Author: Chris Jones, Independent Multichannel Retail Consultant
In the two previous articles (‘Sizing the prize: Using a top-down approach - Step 1’ and ‘Step 2’), we took a look at how to use high-level statistics and broad strategic perspectives to form a top-down estimate of potential eCommerce sales in a target country. It’s obviously important to do this to help build the business case for investing in a focused drive for sales there, for example by creating a localised site.
Estimation processes are often heavily dependent on a small number of key assumptions (such as the weighting-factors used in the suggested methodology), and so it’s always a good idea to have an alternative method to hand to corroborate, or challenge, your results. The most important factor in any such alternative is that is should use a completely different set of key assumptions if at all possible.
Given that we used a top-down approach previously, a bottom-up methodology is a nice contrast, which typically won’t duplicate key factors in the assumption process and which therefore should lead to a reasonably independent alternative estimate. Once you have both top-down and bottom-up estimates available together you can explore why they don’t meet in the middle (or be very suspicious if they do so first time through!). The outcome should be a much more robust business-case assumption that if you were to adopt either methodology alone.
The approach we’ll use here is once again two-step:
In step 1, we’ll estimate the expected visitors. In step 2 we’ll move on to estimate cart-sizes and conversion rates for these visitors.
For the purposes of this estimation, we’ll divide the traffic that arrives on a site into five rather broad types – direct; natural search; paid-for (e.g. SEM, or affiliate) directly driving visits; “effort”-driven (typically social media activities); retention / reactivation – and consider each in turn.
Well start with direct. You are probably already receiving visitors from your target country. Of course these visitors can be switched to a local site, if that is your plan. However your baseline assumption should be that there will be no incremental direct visitors simply in consequence of localising, unless you have statistics suggesting that there is pent-up interest in your brand. Direct traffic is almost always proportional to brand strength; if your marketing budget intends to build your brand locally, you can plan for some increase, but not otherwise (incidentally overseas visitors that do take the trouble to track down your UK site often convert exceptionally well – they really want to buy your stuff. You cannot assume that this will be true as you cast the net wider).
Next up is natural search traffic. If you localise / translate your site, you should expect to capture a good proportion of the local traffic for each keyword. However it takes time to get anywhere near as good at SEO as you are on home ground. A sensible planning assumption is that your keyword ranking will be two places lower in a foreign language, or one place lower in an English-speaking country, than you are used to at home, and that probably it will take 3-6 months even to reach this level.
The chart below (taken from The Multichannel Retail Handbook 2016 Edition) gives an indication of the consequences of this. If you are used to being ranked #2, but will now be ranked #4, then you’re only going to capture half the usual share of traffic. If you are enjoying the benefits of top ranking at home, then the drop-off will be unpleasant.
However this is of course only half the story. A very important point to remember, which it is surprisingly easy to overlook at the planning stage, is:
If there’s no traffic, you can’t capture it!
If the bulk of your home traffic is driven by searches for brand-terms (e.g. “PGTips”, “PGTips Tea Bags”, “PG Tips Tea” etc), and there’s zero volume for your brand in the target country, you might capture 100% of zero, whether via direct, natural or paid traffic! If, on the other hand, your home traffic is driven by generic terms (e.g. “English tea”), then you’ll be competing with already locally-established players.
These same principles are also true of paid traffic. In another article (Global SEM Costs Part 2), you can find a proposal for how to roughly estimate the likely traffic you can get for generic terms without undertaking a lengthy translation exercise first. For brand terms, the fastest method is to take a look at Google Trends, although the Keyword Planning tools will also do the trick (and is much easier to turn into a spreadsheet!).
In the same article, you will also find a comparison of the cost of paid traffic in various different countries. In general, the US is more expensive and most other western countries are about 60-80% of the cost of the UK – more detail in Global SEM Costs Part 2. So you can expect to get slightly less / more bang for your buck depending where you are targeting.
What I’ve termed effort-driven traffic – for example as a consequence of activity on social media – is less brutally (and often depressingly) arithmetic. Language is obviously a big issue, and the overall story might be rather similar to that for direct traffic – you can move existing fans over, but attracting new ones depends on other marketing activities too. However social-media type activities have a secondary benefit of being marketing in the true sense. What you obviously can’t assume is an immediate leap to a level in the new country proportional to its population, even if you do decide to invest in local language, local thought-leaders or local celebrities; what you might be able to assume is that you can get there over time, perhaps using the same 10%:20%:40%;70%:100% growth rate over five years that we used in the top-down estimation processes.
Finally we have retention / reactivation, both directly by methods like Google Remarketing (which isn’t fully available in all countries – do check the rollout phasing if you’re considering somewhere relatively obscure) or tools like email campaigns, or the more indirect “second purchase” which is implicit in many of the figures above. The obvious point to make here is that until you have visitors, or buyers, there’s nobody to retain or reactivate. Your planning assumption for year 1 should probably be close to zero traffic from this source.
In each case, it’s possible to make a reasonable comparison with data that you already know - the traffic in your home country – and consider the factors that will make it less in a new one. The overall messages are:
Once you’ve estimated the likely traffic you might be able to attract, we need to consider whether you will be more or less effective in converting it, which is the subject of the next article.