Author: Chris Jones, Independent Multichannel Retail Consultant
If you are intending to take your ecommerce proposition international, a key element of your planning and budgeting is likely to be marketing spend. Traditionally, when entering a new market, you might have expected to need to spend a higher-than-back-home percentage of forecast sales ‘building your presence’.
In practice, of course, in the online environment, a good proportion of marketing spend is very direct: it’s pay per click. It isn’t indirectly ‘building presence’ it’s directly ‘building sales’.
So, if you are indeed at that budgeting stage, which countries are going to bring the best return on pay-per-click investment? Comparing pools of keywords isn’t useable as a methodology, because of course it inevitably has language bias. It’s no good comparing the cost of the term ‘men’s shirts’ in the US with the cost of the same term in France, because French customers search in French not in English. From personal experience of travel in more than 90 countries, the only universally used English words seem to be “Hello Mister”, “OK”, “Manchester United” and “F*** Off”, which aren’t really a terribly useful sample for trying to benchmark retail PPC costs.
One possible solution to this challenge is to try considering brand names which are so global that they might reasonably be used to compare across languages and countries (we’ll look at a couple of other techniques in later articles in this brief series). Accordingly, I’ve taken a set of 30 global brand names, such as Nike, Samsung or Lego, and used Google’s Keyword Planning Tool to attempt a comparison of SEM costs across a sample of countries which might be interesting targets for UK retailers.
Here are the raw averages, shown as a percentage of the cost of the same terms in the UK:
When you look behind at the raw data, it becomes evident that we’ve made a reasonable job of eliminating language bias, but not cultural bias: some brands are relatively more popular in some countries than others, which distorts the picture somewhat.
One way around this is to weight the averages according to the popularity of the search term. Here’s the resultant chart:
This does balance the results a little, but it still seems improbable that countries like Germany or the Netherlands are so much cheaper than the UK.
A further dig into the data reveals a strange anomaly. The suggested bid for consumer electronics brands is much higher in the UK or US than it is in continental Europe. ‘Samsung’, for example, is apparently 6 times as expensive in the UK as in Germany. Otherwise, there is reasonable consistency. So here is the same chart with the consumer electronics brand names removed:
What can we conclude?
Firstly, although there are variations in their scales depending on how I’ve cut the data, the graphs are all essentially the same shape. Some countries, such as the US, are apparently expensive places for SEM. This probably makes intuitive sense: it’s a populous country, online retailing is well developed and therefore competitive, and it’s a rich country so retailers can afford to pay well for SEM keywords.
The apparent low-cost of Japan or Korea is almost certainly driven by language/alphabet issues, plus the fact that Google is by no means the dominant search engine there. However this doesn’t seem to apply to Russia, where Yandex has higher market share than Google; a possible explanation is that well-off Russians who can afford to buy non-Russian brands are more likely to use Google (and English).
Nevertheless an important second conclusion might be that there are opportunities to grab customers cheaply in some places:
Finally, we can probably conclude that this methodology is reasonable, but not perfect. For example, I suspect the apparent cheapness of France is an anomaly somewhere in the data or methodology. Nonetheless, it does seem to represent one reasonable way to estimate these costs for early budget/planning purposes without engaging in a full-scale exercise via a local agency.
Read our follow up article on Global SEM Costs Part 2.