As promised, here’s the first in regular monthly updates on how Search Engine Land is growing. I hope the traffic statistics look will be both interesting to the curious and informative about how various places can turn into traffic generators.
For December 2006, we had about 93,000 page views, or about 4,400 page views per day. Actually, those are "AdViews," the number of times ads were shown as reported by our ad serving software. We only began showing ads from December 11 onward, so I have to turn to our Google Analytics statistics to talk about the entire month. Let’s dive into them starting with a chart:
The chart shows you the overall trend of visits and page views for the month of December. Note that for the first two days after we launched, December 12 and 13, we have no stats or minimal ones because our stats tracking code was accidentally left off as we brought the new web design up.
Overall, we had about 135,000 page views for the month, conservatively rounding up for the missing days — or about 4,400 page views per day, matching what our ad server reports. Visits were about 85,000 for the month or about 2,700 per day.
The two big spikes are traffic from Digg. I already covered the first time we hit Digg in December. The second time was for my story on the disappearing sex blogs at Google. FYI, we also hit Digg a third time just after the New Year, for my 14 "Is Google Evil?" Tipping Points Since 2001 article. That also resulted in a big traffic spike.
There’s no denying that Digg is a huge traffic driver, especially for a new site. Search Engine Watch, my alma mater, was Digged occasionally. Given the huge traffic it received from other sources, Digg made a spike but nowhere as high in proportion as I see with Search Engine Land.
That’s something I want to correct, of course. Problogger had a nice look recently at how Digg spikes can help build a traffic up over time. That will be great if it happens here. However, that’s not what I’m building Search Engine Land upon.
I don’t want Digg to be my primary traffic driver any more than I want to have a site completely dependent on Google sending traffic. A single source simply leaves you vulnerable if something goes wrong with that source. More and more sites are reporting that they’ve been banned from Digg, for example. That’s dangerous for a new site, if their traffic strategy is all about Digg.
So for me, Digg traffic is a nice surprise, the frosting on the cake — but I’m baking a cake with many traffic source ingredients. I’m successful if the overall line between the spikes continues to rise.
Of course, December was a tough month to look at traffic patterns. Many people aren’t online as much during the holiday period, at least for a "work blog" like ourselves. In addition, we didn’t formally launch the site until December 11. January will be our first real "normal" month.
Still, there’s plenty to learn from December. What were those other ingredients in the traffic cake? We had 1,417 overall traffic referral sources. Here were the top 25:
|Direct Navigation (people estimated to have typed in our URL directly)||
|Google (search related traffic)||
Google Personalized Home Page
|Search Engine Watch Blog||
|Search Engine Watch||
|Search Engine Journal||
|Search Engine Roundtable||
|Search Engine Guide||
There were some big differences from my preview of December stats just after we launched. Digg, I’ve already covered. Direct Navigation I didn’t mention before, because I’d used a slightly different chart from Google Analytics, Referral Conversion. That shows all the traffic where there’s a known referral source as reported by your browser (IE, the browser tells your web server the last place it came from). Instead, the chart above is the Source Conversion report. That shows all sources sending you traffic — referral links, search engine traffic plus direct navigation.
Direct Navigation From Press Mentions?
One thing I was curious about with Direct Navigation was whether press mentions made any spikes. I get a lot of calls from the press. Some of them result in quotes and mentions of the site. Typically, these online mentions won’t have direct links. But does having the name of the site out there alone generate more visits?
One recent article I remember was USA Today looking at Google’s traffic. That came out on December 28, so potentially I should have seen a rise. And the chart says?
Indeed, there was a spike. Direct Navigation sent 806 visitors on the 27th, the day before the article. That dropped to 682 ,the day of the article. Then it leapt to 1,536 the day after it ran (and 940 on Saturday, December 30).
So there’s your proof. Press mentions do pay off, even without links. Well, maybe. Why didn’t the traffic go up on the day of the article? One reason could be that the story also hit the wire services, so it might have made it into other papers over time. Plus, people might not have gone online to check out the site until after reading it.
Also complicating things is the fact the Direct Navigation spike came around the time both Digg and Boing Boing sent a lot of traffic. That spike might be due to people simply hearing about the site from these sources and visiting it directly, perhaps after an initial visit.
Going forward, I’m going to watch Direct Navigation traffic more closely after a press mention, to see if I can pin that down more definitively.
Our third big driver of traffic, searches on Google, I’ll deal with separately below. Similarly, I’ll deal with Google Reader and other feed reading services in a section on feeds.
Drill Down On Traffic Sources
Boing Boing was a new traffic driver for us, all related to the sex blogger story. I think there’s also a lesson here on how traffic can result because of an active participation in your community, as well.
