Ecommerce lives and dies on its conversion rate. That's true for all channels, not just SEO. Get the three key areas covered here right, and you'll be able to conquer those new territories while not just protecting, but actually improving your main territory performance.
What does an expert big-site Ecommerce SEO do to expand into new countries without harming existing rankings?
Let's assume we have to handle expanding into Canada, U.S., Germany, and France. It's your job to ensure the new country performance is just as good, and that performance in the current country isn't affected during the roll out.
Design & Build to Succeed in Foreign Languages & Territories
Domain architecture is the first port of call, because you'll need to get your changes in the technical queue ASAP.
First up, we want to harvest up all the country TLDs for the brand if we don't have them already. We will be using them, but not to perform in the SERPs or host content of any kind, instead we'll use the following structure.
Taking Canada as an example territory, we'd use: ca.bigdomain.com/en/ & ca.bigdomain.com/fr/, and so on for all relevant language permutations, with the root subdomain verified in Google Webmaster Tools as local to the country to be targeted.
Why keep everything on the .com TLD? So we can perform out of the box on the top level generic terms that are critical to the success of our campaign, yet still target multiple listings thanks to Google allowing subdomains to perform in the SERPs independently.
If we've got best practice eCommerce URLs in place already, then it's a simple case of translating our existing, non-duplicating architecture to identify all the key pages we'll need for each territory:
- ca.bigsite.com/fr/ (French language homepage, Canada)
- ca.bigsite.com/fr/translated-full-product-name-matching-title/
- ca.bigsite.com/fr/translated-top-level-category-page/
- ca.bigsite.com/fr/translated-top-level-category-page/sub-category/
- ca.bigsite.com/fr/translated-top-level-category-page/offers/
Because we're localised in Google's Webmaster Tools, if we declare our languages as an HTML attribute and in the head, we'll ensure the right page is performing in the right country for the right language. Keeping control of the landing pages performing for each term is critical to achieving success in new territories.
Oh, and those country-level TLDs? We've set them up as vanity URLs to 301 to the appropriate subdomain pages, so we can run offline campaigns through them, where the shorter the total URL the better. By making the 301 preserve any additional query parameters or directory info, we can also segment out the performance of those offline campaigns in our main site analytics.
In our .htaccess on the vanity domain we'd use something like:
RewriteCond %{QUERY_STRING} ^(([^&]*&)*)id=radio&*([^&].*)?$
RewriteRule ^/$ http://ca.bigsite.com/fr/radio?%1%3 [L,R=301]
Avoid Content Translation Penalties in Google
Looking at our four new target territories (Canada, U.S., Germany & France), two would use essentially the same content developed for the current English-language market. If we're also after the English speaking communities in Germany and France (often the case for high end products targeting Ex-Pats), then we have 4x pages for every live page on the site requiring essentially the same content.
Your first thought would be that if Google allows site owners to localize parts of a domain to different territories, then it also wouldn't treat identical content intended to return in each territory as duplicate content. Right?
After all, what value does requiring sites to generate different content for a product page in England, America, New Zealand, Canada, the U.S. and Australia provide for searchers?
You'd be wrong.
Matt Cutts (and Google's) formal stance on the topic is particularly unhelpful. I'd also hope that Cutts isn't deliberately being disingenuous by suggesting that local country TLDs are best: think of all the additional linkbuilding that would require to get them all performing, and don't get me started on tipping the threshold to escape the sandbox (yes, that does still exist) and perform on transactional search terms.
So what do we know?
Don't auto translate. Matt states in the video that you'll be treated as if you've auto-generated content, and you'll incur a filter preventing SERP performance at best, and trip a +30 or +60 penalty at worst.
In short: we need unique content for all languages in each territory, regardless of language overlap.
This is in fact a good thing, as you can also brief your translators to add unique value to each page, feed back on native searcher habits, and proof content to improve conversion rates.
At the very least you should use this opportunity to address obvious issues like international delivery costs, support quality and location, guarantee validity, etc.
Once you've launched, keep asking your visitors if you're doing it right. Use User Voice surveys like those available for free from 4Q or make your own with SurveyMonkey.
Make Sure You Can Get Paid
To be more precise: make sure you can take payment natively.
Your ecomm payment system needs to take payments in the local currency, using the most common local payment cards & processes. Cross territory payment systems like PayPal and Google Checkout are good, but are still uncommon forms of payment, especially outside of the U.S. You need to handle the right credit and debit cards for each location or your checkout conversion funnel will break down.
I've lost track of the number of European payment systems that won't handle U.S. dollars, or vice-versa, and the plethora of cards (or card alternatives) worldwide is a major challenge.
In Germany, for example, only a small proportion will use a credit card online, with many (40 percent) instead favouring ELV (in effect, a direct debit transaction). The UK has electron cards to make you life a little more difficult, while 2.5 million people in Ireland have a laser card.
In Singapore, 3.2 million users want to use real time bank transfers, while the Italians are remarkably keen (24 percent payment process marketshare) on offline bank transfers to complete online payments.
Each system has its own security checks to overcome, and will likely require third-party authentication from the likes of WorldPay which can severely disrupt the payment process and requires a high level of security in passing transaction data back and forth, so budget to test different options and check to ensure your bank's merchant account can handle future planned international rollouts.
Finally, make sure you clearly populate the right currency options through each territory's pages.
Make sure you don't use just 'Dollars' next to payment in a site that operates in both Australia and the U.S. It's not too much to confirm that you mean Australian dollars on your Australian pages, and U.S. dollars on U.S. product pages, rather than making the customer get all the way through the shopping cart before they discover that the item they're buying actually costs 5 percent more (and, in really poor cases also includes a large shipping fee as a supplementary surprise on checkout).
Source: http://searchenginewatch.com/article/2073589/Winning-Multinational-eCommerce-SEO-Strategies
No comments:
Post a Comment