From the earlier article that I wrote, Social Media Under the Microscope, a lot of conversation was spawned as a result of the data findings. Many questions seemed to have been answered, as many people responded by confirming the same data on their sites. However, there were some new questions created from the data.
The most fascinating questions revolved around defining the difference between different social media technologies, such as blogs, forums, online news sites, and social networking and bookmarking sites. Using the same data, but looking at it in different ways provided some very amazing trends. So, with new & improved charts (complete with fresh new colors) and additional tools to dig into the data, (thanks to ClickTracks) I began the process of analyzing different forms of engagement based on visitor referral sources.
Defining Social Media
Because of the strong differences in engagement and context, I have had to divide the general term of social media in order to properly label and view the visitors from these sources. Because these groups view content in very diametrically opposed methods, they must be separated and defined. I added Web 2.0 customer review sites, since they are social media-based websites.
- Social Media: Blogs, Forums
- Social News: Digg, Reddit, Delicious, Netscape
- Web 2.0: (Yahoo Local, Amazon reviews)
Using one of the more recent events that rippled through the blogosphere, Jennifer Laycock’s run-in with the National Pork Board, a lot of data was built. Jennifer’s original blog post made the front page of the major social news sites, attracted public relations and online news coverage, and made headlines in the mommy bloggers and parenting forums. A data junkie’s dream, this provided compelling data to analyze as there were a variety of visitor referral sources and long term data.
One of the best ways of analyzing visitors is not to get distracted by the big numbers. When building comparisons from referrers, one has to look at the goals of the site. Especially for content producers, making the site “sticky” has to be defined. What makes a successful visit, even if there is no conversion? Any site manager should have to answer that question, as a good customer experience is what makes people come back, even if they do not purchase or become a lead on the first visit. Chances are they won’t. So how do you know if you are taking care of your visitors? This is where engagement metrics are so important.
First, define engagement. Define a successful visit to your website. A combination of time on site and pages viewed were the logical choices for this project, as Jennifer writes a content-based blog. She does sell shirts (which got her in trouble in the first place), so that as a conversion as well, even though it is not her primary activity.
Second, define your audience. This may sound impossible at first, but consider where your audience comes from: search engine queries, website links, direct access or bookmarks. Now, what are those people looking for? Search queries are not that hard to aggregate. I suggest creating “buckets” of keyword concepts. Rather than isolating a specific term and counting visitors, widen the scope and create a catch-all phrase that will capture as many of the related terms to a particular concept as possible.
For example, a site that sells lighting may want to filter the search queries for the phrase ‘ceiling fans’. Rather than waste time trying to capture every single variation of the term, use the single word ‘fans’. This will help you to view the search trends in that vertical rather than getting sidetracked on tracking the specific phrase. Depending on the size of the site, there can be hundreds of related terms within the segment. The more segments you create, the more data you have to compare. Comparison also takes on a new aspect when comparing similar terms within the vertical, rather than comparing all of the terms in one list.
The idea is that by segmenting out each of these key concepts, rather than specific words, you can better identify what each group of searchers is looking for, and then better identify what your site is delivering to them.
In looking at Jennifer’s traffic sources, we saw a significant distinction in the audience:
- Parenting Forums (two popular sources)
- Mommy Bloggers (primary blog referrers)
- Web Marketing Blogs
- Social News Sites (Digg, Reddit, StumbleUpon)
- Search Traffic (four primary keyword verticals in the strategy)
This created a wide range of audience characteristics. On one hand, Jennifer is a recognizable member of the search marketing community; on the other she is a mommy blogger and a breastfeeding activist. This event brought many different groups together, simply by nature of the situation; the corporate giant with an overzealous attorney going after a work at home mom with a blog.
click for larger image
High Audience Engagement
Based on engagement factors, the group that was the most engaged, and with a very respectable conversion rate, were the visitors from Salon.com. These visitors tended to stay the longest, read more pages, and consistently browsed the shirt selection. It was a surprise that this group was the most engaged, but something resonated with them as the article was published right in the middle of the controversy.
Good Audience Engagement
Blogs & Forums
The second most engaged group was a mix of blogs and forums. Typically, the blog referrals stayed longer and read more, but the forums were not far off. This was not surprising, as the highest referring blogs in terms of numbers were other mommy bloggers. The forums were well known parenting forums which brought the article to the attention of other moms and dads. There was also a distinct difference between the mommy blogger referrals and the search marketing community referrals. Obviously, this situation resonated more with the mommy bloggers than the search community. The marketers were more interested in the David v. Goliath match-up and link bombing the Pork Board. The news really spread in the mommy blogger world, as many were personally offended at the attorney’s remarks. These groups, while the engagement was lower than Salon.com, tended to convert and buy shirts at a higher rate – not a big surprise.