The sex blog story hit when I was off for Christmas. When I’m gone, I usually take a day upon returning just to dig out and catch up on things that are going. Then for a big, complicated story, I want to spend more time trying to look at what’s happening rather than just blogging away.
I’d seen the sex blog story as part of my regular review, when I got back. It kept growing and growing until it hit Boing Boing. I reviewed that story, and I couldn’t help but feel that Google deserved some more credit than it was getting. I dropped an email over to Xeni Jardin, who I’d talked with last year several times about Google issues. Could she add a postscript with some of my comments? I thought they’d help provide some balance.
Xeni did (thanks, Xeni!). And there was even a link, which started driving traffic. But getting a link wasn’t my prime motivation. Trying to get some balance and information out was. I routinely comment on blogs or email people on issues relating to search. I put a considerable amount of time into it, actually. I do this because I’m passionate about the space. But a side benefit is that occasionally, that type of sharing gets returned as people link to you.
I went on and then did my own story on the topic. It was a long look at the sites involved and some of the issues. This type of original content is both my chief goal for Search Engine Land and for myself personally. I know we do a lot of referring to what others are writing about. That’s not going away. People write great stuff, and I want my audience to know about that whatever the source. But I personally find nothing better than being able to roll up my sleeves and dive into a story to try and make sense of it for others.
The story resulted in a second Boing Boing mention — plus it also got Digged. Lesson? Good content gets rewarded, something I’ve talked about since I started writing about search engines back in 1996.
Now what’s that The Daily that sent so much traffic? Pretty simple. My guess is that whoever runs it noted that I was listing all my referral sources when I did my stats preview and figured if they could send me enough traffic, they’d get listed. You did.
The Daily has a number of start pages. Perhaps you’d like to see a different woman in a bikini each day. Here you go. And right at the top, there are three links. For a period in December, we were one of those three links. Hey, it’s not all about bikinis. You can also see a different scenic shot each day, or a different animal or just a page of links. But yeah, it’s the bikini page that seems most popular. Of the 2,200 visits from the site, 74 percent came from the bikini page, followed by 19 percent from the scenic page, then the rest from others.
Planet Internet was another traffic driver. This came off a news story over there, referencing my The Lies Of Top Search Terms Of The Year article. Again, good content (well, I thought it was good) can bring in the traffic.
In my preview, I’d already mentioned how Search Engine Watch was one of the top referring sources to Search Engine Land. That remained the case, as the month ended. From the SEW Blog, about half the traffic (47 percent) came from my going away post, with another 17 percent from Elisabeth Osmeloski’s best wishes post. With the going away time now over, I’d expect that traffic to drop. FYI, Daggle, Search Engine Roundtable, Google Blogoscoped and Search Engine Guide sent traffic primarily based on the site launch.
A more typical "spike" from Search Engine Watch would likely be the 148 visitors referred our way from a post about our article on Google Patent Search, for example. And that’s another lesson about original content. If we’re not providing something new, we aren’t going to get the referral traffic. Rest assured, original content will continue to grow. That article was also the top traffic driver for us from ResourceShelf.
Several other search engine blogs I’d noted in my preview still made the top list. Threadwatchers especially loved the news of our SMX search marketing conference happening this June, judging on clickthrough from the site. SEOmoz sent a bunch of traffic, but because of the way Google Analytics drops the ?ID= portion of SEOmoz’s URLs, I can’t tell what particular topics drew interest. Grrr — stop that, Google Analytics! I’ll do a post on this in more depth in the future. The same problem means I can’t tell what Search Engine Journal readers liked.
John Battelle’s SearchMob — a Digg for search stories — showed some legs, with a post there about me sharing site stats sending along 61 visitors. Most traffic from SerachBlog seemed related to the quick news we posted on the Yahoo reorganization. That story also sent traffic from Scripting.com.
Our post no real original content. But I had the release up from what Yahoo emailed me faster than I think Yahoo itself got it into a HTML form that people could link to, plus I quickly added links to more commentary. While original content is key, covering key search news even if briefly is part of our mission — and that, too, can pay off.
Finally, I’ve only listed the top 25 traffic sources. Remember, I said there were 1,400 in all. The other 1,375 collectively drove about 15,000 visits. There’s a long tail to referral traffic, just as there is with search traffic. As a new site owner, I deeply appreciate those links and references from any site, even if they aren’t in the top 25.
Feeds, Feed Traffic & Email
Search Engine Land offers a variety of feeds and email newsletters, as covered here. Getting visitors to subscribe to one of our feeds or an email newsletter is my primary goal with the site. If visitors do either, then I have the ability to continue my relationship with them more easily over time. How’s the relationship building going? Let’s do some charts:
That’s the subscribers to the main Search Engine Land feed since it was offered when the placeholder site went up back in November. The stats are as reported by FeedBurner. Unfortunately, I can’t narrow it down to just the month of December. It’s either one day, last 30 days or all time — so I thought I’d kick off with the all time view.
