If you’ve been following Get Elastic for a while, you may recall an interview with Michael Bonfils on international SEO strategies, the topic of his panel as SES Toronto last June. This year, the search marketing conference has lent me another fabulous SEO expert to interview, Laura Callow from Intuit Canada. Laura will be speaking on the topic of launching a global website at this year’s SES Toronto, and has kindly answered some burning questions about international SEO for Get Elastic.
Linda: What are the top SEO concerns for global ecommerce websites?
Laura: There are a number of challenges facing global ecommerce sites, but 3 high-level issues that seem to be universal include communication, testing, and clear planning and reporting structures and processes.
SEO teams need to ensure effective communication of data, learnings and insights from group to group and from country to country while remaining focused on their own unique primary vision and objectives. Creating a community of practice or a centre of excellence that facilitates this type of communication is crucial. It enables knowledge sharing and minimizes unnecessarily repetitive training, learning and development cycles.
Hypothesis-based organic testing can easily get lost in the marketing-mix due to unclear expectations, and because of less than robust hypotheses. In metrics driven organizations, if business units expect to achieve significant results with SEO based on-page or off-page initiatives, and if they hope that those changes will help them gain additional budget in the next fiscal year, it is important that they propose, monitor and report on those changes with provable, testable hypotheses followed by very clear data-driven outcomes.
Planning and Reporting:
It’s important to have develop a very clear, linear plan-on-a-page that is presented and approved during fiscal planning. This plan should reference and highlight the proposed SEO roadmap for the next fiscal period.
Every plan should start with a vision, followed by clear high-level objectives (or buckets) and include top strategies (or key incremental levers to pull) to achieve those objectives. Tactics are more fluid and can be assigned later. What management needs is a clear summary of what each SEO team/unit plans to achieve in terms of outcomes, and how those outcomes align to both SEO objectives as well as to the overall business unit/company objectives.
Maintaining a robust plan-on-a-page template for all teams globally makes it much easier for management to understand the proposed plan/s and the assigned outcomes and to compare them across geographies. Reporting should follow a similar global template based system, highlighting overarching business based KPIs, followed by team driven KPIs which formalize the clear outcomes expected of unique divisional/geographic objectives.
Linda: A big decision for any webmaster is whether to go with one site for all countries or multiple, localized sites. From an SEO perspective, is one approach "better" than the other?
Laura: The first thing to consider concerns your resourcing limitations. Many global sites use CMS (content management systems) and are template based. If you are planning on targeting Saudi Arabia and the UK for example, you will need different templates due to user behavior, so different localized sites makes sense. (The Canadian, UK and US English speaking online populations also display different user behavior and expectations).
Language and cultural differences suggest that localizing your presence will be to your benefit in terms of appealing to the four elements of Shackel’s Acceptability Paradigm - usability, utility, likability and speed - especially across language and cultural barriers. SEO is to a great extent about usability and likability, therefore country specific content with relevant colloquialisms, languages and user-driven design makes sense.
There is also the fact that we know that in terms of country-based SEO rankings (Google cited), the most effective way to rank in-country is to have the relevant ccTLD (country code top level domain). Other ways effective rankings can be achieved are via hosting in-country regardless of ccTLD, and/or by geo-targeting in Google Webmaster Tools – which can be done at a subdomain level or at a folder level. It is also worth noting that in many cases a relevant ccTLD is more likely to receive friendly link-juice from other in-country sites due to the perception of affiliation and/or nationalism, and any SEO will tell you that backlinks are the life-blood of SEO along with unique, relevant, quality content.
Linda: For a site with, say, 15 localized country sites, would you recommend each site have its own SEO team, or can efforts be centralized?
Laura: Different countries, even those with the same official language, can have very different culturally driven user habits, expectations and behaviours. It is always possible to achieve economies of scale by sharing data and winning test versions or copy, but those should be tested in-market without preconceived assumptions. SEO starts with - and is significantly impacted by – keyword research and an understanding of the local competitive and search behavior landscape. Marginalizing those impacts by using a single SEO team can marginalize your returns.
Linda: We know nothing can substitute for a local speaker when writing copy, but that can be expensive. What are some ways to get high quality translation/copywriting at a reasonable price?
Laura: If you’re going global and expect to achieve global returns, you have to speak the language of your users on a country-by-country basis. In fact, in many instances it comes down to dialects. You can go small and use free or minimal cost translation services and you may perhaps get lucky. It is often advisable to start smaller in your net gain expectations to get it right first (or fail fast) and refine, than to assume ‘they will get it’. That mentality smacks of ‘build-it-and-they-will-come’. Your brand may be big, but if you don’t speak the language of your users and understand their unique needs and wants, it may be a very costly mistake.
Run a simple A/B test when you launch if your choice is to go ‘minimal translation cost.’ Make your ‘Control’ an existing page (with potentially/proposed significant revenue) and make your ‘Recipe 1’ the exact same page but with high-end translation. See the results for yourself. It varies by brand, by industry, by market and by geo-target. It may be there is no need to expend the extra dollars. It may be that there is a significant gain to be made long-term by investing up-front to meet your user’s needs and expectations. Don’t minimize your potential revenue by assuming no/low return on real translation. Test it!
Linda: What is the number one SEO mistake you see with global ecommerce sites?
Laura: It’s hard to say what the number 1 mistake is, but I’ll include 3 here that make up a big majority of potential misses and mistakes.
The first is unrealistic expectations due to bringing SEO in reactively. SEO inclusion, application and integration should be elemental, incremental and up front from the research and IA/UxD (information architecture, user experience design) stage and should include effective keyword research, partnering with usability from the initial planning stages.
The second involves assuming that global sites will auto-generate backlinks. While big brands and global sites do have the propensity to seem to minimize the need for backlink acquisition strategies, the backlinks acquired may not be to country specific sites. If they are, they are likely to include nearly entirely brand-based anchor text and point in the majority to the country-specific home/index page.
The third involves analytics and reporting. SEO doesn’t end when the site goes live. Tactical refinement and adjustment based on analytics analysis is vital to keep the dollars flowing and to maximize the bottom-line impact of SEO initiatives.