Microsites. A Bad Idea Most of the Time.
This morning, TechFlash noted that Drugstore.com is expanding its microsite strategy around product categories. The growing list now includes:
Possibly you can already tell based on my headline that I think this is not the awesomest strategy ever. I could call it short-sighted, misguided, any number of other prefixed words, but perhaps I’ll just tell you why microsites are often not the best strategy to pursue. Before I get to that though, I want to point out that I find microsites a bad idea most of the time. Sometimes they are a great idea. Although the more of them you have, the less likely this is the case.
But, you protest! The TechFlash article said that beauty.com helped drugstore.com post a 20% revenue for Q1 2010. Sounds awesome. Microsites rule! Except that beauty.com is the one domain from the list that isn’t actually a microsite. It’s simply a vanity domain that redirects to the beauty category of drugstore.com. So certainly, by focusing resources and awareness on that category, they’ve likely managed to increase sales, but that increase has nothing to do with microsites.
So what’s wrong with microsites? Let’s tally up the ways.
You lose brand identity and audience engagement
You spend significant corporate energy on positive brand perception and awareness. And then you start over completely from scratch with an entirely new brand. Woo? If you are reaching an entirely different audience and your current brand would be confusing, then you may in fact want to build out a new brand, but that case, you probably won’t be launching a microsite, you’ll launch a full site. In most cases, microsites are subsets of or promotions for the main site, with exactly the same audience. Do you really want to work at building up multiple brand identities? And do you really not want to benefit from the brand building in one category for another related category? (This comes especially important with ecommerce sites, such as those drugstore.com operates. Even today, we don’t want to hand over our credit card information to just any site.)
Brand awareness has a search impact as well. As I note in the searcher behavior chapter of my book, searchers quickly evaluate the search results page to determine which result to click on. Many things go into that evaluation, but certainly brand recognition helps in evaluating credibility and perceived value.
You lose the ability to leverage your audience
Let’s say you launch an awesome site with a fantastic user experience, great products, and unrivaled customer support. For instance, let’s say you’re Zappos. Someone writes up a positive article about you in say, the NY Times. Readers start clicking over to your site. They see you sell running shoes. They just read about how great you are, so they feel confident about purchasing some products from your site. But maybe those same readers also need some clothes to go running in. If you had a separate runningclothes.com microsite, you’ve just missed a great opportunity to reach a targeted and motivated audience.
You confuse people and search engines
Oh, I won’t have that whole NY Times reader problem, you say. I’ll just keep a complete copy of my runningclothes.com content on my main site too! That way, I can reach the audience for my main site as well as get all the additional audience potential of the microsite. Oh really? First, that’s just confusing. If someone becomes accustomed to shopping for athletic clothes on your main site and then clicking over for shoes, but then one day they end up on runningclothes.com and everything looks the same… and yet the shoes are gone — that’s just not the experience you want to give users.
But the problem really comes in when you add search engines to the mix. Which version of the pages do you want them to index — the version on your main site or version on your microsite? Likely you’re going to say the microsite. (Especially if you’ve built the microsite because you think keyword-rich domain names have great search potential — read on for more on that, by the way.) But the search engine is likely to index the version on your main site because that site has been around longer, has more links, and has more overall credibility with the search engines. No problem, you say. You’ll just block those pages with robots.txt. Well, OK. You can do that. But then you lose all search engine value from any of the external links those pages may accumulate. You also lose the search value of the internal link structure. That’s not awesome.
My guess is that drugstore.com has Zyrtec on both its main site and its allergysuperstore.com site. Along with all the user reviews, product details, and directions. Drugstore.com ranks #37 for [zyrtec]. allergysuperstore.com doesn’t rank at all.
You may have to spend substantial additional resources
The microsites run by drugstore.com all use the same template and content management system. So it seems like low engineering overhead to maintain them all. But wait. As you build out the content of both sites, you have to decide which content to put where. And decide how to spend marketing, PR, and advertising resources. When you issue a press release, which site do you talk up? All of them? What if you have 20? And you likely are doing social media. Do you now maintain 20 Facebook pages and 20 Twitter accounts? I’m tired just thinking about it.
And if you’ve built the microsite specifically for an advertising campaign, what happens when the campaign is over? Do you maintain the site? Abandon it? Take it down? This question gets more complicated if the microsite included a social networking element. You’ve gotten your audience engaged, now what do you do with them?
During the 2009 Super Bowl, Jack-in-the-Box aired a commercial that showed Jack getting hit by a bus. They launched the microsite hangintherejack.com as part of the campaign. I’m not sure what happened with the lifecycle of the site, but that domain now redirects to jackinthebox.com, so whatever assets they built up there (both in terms of content and audience) have just been thrown away. (They did better with the Twitter account launched as part of the campaign. That wasn’t campaign-specific and it still being used by “Jack”.)
You cobble your search acquisition efforts
A big part of ranking well in search engines continues to be the strength of the external links to the site. If you maintain multiple sites, then you are diluting that external link value. If five people link to your main site and five people link to your microsite, each site is competing for rankings against the rest of the web with those five links. Instead, you could have one site competing with ten links. Anything that you do for offsite search engine optimization, you have to repeat for each site.
It can be difficult to match promotions to search visiblity
One common case of microsites is when a company launches a new promotion. It seems to make perfect sense to launch a microsite as part of that promotion. You can tie branding to the promo and it can be a lot easier to outsource the development of the site to the agency that is managing the promotion creative than to try to coordinate in-house resources and add a section about the promotion to the main company website.
The trouble comes in when that promotion sparks search interest (which it undoubtedly will). I’ve observed this with the Super Bowl commercials in both 2009 and 2010. In 2009, several sites, including Hyundai and Sobe advertised taglines that had corresponding microsites, but those domains redirected to the main domain. Advertisers expected that viewers would type the URL into a browser address bar, but instead, many people typed the tagline or domain into a search box. Since the domain didn’t actually exist, the advertiser didn’t show up in search results. You can see this, for instance, with Hyundai’s Edit Your Own campaign.
Another problem with launching a microsite at the same time as an ad campaign (even if you don’t redirect the URL) is that you don’t want to launch the site until the ad goes live, but you want the site to be visible in search results as soon as the ad goes live. And unfortunately, you can’t have both. The hangintherejack.com site noted above experienced this issue. It wasn’t indexed in Google until six hours after the commercial aired (and the site launched). For a site to be crawled, indexed, and ranking within six hours of launch seems pretty quick. Unless you’ve just spent millions on a Super Bowl commercial that’s caused the audience to search for the site in Google. You can, of course, mitigate this problem by buying paid search ads. But this blog post isn’t about how to work around microsite issues. It’s about why microsites can be problematic.
But, I can hear you asking. Wouldn’t an advertiser always have this problem, even if they just launched promotion-related content on their main site? Well, yes and no. At the very least, the domain is already known and being actively crawled by the search engines, so you increase your chances of a quick crawl of the new content, particularly if you link it from your home page as soon as it goes live. You can also launch the pages early (without all of the promotion-related content) and ensure the pages include the words that correspond to the queries the promotion will likely trigger, then swap out the content when the ad goes live.
For ad campaign-related web content, you always have to think through the implementation to ensure you leverage search interest, but your options are more limited when you’re dealing with a microsite.
You don’t get the search engine value you think you get
This is the crux of the issue in the case illustrated by drugstore.com. They aren’t launching microsites because they are working with an ad agency on creative for a campaign and it’s too difficult to get internal engineers to add content to their website. And they aren’t building a completely difference business for an entirely new audience. They’re launching entire business verticals for the same audience as their primary site on keyword-rich domains. Why? It can’t be for the type-in traffic. Even those who specialize in the domain business will tell you that type in traffic is on a serious decline. We can see this with the Dockers commercial from the 2010 Super Bowl. A URL was the number two spiking query on Google that day.
People use search engines as primary navigation for the web even when they already know the web address.
Generally, when I work with companies who want to use a bunch of keyword-rich domains, it’s because they think there’s some inherent search engine value in the domains themselves. This assumes that the brilliant PhDs at Google think to themselves: “Huh. This domain is cheaponlinebooks.com. It totally must be the most relevant result for [cheap online books] queries. After all, the words are right in the domain name!” However, as it turns out, this has been a technique used by spammers since the days of stone tablets and chisels. Or, OK. Since at least 1995. The search engines are onto it. (Well, maybe not Bing quite yet.)
There is so much super valuable content on domains that aren’t keyword rich and there is so much spammy, crappy content on keyword-rich domains that Google just doesn’t find it useful as a relevance signal.
Keywords can indirectly help when they’re in the URL because you’ll get anchor text credit for any URL-only links. But that really has nothing to do with the domain, so why not just use keyword-rich URLs on your main domain and get those benefits without incurring all of the drawbacks of microsites?
People also sometimes think operating multiple domains will help search engine rankings in other ways, such as that you can link to yourself for instant PageRank credit! Or that you can dominate the search results with all those domains. I hate to be the one to break the news, but search engines are on to those things too. Over time, search engines generally can figure out when sites are part of an owned network and then treat them accordingly (which is similar to how they would treat the content if it were all part of one site). And if you now want to ask how do they know so you can figure out how to hide it, then you’re getting dangerously close to thinking about search engine manipulation. Maybe you should read this and then come back.
Certainly, you’ll find many out there who swear up and down that having keywords in the domain makes a big difference. I think mostly this isn’t the case. That any examples of keyword-rich domain names ranking well are also a case of the content on those domains actually being the most relevant result for a set of queries. Even if it did work, it would presumably only work for exact match, so you’d need a lot of domains to make up to set of queries you really want to rank for. That sounds exhausting. I also think that this is the kind of ranking signal that’s likely being tweaked all the time, and even if it works for a time, it’s a poor foundation for a long term business strategy.
But just as importantly, once you start focusing one building your business based on perceived signals in the search engine algorithms, you’ve lost sight of why you’re building the business in the first place and of your customers and while this may seem like a minor diversion, it may take you down a completely different path than the one that’s based on building substantial user value.
Suddenly you’ve got a set of spam tactics rather than a business model.
So do keyword-rich domains have any value? Maybe. If you are starting a brand new site and can pick any domain name you want, in some cases it may make sense to go with a keyword-rich one. It will be memorable, easy to type, and will encourage useful anchor text. It might also encourage click through for URL-only links as it may be more obvious what the site is about. And if you have or can acquire a bunch of keyword-rich domains related to your industry, you may as well redirect them to your main site to capture any type-in traffic they happen to get (although don’t expect any SEO benefit from this).
But launching a whole bunch of keyword-rich microsites related to your industry in hopes that you’ll get all those microsites ranking separately for variations of query? Probably not the awesomest idea ever.
Updated 10/9/11: I’m not sure when this change happened, but most of these domains are now redirecting into a section of drugstore.com.