You know how Eskimos supposedly have the highest number words that mean “snow” than any other language. OK, maybe that’s an urban legend, but as an SEO, how would you optimize a site for such redundancy? If you’re working on an eCommerce site, you probably feel like you have that problem all the time.
“I have 30 collection pages here that would be a good landing page for shirts!!! What do I pick? Where do I start?”
Do you decide to try to get all the pages to rank for shirts, and hope that one comes up well in the rankings? Do you pray to the duplicate content filters gods and just wait and see what ranks? Do you beg the catalog managers to change the collections (which is probably an uphill battle that could have a dangerous effect on usability and sales). Or do you get more proactive in your SEO efforts?
Having 30 pages that speak to the same thing gives you a lot of opportunity, after you trim the fat. How are the duplicate pages created? Are they based of parametric filters? Sorting? Internal search queries? Which pages get the lion’s share of the traffic? What have the best conversion rate or least amount of bounce? Which are really the important pages in your pack? Determine the fat and the lean; make this step one.
Once you figure this out, you should work to close these paths. Granted, on a dynamic platform you may not have the ability to add per-page meta robots, nofollows, or a tightly tuned canonical tag. What about Webmaster Tools, and their cool parameter-blocking feature? It might be an option.
Let’s be honest – in enterprise SEO, it’s really not likely that you’ll be able to tighten the site to your exact specifications, no matter how big (and hard working) your team is. In enterprise SEO, it’s about locating the biggest holes and plugging them first. It might be more of the 80/20 rule, or a matter of specific initiatives to tie with other channels. Just don’t expect to eat the whole pie.
So back to our opportunity. Once you’ve cut the fat and done your best to get it off your plate (which may require you to monitor Google to see if their internal duplicate content consolidation is working), you get to have fun. Let’s say you’ve been able to identify five pages that would be good candidates for optimizing with the term “shirt.” Let’s go a step further and pretend they’re the same types of shirts – tee shirts. Do your keyword research. Grab up the best terms. Compare your current placement for each term, against your competitors. Look at PPC data (helps you understand demand and opportunity). Look at each page’s back link portfolio (or hit up Linkscape) to get a sense of what page is likely already juiced up.
Pretty soon you’ll find that for your five pages, you have a handful of similar terms that need a home: tee shirts, printed tee shirts, cotton tee shirts, etc. That’s right, folks. You’re probably going to be playing with the long-tail… the SEO devil’s playground. Expect to keep notes while you plug some content in, and fuel with backlinks. Wait a few weeks (or months depending on how strong your site is), and measure. Was there MOM/YOY growth (a lot of eCommerce is seasonal)?
The goal here isn’t just to avoid cannibalizing terms, but not to cannibalize themes as well. You really want to theme these five pages out. For duplicate page A, start theming it about printed tees. The cool style, what people like about them, the variety in full body prints, etc. For duplicate page B, start theming out the cotton blends and how they’re durable. Get creative, and start dabbing your uniqueness all over each canvas, using different colors.
Once your done, sit back and watch. Start measuring. Keep a log. You put in a lot of work that hopefully benefits depending on how well your decisions were made at step one. Maybe you find it didn’t pay off as well as you’d like, but I assure you, after doing eCommerce SEO for 10 years, it takes practice getting your methods down. And because ROI is your challenge, you might find it’s also you’re best friend since it’ll keep you happily employed.
Google, always trying to push social search (despite early missteps), has a good one here. The new +1 button, rolling out now (when you’re logged into your Google account and choose to opt into the experiment) lets searches know how popular a listing is, ala StumbleUpon’s plugin.
After speaking with Google, this is not yet affecting rankings or quality score, but they said it will. Straight from a product manager’s mouth. He couldn’t tell me when – they need to evaluate, I’m sure.
But this is a useful thing to reinforcing CTR on a brand, not to mention owning more search engine result page real estate. Thanks to social additions, my listing now pushes other sites below the fold even more than before. If I own a few top listings, I can ultimately “black out,” or push a fourth or fifth under the fold:
Is it like Facebook’s Like button? Definitely, especially since soon you’ll be able to add a +1 widget to your websites. But this has something Facebook doesn’t have… the ability to “like” something in Google results. That’s pretty big.
More info here.
