google Analytics can do. They have so many different ways to look at your data, and because you can “cut” each of these pages by adding segments to look at in isolation, it is immensely customizable for whatever you want to look into at any given moment.
The measures I described here are just the few things that we have found most useful as beginners, and have kept coming back to over the last couple of months. I’m sure we’ll change things up as set new goals for ourselves AND learn more about Google Analytics.
What measures do YOU use? What tips do you have for us?