Overall, the feed is showing a pattern I want to see, spikes each week that keep getting higher. It plateaus a bit during the holiday period, which is understandable. Fewer people are hitting their feeds then, and subscribers only register as they make an active call on their feeds.
Over at Search Engine Watch, our feeds grew to hit the 15,000 level over time. So that’s my goal with Search Engine Land, hit at least that level or try to go past it. I know there’s at least that many people interested in search news on a regular basis.
(FYI, any Search Engine Watch stats I mention here or going forward are all in the public record, things that have been reported on Search Engine Watch in the past or currently. So I’m not passing along anything confidential, in case you were wondering).
I was figuring you had loads of traffic already with most SEW readers converting to SEL?
The reality is that Search Engine Land is new, habits are hard and readership is broad. This site has a lot of goodwill from the community, but it still has to prove itself over time as deserving a larger readership beyond just the core search community. In addition, it’s one thing to tell people you’ve moved and another thing for a chunk of the audience to make the actual effort to add a new feed.
In short, Search Engine Land is very much an underdog. And that’s fine – I’ve always liked underdogs. My goal has always been to hit that 15,000 level not in the first month but by the end of the first year.
By the way, I mentioned a "core" search audience. Look back at the feed, and you’ll see how it quickly spikes up to the 1,500 level. I think that may be the hard-core search audience that’s out there, the influencers. Or maybe it’s just 1,500 people who like me and the others here on Search Engine Land!
Where are these people coming from? FeedBurner breaks them down like this:
Google Analytics tells a different story, but we’ll come back to that. Let’s look at the SearchCap feed, first:
SearchCap is a post we do each day recapping everything that was on the blog. You can take it in feed form, but it’s also offered as an email newsletter. Given its alternative newsletter format, it shows two key differences from the Search Engine Land feed:
- Weekend drops are minimal
- Most people get it via email
Look at the Search Engine Land feed, and you’ll see there’s almost always a midweek rise, then a plummet at the weekend, then a new midweek rise. That mirrors a regular work week, where you tend to have a lot of people online during the week — and most especially in the middle of the week — and offline on the weekend. The feed stats, as I said, reflect feed calls. You only have a "subscriber" if that subscriber issues a call to get your feed.
Newsletters are much different. Once you have a subscriber, that subscriber stays with you every day, even if they are offline. You only lose them if they unsubscribe. So the stats are more regular.
In fact, the drops you see for SearchCap are only due to the relatively few people who take SearchCap by feed. They contribute to the tiny midweek rises. If they weren’t there, SearchCap would be mainly a steady rising line.
Of the roughly 2,000 SearchCap subscribers we have, about 1,600 of those take it via email newsletter. After that, Google Desktop is the next biggest draw, 112 subscribers.
As with the Search Engine Land feed, I also have a goal for SearchCap. That’s to hit 35,000 email readers in the next two years. Over at Search Engine Watch, the daily SearchDay newsletter is around the 35,000 mark in readership. That largely built up over a course of four years, as Chris Sherman took the helm of it back in early 2001. Once that figure was hit, it stayed largely stable over time. That makes me feel there’s a core email newsletter audience of that size. But gaining email subscribers is harder in these days of increasing spam and feed alternatives. Two years might be too optimistic. We’ll see!
That leads me to Search Month. I ran the Search Engine Report monthly newsletter for a decade. Now I’m doing the new Search Month monthly newsletter for Search Engine Land. That takes me from a readership of about 90,000 to 100,000 to a current one of about 700.
Ouch! I don’t know if we’ll ever hit the 100,000 level, something that took Search Engine Watch about five years to do. I’ll go with a more realistic goal of 50,000 over the next three years. One thing that helps is that the first issue is out. Check it out — I think anyone who wants a regular monthly recap of search news will find it useful.
By the way, both SearchCap and Search Month are full-text feeds, unlike the Search Engine Land feed. The main Search Engine Land feed may go full-text as well. There are some issues to consider there, and I’ll explore these more in a future post. But I’m leaning that way for a change in the near-term.
Now back to Google Analytics. Remember my big chart of referral sites above? Here’s the breakdown of sites that are newsreaders or email referral sources:
Google Personalized Home Page
That combined Google Reader and Google Personalized Home Page figure? Google Analytics actually reports that as "google.com [referral]." But when you drill into it using the Referral Conversion tool, you can see that about 2,600 visitors are coming from Google Reader (google.com/reader/view) and about 1,400 from Google’s Personalized Home Page (google.com/ig).
