Bloomberg recently published an article discussing household name brands that have shut down their Facebook stores due to under-performance. Retailers at Gamestop, Gap, J.C. Penney, and Nordstrom have all launched, and since abandoned, initiatives to sell through Facebook “tabs” as an e-commerce channel.

I can’t say I’m surprised by this data. The “F-commerce” model goes against a marketing concept that I believe in strongly: the “hub and spoke” model.

In this Hub and Spoke model, destinations like Facebook are but one of many “spokes” where online marketers need to be in order to influence, persuade, and satisfy prospects. A common set of spokes would be e-mail marketing, social media sites, mobile, display advertising, PPC, SEO, TV, etc.

The “hub” is the core website, e.g. www.YourBrand.com. All offers and calls to action occurring in the spokes drive traffic to the hub, and the hub is designed to service or convert all prospects or customers, regardless of which spoke sent them in.
Assuming you agree that a Hub-and-Spoke Model makes the most sense for an online retailer, the real question becomes: How do I create the greatest “Hub” the shoppers have ever seen?

I’d like to finish this post by listing a few pseudo-philosophical ideas on how this can be accomplished over time:

1. Segment web analytics by “spoke”: If your website is the hub of your multichannel marketing efforts, you should RARELY be analyzing web analytics data without first segmenting by traffic source. Your traffic sources = your spokes. If a spoke in your model exists that you can’t segment on, you should be looking for ways to solve that problem. For example, you might say that tracking your TV ad spoke is impossible. This isn’t true if you employ vanity URLs, redirects, and campaign tracking parameters. Segment on those tracking parameters and BOOM! you’ve got a slice of site performance data specific to traffic driven into the hub by TV ads.
2. Hang out in the spokes to understand your audience: If you set up your various spokes of, let’s say, YouTube, radio, and PPC, you need to be investing time in “hanging out” with the prospects who you’re reaching via those channels. Who is commenting on your videos on social media sites? What are they saying? What lingo do they use? What is their sentiment? You get the picture. The data you collect listening out in the spokes should inform your optimization efforts at the Hub.
3. Test and optimize by traffic source: Optimization efforts like A/B and Multivariate testing are en vogue right now, but if you’re running tests on your aggregate traffic (all the spokes at once) you may not get the best possible results. If you have enough traffic coming in from your spokes, restrict the experiment group to only those entering from a specific source/spoke. You’ll know more about them via segmented analytics and listening efforts, have a more specific hypothesis, design a better test, and see more radical results.
4. Optimize from the Hub out: Based on what you learn in hub optimization testing about what messages, offers, or designs work best for what marketing channels, use those learnings to go out and tweak your marketing touch-points out in the spokes. For example, all the tests where you land the user in the shopping cart with the offered product loaded in bombed. So, now you go out to the spoke with a slightly different call to action, and land the user on a product detail page or a category page.

Do the four things above, and I believe you’ll have a very effective hub-and-spoke marketing architecture that can be continuously improved.

What about you? Do you think f-commerce stands a chance? Do you subscribe to the hub-and-spoke concept? Any anecdotes about marketing optimization in the hub and spoke model? Let me and my readers know.