Friday, February 24, 2012

A Universal Index for Measuring Social Engagement

When talking about social media, the conversation usually focuses around two broad strategies. The first is to work in the medium as a pure social channel, where engagement metrics define success. The second is to treat Facebook, for example, like a traditional broadcast medium, where the focus is really on the size of the fan base and brand-sourced content.

While I believe that many brands preach the former but practice the latter, even the most well intentioned organizations face an uphill battle when trying to determine the success and value of their efforts in social.  Social ROI, for example, is too often judged by direct conversions, and I’m not sure this is the right metric (especially as Facebook commerce struggles as a whole). At the very least, attribution reporting, which tracks how various points of contact contribute to a final conversion in any channel, must figure into the picture of social ROI (See the “Assisted Conversion” report in Google Analytics for an example). Either way, it is an important discussion that deserves a deeper dive in a later blog post. For today, let’s set it aside and focus instead on the concept of metrics for social engagement.

There exists a common core flaw in the metrics that are generally used to measure engagement. The flaw is that nearly all reporting I see is based on raw counts of engagement interactions. Examples include the quantity of Likes, Shares, Retweets, etc. These metrics would all be fine if we consistently worked with the same quantity of followers, but we don’t; as adoption of social channels and mobile devices continues to climb, it is a safe bet that even the most dysfunctional brands will continue to attract more followers. In other words, even brands that hardly try will see their fan bases increase and, as a result, their raw engagement counts grow.

Here are two specific challenges inherent in using raw counts for social engagement:
  • Inaccurate performance comparisons between two time periods like year-over-year or month-over-month
  • Relative inability to compare brands or competitors of differently sized fan bases 
In a nutshell, what we need is a universal metric that would give an apples-to-apples comparison for engagement performance, regardless of the size of the fan base.

The good news: I’ve got one that works especially well for Facebook. DMW calls it the Facebook Engagement Index (FEI). It is a simple but powerful metric that allows regional amusement parks to compare themselves to Walt Disney World, for example, and for brands to measure their engagement efforts vs. the same time last year. The concept is to measure engagement by “per fan metrics” instead of raw counts. That is, divide the raw engagement counts by the quantity of followers. What you wind up with is a consistent range of percentages that universally defines social success (or failure) in Facebook.

We have been using FEI with our clients for about two years now, and have recently begun monitoring other brands as well. I usually find that an FEI index of 2% to 4% counts as “average”, with an index of 6% to 10% counting as “very good”. I occasionally see brands reaching an FEI above 10%, but this is uncommon and almost always temporary.

So what can we learn by using the Engagement Index? For starters let’s compare Engagement Index over time against quantity of fans. In the past two years, DMW has learned that nearly all brands will form an “X” chart here, where engagement starts high and fan count low. Unfortunately, the common pattern is that these trends then switch places over time, with engagement diminishing in converse relationship to growth in fanbase. Although it’s something to avoid, here’s a common chart:

The goal is to try and keep that Engagement Index trend nice and level as your fan base goes up. That is, maintain the quality while increasing the quantity. One of our clients, Center Parcs UK, is superb in this category. Here’s a sample chart from them:

See that level blue dotted line? It means that they have grown a fan base that remains consistently engaged despite its size. They have gained Quantity without sacrificing Quality. It should be the goal of all brands to have an Engagement Index chart that looks like this one.

Next, for a brand-to-brand comparison, let’s look at three sneaker brands over the course of the past four weeks....

Notice that Converse (in green) is, by a long shot, the most popular of the three brands with over 2 million fans. But look at the dotted green line, their Engagement Index trend. Floating in the range of 1.5% to 3%, they are just barely getting into the realm of “average”. Summary: Converse has a built broadcast channel and probably has a good one million followers who couldn't care less about the brand.

As a point of comparison, let’s have a look at Adidas (the blue lines). This brand has a healthy but comparatively modest fan base of just half a million, but in terms of engagement they are punching far above their weight. Even discounting for their temporary meteoric rise into the double digits, they are usually around 5% or 6%. In terms of engagement, they are running circles around the other two brands. Great job!

So I ask you: Who is doing better in social? The brand with a huge quantity of disinterested followers or the brand with a modest quantity of engaged fans? There is plenty of room for debate, but my money is on the latter. I believe it easier to grow engaged fan bases over time than to prune dead weight away from an already grown community. What’s your opinion?

If you'd like to learn more about DMW's Facebook Engagement Index, please get in touch via the links on the right or via the comments.

Tuesday, February 7, 2012

New Faces but Old Games for Major Hotel Websites

Waldorf-Astoria, Four Seasons, and Marriott’s AutographCollection have all seen website updates in the past few weeks. All three of them have done a relatively good job in optimizing for tablet users, but some have done better at addressing the thornier issues of integrating social sharing and reviews.  Let’s take a closer look and learn more…

(Note: I am limiting today's review to traditional websites only; a similar review of mobile sites will come soon.)

This site wisely maximizes the impact of its enormous panoramic photographs with hideaway menus for booking (up top) and other details (along the bottom).  This cosmetic change is good for tablet users, but the website appears to be completely devoid of reviews or other guest-generated content. The same appears to hold true for the property-specific websites that link from here. While it would be ideal to see reviews in a variety of locations on the site, one good suggestion would be to include reviews and ratings for the "Experience" venues: golf courses and spas. Hilton would actually be able to take a leadership position with such a move by expanding the review game into specific outlets and amenities. Doing so would provide utility to shoppers, therefore increasing traffic and conversion.

Regarding sharing features, the controls are all tucked away on the property sites and don’t seem to exist on the brand site itself. Browsing guests are hardly encouraged to share. More attention is needed here to drive awareness.

Bottom line: Potential guests require third-party reviews to help validate that the actual experience matches those alluring photos. There is a lot of quality here, and some good thinking, but I cannot imagine this website performing to expectations without review content. Overall, it seems like a great facelift but unresponsive to the significant changes in shopping behavior that have developed in the past few years.

Like Waldorf Astoria, this site does a great job with big-impact photos and unobtrusive design. This site goes a step further than Waldorf-Astoria, however, by including “Reviews at a Glance”, right beneath the fold on the property sites. This widget also includes Facebook and Twitter content and links to those assets. The TripAdvisor content is more like “mini testimonials” than actual “reviews” (there is no score, for example), and there is nothing to indicate that these testimonials are not cherry-picked. Still, it's a step in the right direction.  The integration with the site, versus a generic plugin, works well visually.  

Within the property pages, sharing functionality is a step up, too. Share buttons are less tucked away and for the Accommodations page, at least, the share allows you to select a thumbnail of your choice. This functionality seems, oddly, to be absent from the Photos & Videos page, however (at least for Denver).

A few areas of concern I would have here: The wholesale handoff of traffic from the Reviews widget to the TripAdvisor, Facebook, and Twitter pages seems ill advised. I think this is particularly true given the very limited review information that the widget provides. I do not think that shoppers’ needs for third party reviews will necessarily be met by this widget and that they will therefore still leave the site. I’d also suggest that bringing in reviews from platforms in addition to TripAdvisor would be advisable.

Here's a a nice-to-have: Wouldn't it be great (ie: "useful") to see a “By User Review” tab on the “Find a Hotel or Resort” page?

Bottom Line: A good looking website with good sharing functionality and some of the review info that shoppers need, but not necessarily enough to keep them onsite. In the future, they might find that they can meet shoppers' needs and their own needs (converting traffic to reservations) by using plugins to more completely integrate reviews and social content into the site itself.

What a great looking website. The presentation layer, including its video content, is nearly perfect for high-bandwidth tablet use. This site has a “personalized view” function that should, in time, encourage sharing and enable first steps towards social CRM (SCRM) via Facebook sign in. Individual property pages have “Love” and “Share” buttons at the very bottom of the page, and some properties seem to have a call to action for shoppers to add their photos to the hotel’s Flickr account. 

While the Personalized View feature is a good one, the call to action for sharing on property pages seems only adequate. A better choice would be to have a share button right near the big photos at the top of each page. But on the flip side, all property pages seem to have a Google maps local listing, which will boost optimization in Google search results.

Bottom Line: Sadly, this site also appears to be lacking reviews. Shoppers will go to another site to validate the promises made by the brand, and a certain percentage of them will book other hotels or through an OTA channel. 

And the winner is…
Unfortunately, I’m not prepared to give any of these new sites a “winner” label. They are all physically beautiful, but are lacking in best practices for the new era. A truly winning website in this space should have, at least, the following features:
  • Ratings and reviews from multiple platforms (not just TripAdvisor) located within the site
  • A comprehensive sharing strategy that extends deeply into the website and that includes Google +1
  • If the brand believes its social content to be influential to shoppers, that should be embedded within the site as well.
And I would give bonus points for...
  • Including ratings and reviews in the booking engine pages  (the hotel equivalent of the shopping cart), not just the presentation layer
  • Experimenting with validated reviews, as Starwood have done
  • Any use of the game mechanics that are used by to such excellent effect.  
To be fair, it is hard to innovate for such big ships. In fact, I increasingly suspect that Marriott, Hilton, and others ponied up to participate in specifically so that they would have the opportunity to innovate in an environment free of the excessive baggage of their technology and operations platforms. But, until we see evidence of any of these brands breaking out of this rut, the OTAs and review sites will continue to out-innovate and out-maneuver the brands that they are selling -- great new photography notwithstanding.