Updated: Feb 23 2011
Customers still rely heavily on search engines to find web-based mobile sites. It’s not unlike traditional SEO in many technical ways (Google still cares about the keywords and the links), but is very different when optimizing for user or customer value. To optimize for search engines on behalf of the mobile user or customer, you have to think about what the mobile searcher is looking for when searching on a phone. The answer: relevancy, speed, and good usability. Identify your landing pages that are best suited for them and think about how you can optimize for the phone. We’re trying to attract mobile users in addition to desktop/laptop users. But mobile users have a larger sense of urgency.
Phones are not used like desktops and laptops. They’re not even like iPads. Customers on the mobile are on the move. Assume they’re short on time; they may quickly be approaching a bus stop, or walking into a store. Maybe the light just turned green (scary but true) and they need to get back to driving. When we optimize for a mobile page, we need to identify and provide the key answer in the title, meta description, and body copy with as few words (and keywords) as possible. We need to be much more concise and specific so the mobile user can identify the best results faster. We need to spend closer attention to the query intent. If that means more specific mobile landing pages (and less general, high-keyword frequency pages), so be it. Granted, that goes against some traditional SEO strategies. From the little data that’s been revealed from Google about differences in the way they approach mobile sites, it’s our best hypothesis that they’ll continue reevaluating your keyword choice from a mobile perspective. You already get personalized, GPS powered mobile results from Google sniffing your smart phone browser now, so this isn’t really a stretch.
The mobile searcher is likely searching for a quick one sentence answer. Or a price. Or a location. Or a quick review. Microformats and location tagging will likely take a larger role. Mobile users don’t want to zoom in/out of a page all the time (if their phone even enables it); they’ll often back out and view other Google results for the best visual snapshot (even if it’s not the most relevant page to the query). Usability plays a different, but equally as important role as it does now. In general, if our goal as SEOs are about driving qualified traffic from the query all the way to the shopping cart, sometimes we need to be focused on design and usability.
Old school technical SEO still needs to be a factor. Most developers create a different URL for mobile sites when it’s not necessary. I see the “m.” subdomain used. If you share your mobile link through an online social channel, you’re sharing the m. version. If your logic properly redirects a user through that link to your desktop version, you’re still being served a redirect. Some loss in link juice there even if its a 301. At least use an /m/ directory and turn off internal linking user agent switching so you can get some links that help your overall domain authority. Currently Google has their normal Googlebot, and Googlebot-Mobile which crawls content for traditional phones – not smart phones (with the exception of a recent iPhone Googlebot that’s been testing). Google believes that smart phones can see the web just fine and doesn’t need their own bot. If that’s the case, there really isn’t much reason to create a new URL anyway if the content you want a person to see on the phone is the same as the content on the desktop. Just create a different CSS sheet to create a more mobile layout.
Mobile will only continue to grow. Additionally, more iPad-like tablets are slated to come out, which blurs the lines a little more between what is a mobile device and what is a desktop device. Google will continue to take the non-desktop search and web experience seriously. So should we.
The major web platforms are looking at targeting users with local functionality. Many see this as a major growth opportunity in 2011 due to the higher use of smart phones. Google is especially focused in this area as of late, arguably more than ever before. As online retailers, who may not have heavy connectivity with their brick and mortar counterparts, local SEO may not seem like something that provides much – if any – online traffic. But it does. Especially with recent Google changes.
If you haven’t noticed, Google changed the way they display their local searches. They appear to show up more often, and resemble traditional natural search listings. The result is that other non-local listings are getting pushed down under the fold, and more local listings are being clicked.
Each listing provides 2 destination links: the main link (which leads to your main site), and a places page.
When you show up in the local searches, the main link provides pretty good traffic. In most cases, the searchers that click a local link were looking for local information. The destination of this link doesn’t satisfy, but it’s a chance for your homepage to capture the users interest and maybe persuade them from getting off their couch and driving to the store to buying online.
The other link, Places (formerly called Local Business Center), is a nice thing to have because it provides opportunity to really sell your local store. You can provide an exclusive coupon, or promotion. Within the Places page, there’s yet another link that you can control.
There’s opportunity with this link since you control it. Maybe design a landing page displaying synergies between your web and brick and mortar stores. Can you buy online and return the product in the store if you’re not satisfied? Promote that here. Do you have exclusive in-store printable coupons? Display that here. Experiment with this traffic, and develop something special knowing that these are local-minded shoppers (at least they were at the time of entering their first query into Google.