Both of these are feedreaders. Neither is showing up in my FeedBurner stats to any significant degree. That’s probably down to Google not providing readership stats the way some of the other feedreaders do, but it might also be that FeedBurner needs to make some of its own adjustments. I’ll be checking with both companies and do a follow up in the future.
Also keep in mind that Google Analytics is showing page views, while FeedBurner is showing feed views. A chunk of people might not be loading up our pages, simply reading whatever is in the feeds. FeedBurner can track those people, while Google Analytics is tracking those who come to the site or load content directly from the site.
News Search Traffic
As a news site, Search Engine Land naturally wants to tap into a key vertical search space that’s suitable for us, news search. In my preview, I’d noted how the site had already been picked up by Techmeme. Traffic from there continued on, ending the month with 1,500 visits.
That’s nothing compared to Digg. However, Techmeme is heavily read by influencers, people I know will be linking to us or making mention of our stories. I love being part of it and appreciate the traffic.
Indeed, out of curiosity, I checked my new visitor to returning visitor stats. For Techmeme, about 40 percent of those coming are return visitors. For Digg, that’s about 4 percent. Now in raw numbers, that 746 people from Digg as returning visitors to Techmeme’s 600. So Digg still can produce a net gain, but the audience at Techmeme is clearly much more different and much more in tune with what we’re writing.
In my preview, I also discussed how much I missed being part of Google News. Later in December, when I felt we had enough content to prove our worth, I formally made an inclusion request using the Google News inclusion form. We were added about a week later. In that short period, Google News sent over 1,200 visitors.
FYI, I plan to revisit both Google News and Techmeme from an inclusion standpoint in the near future. So stay tuned.
Keyword-Driven Search Traffic
Now for dessert, search marketing stats. Here are the top terms that drove us traffic in December
|search engine land||
|tracking santa 2006||
|norad tracking santa||
|norad tracking santa 2006||
|santa and norad||
|santa on norad||
|tracking santa norad||
|top google searches 2006||
|tracking santa on norad||
|mobile local search for the rest of us greg sterling||
As was the case in our preview, our top terms tended to be those relating to our name.
The terms google and goog are mostly driven off of Google. Google Analytics unfortunately doesn’t break down where on Google these are coming from — Google web search, news search, blog search, etc. I hope they’ll change that or make it an option in the future. But from spot checks I recall doing, these seem to be from Google News and Google Blog Search. That underscores the importance of vertical search. Hey, I’d love to rank tops in Google web search for "google." But there’s plenty of traffic for the term in other types of vertical search areas. Don’t neglect those.
You see a bunch of terms relating to tracking Santa. These come mostly off Google web search, from a post I did in December
Another term that leaps out is "search engines." That’s us ranking in the top results for search engines at Ask. I’m used to seeing traffic for that term when I was at Search Engine Watch, since after being a resource for 10 years, the site’s in the top results of Google, Yahoo and Ask for that phrase. I’d obviously considered "search engines" as a key term I’d like Search Engine Land to be found for. However, I didn’t expect this would happen for at least months.
Instead, sometime in December, we hit the first page of results at Ask. Currently, we’re on the second page. I’ll follow up more with Ask about this, because the first thought many may have is that Ask has simply altered the results to get Search Engine Land higher. Alternatively, Ask has long talked about ranking sites not based on all links from across the web but links within a community of sites. In that case, it makes more sense for Search Engine Land to do better, if it has quickly gained enough trusted links from within a relatively small search community.
Just to benchmark, at Google, Yahoo and Microsoft Live, Search Engine Land is not in the top 100 results for that term. Here’s hoping that will change over the next few months. Any link donations to help make it happen are happily received!
In my preview, I talked about how paltry my query stream was. Overall, it’s not gained much, but it’s still early days in the content building business. But then again, don’t forget the search tail. The terms above make up about 3,000 visits, about half the search related traffic. Another 3,000 visits came from the remaining 1,900 or so terms not shown.
Two more issues on the search term analysis. Remember above, I said Google sent search related traffic of 6,172 visits? That’s also precisely the same amount of traffic of ALL search related traffic that search engines have sent, according to Google Analytics. The figures shouldn’t be the same. The figure for ALL traffic should be Google’s traffic plus some of the other search engines, making it at least slightly higher. I suspect that Google Analytics is counting things differently using different reports, and I’ll try to track this down further in the future.
Meanwhile, here’s the breakdown on traffic sent by search engine, according to the ALL search engines report:
- Google: 5,618 visits
- Ask: 187 visits
- AOL: 117 visits
- Yahoo: 114 visits
- Live: 10 visits
- Others: 126 visits
Overall, Google sent 91 percent of our overall keyword driven traffic. For this site, at least, that helps confirm the view of Google being by far the top search engine based on referral traffic.
Finally, what did most people read? Here are the top articles: