Can you explain how an SEO-optimized title on the front page deoptimizes current article results? I’m not sure I follow the thinking on that.

From a local news perspective, I found one benefit of an SEO-friendly headline on the front page is clarity for the reader. Writing a headline like “Town, county taxes set to increase” doesn’t do much to inform a reader, whereas the addition of proper names for the town and county does. That’s a simple example, but I think the exercise makes a huge difference. In fact, I consider it the most important consideration in writing a webhead.

The searchability of that headline is really the second consideration, in my opinion.

You’re right, though: This probably won’t come as a surprise to editors (or former editors, like me) who are (were) plugged into their analytics. In some ways these numbers deflate the “referral model” of news consumption via social media that many argue is becoming the norm. Do you view these stats as testament to the power of a brand’s name in drawing users directly to their site?

Steve, I think Dennis’ thinking is that while a keyword-dense, SEO-friendly headline may be attractive to Google, it may not be as attractive to a human being who’s deciding which (if any) headlines to click on your homepage. Maybe some of the traditional print strategies — a tease, a pun, something to draw the reader in without revealing all — might work better.

Of course, many (most?) large news organizations are already writing separate headlines for the front page, leaving the more dedicated SEO optimizing for article pages.

Understood, but I stand by my point of useability and clarity for the human reader. I think writing a good webhead is part science and part art form. You want to balance clarity, brevity, and archival considerations. The latter is especially true for smaller news organizations that probably don’t have the time to go in and edit the headlines of their archived articles to more SEO-friendly forms.

I can only speak for the content management systems I used during my newspaper stint, but here’s what should be included in any media outlet’s CMS: A/B/C headlines where you can write multiple heads to see which gets the most clicks (WordPress started doing this sometime last year), and perhaps even the ability to compose an SEO-friendly headline and set it to go live on the article once it archives. Maybe that’s already out there, though.

That said, I fully agree it’s WAY too easy to go overboard and write a dense headline in the hopes of attracting those spiders. “Late model blue Ford Taurus crashes into utility pole on Main Street near Thomas Jefferson Elementary School in Example City,” is a very slight exaggeration of a headline someone wrote in my old newsroom last year (and which I quickly pared down).

Though we are not a traditional newspaper, our site sees a much much larger portion of traffic driven to new content by social media, and indeed it is largely those pieces that quickly do well on Facebook or Twitter that perform for us in general. Though there is sometimes tension between the two, we’ve begun to think about social media optimization equally as much as SEO. The good news is that optimizing headlines and other content for social media sharing can result in really good front-page optimization since click-ability, rather than things like keyword density, are emphasized.

Does anyone else share this experience? Is 7% traffic to new pieces from social networks really typical?

One way to optimize your front page is to have it personalized to each user’s interests. DailyMe’s Newstogram platform can do just that.

>>so I wonder if the steady traffic contribution of front pages, like the unswerving love of a loyal but slothful dog, is sometimes under-appreciated because of more spectacular, but less frequent, traffic events.

This is exactly how it is for some properties – and one can only aspire to your eloquence in describing the phenomena in chasing hits versus that of serving your most loyal users.

I’m not saying that one should not chase the news aggregation lottery, as there’s certainly ways you can work that semi-effectively. AND therein lays my view, that the news aggregation lottery is just another channel, smaller and less effective in total than the front page, but surely a channel in itself.

have a great weekend
d. :-)

I’m quite skeptical of the numbers of above even by segmenting the data in the way you have, we see far more traffic come from social media and search.

Hi Vadim,

>>I’m quite skeptical of the numbers of above even by segmenting the data in the way you have.

The data is probably slightly controversial, but it is very difficult to dispute the results beyond questioning whether the segment of two dozen or so large volume publishers, which are included in the study, are a fair representative of the publishing universe as a whole. I personally believe that the publishers included are very traditional in usage patterns. Let me try to explain the results using a fictive but still fair case.

Let’s take the following example of an average 150M page views per month news-media destination (arriving from 30M Visits). A likely distribution of Visits could be something like this:

10M Visits from: Direct (primarily the Front Page)
10M Visits from: Search
10M Visits from: Referrers (such as Social)

This is usually where it goes wrong. When less data savvy publishers look at this picture I’ve seen them go look at something as simple as a top referring sites (or URLs) report in their web analytics solution. This is a good start, but probably not the best analysis to figure out the most important channels for Article Views. What almost always happens in this scenario is that Google ends up on top of that list and from there on a belief of being ‘search dependent’ starts.

This list is not an optimal analysis as Search is being dominated by one player (Google), Referrers are highly distributed among multiple sites (from Social sites to News Aggregators), and Direct traffic is essentially excluded (it is not a referrer in the typical web analytics sense).

Let‘s make a few additional (and very standard) assumptions for our 150M publishers above. The first assumption being that 7% of all search referrers are branded terms that take the visitor directly to a front page – in my book that is not search, that is navigational traffic. There is no real difference between a visitor writing “San Francisco Chronicle” in Google and that of going directly to sfgate.com. For referrers, a link with the anchor text “San Francisco Chronicle” linking to the Front Page of sfgate.com is again direct traffic to me, let’s make a cautious assumption of 3% of arriving on from branded anchor texts.

If we move the visits around based on those assumptions, we have the following visit distribution:

11.0M Visits from: Direct (primarily the Front Page)
9.3M Visits from: Search
9.7M Visits from: Referrers (such as Social)

Let’s make the assumption (widely accepted) that the average page-views-per-visit is not the same from channel to channel (equally loyal), and that those coming directly to the destination because of the brand are the most loyal. We could do a decent average of 7 PPV for Direct, 3.5 PPV for Search and finally 4 PPV for Referrers. Sounds fair?

If we move back into Page Views we get the following distribution:

77,000,000 Page views (11.0M * 7) from: Direct (primarily the Front Page)
32,550,000 Page views (9.3M * 3.5) from: Search
38,800,000 Page views (9.7M * 4) from: Referrers (such as Social)

This is approximately 150M (the size of the property as stated) and a distribution that concludes that 52% of all page views can be attributed back to the Front Pages (plural). In other words about half of all the page views on any given day can be attributed back to the front pages.

The Data in the Article itself concludes that of the 64% article views in the current article segment, 78% could be attributed to the Front Pages – which is 50% (64% * 78%). In other words about half of all the page views on any given day can be attributed back to the front page.

In conclusion it is natural to be skeptical, I was too, but the data tells us a different story than what we’ve heard the industry at large push over the last half decade. There is simply no escaping it – The Front Page is very likely the most important channel for most publishers today!

cheers
d. :-)

Thanks for the thorough response, and a well articulated one. However, I think the number by itself may be a bit misleading based on how most publishers analyze their traffic and as you mention would likely be quite different for non-traditional publishers.