I was recently asked in an email why I consider SEO a marketing channel. Among several things, good marketing and advertising work to get messages out about the value of an item, and provide you with information. Most subscribe to this definition. Marketing helps those who are interested see if they really want and need it, and helps inform producers.
So does Google.
SEO helps those people who have interest, and are qualified enough to make a digital inquirey, find this information. SEO also helps create that two way, open engagement that more people are expecting of the maturing internet.
I work with a lot of huge brands, typically in the ecommerce space. It all holds very true for them. Doing SEO work is about caring for the customer more than the product. Hopefully the product was made with an audience segment in mind; SEO is bridging the gap using the internet’s elected hub – Google.
Yes. It’s textbook marketing taking you back to college. But it’s breathing on land now, and doesn’t require gills. The nervous system hasn’t changed. The song remains the same.
Once you make the site technically crawlable and findable, you need to make it work. Sure, you can pass it off to merchandisers or usability or any other group that should have an interest in what to do with the search traffic you deliver, but they won’t know what brought them there like an SEO will.
As far as I’m concerned, marketing is part of the broader definition of SEO in the modern age, still keeping it your most powerful acquisition channel by far… If done right.
According to Hitwise, 81% of the searches done on Bing and Yahoo resulted in an actual visit to a website. Google only showed a 65% rate. This suggests that either Bing/Yahoo is more relevant and providing the best results more often for the bulk of users, or that people search differently with Google. I’m assuming the latter.
I think most people who use Google expect to do a little digging. I think the results you’re given require you to refine your search, and as a Google user, you’re used to that. You’ve come to expect that.
Andy Beal at Marketing Pilgrim says, “Google offers more opportunities right upfront to refine the search by time, type of result, even result loca tions. Because of this, I’d bet many people take a second or third try at finding exactly what they want before they start clicking through.” That makes sense. I also think without the options, google users would be more apt to refining their search anyway.
I believe Google’s results are more detailed in nature and require your queries to be more specific as well. I feel like I get broader, safer results out of Bing. Thats what theyre going for per their marketing, but it feels a little “Fisher Price” to me. Not my style. Maybe Bing users are more casual.
Google and Bing have segregated the search audience. Like democrat and republican, NFL and MLB, or beer and wine, the two parties are different, and will continue to be shaped by the structure of the engine to some degree. It’s interesting, really, just how big a role search engines play, and what we can tell about people who use them. It’s not just an information retrieval system, but an extension of your brain. Much like a car.
Usually when we learn from examples, we learn from someone’s success. Sometimes it’s good to look at someone who did it wrong. The shortest job stint in my life was for one year, but despite my unhappiness, I did learn some good primary lessons that are still effective today.
Let me introduce you to Slim (name changed to protect the careless).
Slim wasn’t much of a businessman. He started out as a blue collar type with a hobby collecting a certain kind of collectible. As the market swung, Slim’s hobby started turning into a passion for a lot of other people in the area. Slim began supplying merchandise related to this hobby at local shops. He was soon able to open his own store.
As the internet and ecommerce grew, interested searchers started using Google to find retailers who were selling these collectibles online. The small town shoppers who loved the brick and mortar store weren’t the only audience Slim could reach. To his credit, he partnered his physical store with an ecommerce store. Opportunity abound!
I worked for Slim doing SEO. His pay rate was insulting, but because he was becoming semi-popular in his genre, and I was able to negotiate a small commission on sales, I thought it was worth a shot. I should have recognized the cheapness as a sign of things to come.
I found a new feature on Google. If you click advanced search on the search page, you have a ‘reading level’ option. In the drop down, you can choose to annotate your search results. When you search, you can now see the reading level Google thinks these (and your) pages are at.
As a father of a six year old, this seems interesting. It could potentially help me find pages that he would understand and enjoy. Unfortunately it doesn’t seem to work as I had hoped. Apparently my about me page is at an intermediate reading level. Hardly. So is sesamestreet.com. Wow.
Maybe this is just more Google fluff, and maybe it will improve. But it has me wondering about the signals and algorithm that determines this labeling. Are there any clues here for the reverse engineers to understand more how Google thinks? Likely, Google would be very careful putting this out… but still.
Update: Looks like Google gave us a little insight. Looks like a model was built off decisions made by teachers. I’m now thinking this is simply another algorithm strand layered into the Google rope.
Hey all – my new article is up at Search Engine Land. Enjoy.