Closely related but less engaged were the searchers. Of course, the search traffic around the event lagged based on the search engines. However, there were changes in the Google index within hours of the first post. Two days after the original blog post, Jennifer’s site, TheLactivist.com, ranked #6 for the search term, “National Pork Board.” The engagement rates varied based on the terms they were searching for, which provided a very insightful view into the engagement by search term “group”. The conversion rates were lower than the blog referrers, but there was also differentiation among the conversion rates based on search terms as well. Based on what people are searching for, they view a site very differently, which leads to understanding the context of the search and the searcher’s expectations.
Low Audience Engagement
The lowest engaged group is the social media group, whose primary demographic is twenty-something tech-oriented males. However, the behavior shown here (lowest time on site, rarely more than one page view, no conversions) is not specific to this site. It is an occurrence on almost every site that is subject to attention from the techie world. From Slashdot to Digg, the attention from techies who share stories from around the world is nothing new. Consistently, the referrals from those types of social news sites all follow a trend of low engagement and rare conversions.
The “Long Tail” of Referrers
Most web marketers have heard of the Keyword Long Tail effect. See Keyword Long Tail for more info:
However, one of the more exciting things noted in the analysis was from the blog and forum referrers who sent traffic from the initial visit link. As time went on, those referrers tended to link to Jennifer’s site again and again, especially as she broke new stories specific to the parenting and breastfeeding communities.
This is the critical long-term data observation.
TheLactivist.com attracted attention from a wide variety of sources, yet the primary message is breastfeeding rights, parenting, and activism. Those blogs and forums, specific to that audience, that initially found her site from the Pork Board suit, continued to link to her site because it was relevant to their message and audience.
While the Digg, Reddit and other socal media did the “Flash Mob” thing . . .
Her other referrers, many of whom found her site from the Pork Board story, continued to send visitors and link to her blog. This is the long tail effect for links.
When new bloggers and opinion leaders find your website and it resonates with them, they tend to link to it more often, thereby sending more people over the long term than in one specific instance. Interestingly, there were other stories that drove more people from these blogs to her website than what was drawn by the original article!
Context and Competition
Here are the keys to developing this effect for your website. Context and Competition for Attention.
The principle of context is simple, we each practice it every day. When mommy bloggers and parenting forums linked to Jennifer’s story, the context was clear, they were mothers outraged at the comments of the Pork Board lawyer. Jennifer’s blog was sympathetic for them, as she was a work at home mother being harassed and bullied by a corporate giant. Therefore, the context of the link in those blogs and forums was very high. Conversely, the competition for the reader’s attention was very low. When a blogger links to another site it is usually supported by surrounding information that is relevant, powerful and the next best thing to a word-of-mouth referral. There are very few other links competing for attention, and when a link is provided in a clear context, there is no competition.
The next level in context and competition is search referrals. Searchers have queried a topic and they are evaluating the result page to find the most relevant choice to click. The context is usually fairly good, depending upon the searcher’s terms, and in an ideal world, all of the choices are relevant; however, they are all competing for the searcher’s attention. The more results on the page, the more scrolling, and the more results pages viewed, the more the competition for the searcher’s attention increases and the less chance that your site will be clicked.
Finally, social media. The competition for the reader’s attention is huge, as the homepage of Digg alone has about 30 news and topical links to choose from. Contextually, those links have little to nothing in common. This is the place for distraction, something different, and discovery of news not otherwise found. Most users are not looking for anything in particular, politics excluded, so it is all about what catches their eye. The competition for their attention is very high, and the context is very low for specific subjects, so it logically stands that their engagement rates for any site will be very low.
The biggest take-away for me was the impact of a good link from an online news source. The visitors and the weight of the authority provide a significant benefit. Truly, a show of quality over quantity. Online Public Relations and reaching out to online news outlets is still one of the most viable methods of creating awareness and traffic for a website.
The second take-away is the recognition that a long term strategy that focuses on your target audience will be the most viable and profitable. Chasing after short-term repetition from social news sites as a means of marketing a website will lead to detached visitors who see no consistent context to your site or your goals as a business.
The tried and true focus of building a business by a long-term focus on your target market is borne out in the data. Building a relevant site that connects with visitor needs is the most beneficial and valuable. A strategy that engages visitors by the contextual relevance will build sales and leads and long-term reputation at the same time.