It’s hard to report back from a conference with an overview: if you weren’t at the conference, unless there was a huge buzz, few people are going to spend time trying to catch up after the event. This, therefore, could be the post that no-one reads. Which is a shame, because International search is difficult. There are few specialists out there, so best guess is even harder than with genuine ‘local’ search. International search is broad church, the lessons from one site may not apply to all, the number of approaches is huge, and the range of budgets is vast. An already small – international- conversation is further fragmented by the fact that Google’s algorithm rolls out in different countries at different speeds. It would be helpful, in my opinion, for us to start differentiating between types of sites. We need a new lexicon to describe and differentiate between the types of sites that we manage. There is a vast difference between the e-commerce brand that can immediately generate an ROI in sales from its website and the one which is there for lead generation. There’s a huge difference between a FTSE 100 company selling worldwide with a huge budget for lots of sites and external support, and the site for the owner manager who’s the international expert in their field, needing to be found on a single site. So here’s the rub. Best practise in international search is to have separate sites and work locally on things like links, social signals, servers etc. But frankly, the budget for this for most companies just isn’t there. As Guy Redmond (see below) pointed out, to reach 90% of the world Internet users you need at least 21 languages. Even within the language spectrum there are differences in usage in different countries. So frankly, international search is always going to be a compromise. (I’d love to see more companies taking an interest in international search and more specialists evolving.) The biggest message was use your data and learn from it. But that wouldn’t make much of a story, so here’s six of the best from the International Search Summit in London 2012:
Number one: Test, Measure, Rinse and Repeat
The importance of finding and analysing data was a constant theme, but it was Guy Redmond of an unspecified FTSE company who articulated the process – test, measure, rinse, repeat. As Bas van den Beld, State of Search, pointed out, people’s behaviour varies with their culture. A global site may be a local site, but examine how people are behaving in your target countries and learn from the differences. Things roll out at different speeds in different countries. By keeping an eye on the data/what’s happening, you may well be able to predict what’s coming down the line, or when a particular tactic is about to lose favour with a particular search engine. For example, Google tends to roll out in the US first. Old SEO tactics may well still work in some other countries, but fail to pre-empt the changes at your peril.
Number two: lean on people
Overall the message was to lean on people. Get your data, and then ask the questions about why. (Hat tip to Jack Porter Smith of Peach Digital for this one.) For example, get users to feedback how easy your site is to use for visitors coming from another country, and use as many people as you know, or can cheaply engage, to give you help with translations. If you work for an agency, be honest with the client and ask them for help with messages in foreign languages, translations and key terms. In different cultures, you’ll see different behaviours. Don’t stress. Learn from the data, research what it might mean, and act accordingly. I learned across the day that Russians, for example, are sceptical about ads, and deal a lot in cash on delivery. The Japanese are sceptical about ads. And PayPal is preferred in some countries. So the fact that no-one is buying in Russia is as likely to be about your sales/service structure as about your product not being wanted there. Lean on people for information!
Number three: Devices are different.
Astoundingly obvious, rarely articulated: Stanislas di Vittorio of eSearch Vision pointed out that the customer journey will always be different according to device. In complex sales visitors may come often, and therefore visit from multiple devices. The mobile user, for example, may well convert offline. So in measuring conversions, we need to try and understand what the consumer is doing. And this may well be different in different countries, even from similar devices.
Number four: Get over Google
Most of this thinking should be attributed to Guy Redmond. It’s not all about Google. In the USA, Japan and others, Yahoo and Bing matter. Yandex is one to watch. Baidu needs local languages and local links. Google’s guidelines matter. Local laws matter. A site which hasn’t adhered to the cookie laws in the Netherlands, for example, won’t be served, according to Bas Van den Beld. See what’s happening and act/react accordingly.
Number five: Get over your domain issues
Should you have a folder structure, a sub domain, or cctlds? SAP, IBM, Nike, and Apple each use different approaches, successfully. Having listened to the discussions, I pretty much concluded that it makes little difference, as long as the search engines have a clear indication of where to go.
Number six: Send signals
To get a page served in a particular country, you need local, relevant content on your pages. And links back to your page at a local level. Mark the pages up: give the search engines a hand. What clues can you offer? The list I assembled across the course of the conference included: URLs; local addresses; geotagging; inbound links from the country the page needs serving to; server location; site maps; href lang tags and canonicals; language; currency and address formats And, as Bas van den Beld pointed out, one of the clearest signals you can get is social. If people are talking about you locally. Seek out those local advocates and get them talking.
And to finish
There are two presentations, both excellent, that I haven’t focused on in the above. Shahid Awan of Cheapflights gave a presentation on managing multi-national agencies – really interesting for anyone managing agencies – from which I concluded: use agencies for their own particular strengths; be slightly disloyal and use someone else for the things they’re stronger at; but come back when you need those strengths again, learning all the while. But I’ll finish with Marty Weintraub of aimClear. If you’ve read Wind in the Willows, Weintraub reminded me of Toad of Toad Hall – gloriously relishing a new toy, just because he could. Boundlessly enthusiastic, seeing no harm in playing with the latest greatest toy, his undoubtedly black hat advice could almost certainly land you, like Mr Toad, in jail (Google jail at least) if not handled with a great deal of respect, care and intelligence. His presentation in multilingual and multinational reputation monitoring reinforced the ‘use the data’ message, and he offered up a live demonstration for acquiring data that was both stunningly simple and deliciously naughty, breaking the terms of service on just about everything, scraping, filtering and firehosing. I’m not going to repeat the advice here: for sheer daring alone, if you get the opportunity to hear Weintraub speak, jump at it. Poot, poot!