Measuring the direct ROI of localization is hard because, while it unblocks users, localization alone does not guarantee success in other languages and regions. There are all sorts of things that can hold you back, from regulatory barriers to in market competitors who’re already established.
One thing you can do to estimate ROI is to look at revenue on a country by country basis. Then for each country assign a weight to represent the percentage of that revenue that is attributable to the product being localized. In regions with a high degree of English proficiency (Canada, UK, Scandinavia), localization isn’t going to have as big an impact as it will in regions with low English proficiency.
Let’s work through an example. This company currently generates $20 million in revenue, 40% of which, or $8 million comes from non-English countries. The majority of that revenue comes from Brazil, France, Japan and Korea.
Here we assign weights to each country to reflect the extent to which localization unblocks users. In regions like Japan and Korea it is reasonable to assume that most users would not use the product if were not localized. In a country like France, the effect may be more muted since many people there are bilingual, so I used a lower weight there. We then estimate that of the $8,000,000 in revenue from non-English countries, $5,500,000 was probably due to the product being accessible in local languages.
The next step is to look at the cost of the localization program. See:
I’ve found that most companies do not have a giant amount of material to translate. In a typical scenario they will have a mix of in-app content, help center and marketing and website material to translate. 250,000 to 500,000 words for all of these touchpoints is a good ballpark estimate to start with. Translation services will account for the majority of the cost, and most of this content can go through hybrid AI|human translation workflows, at a cost of 10-20 cents per word (the real cost can be lower but let’s be conservative for this exercise). This works out to a cost of $25,000 to $50,000 per language, or $100,000 to $200,000 for these four languages.
Next add $50,000 per year for translation tech vendors and $250,000 per year for staff to manage the program. That all adds up to $500,000, which works out to a 10x return. What is particularly important to note is that this return will only increase as the company grows. One of the other effects of localization is it will increase your AGR (annual growth rate), and that is a compounding effect.
Another way to approach to justify a future localization project is to look at how many net extra seats do you need to sell to breakeven on the localization spend. Most SaaS products charge around $100 per seat per year at the low end. Using the example above, we spent $500,000 for four languages, or $125,000 per language with everything accounted for. Let’s say the cost of goods to operate the service is 25% of revenue, so we clear $75.00 per customer per year after factoring that in. With just over 1,600 net new customers per language, we breakeven and everything beyond that is gravy.
This is a simplified way of estimating ROI for localization, and while there are some assumptions that go into this, it is a fair way of representing the value it delivers. It is also an approach that C-level leaders can understand. One thing that frustrates me with localization is it is often viewed as a cost center, and treated as such, when it is one of the cheapest and most effective growth levers for a company. Even fairly early stage companies can benefit from this, and even if they don’t plan to expand internationally, they can add support for languages spoken in their region (Spanish for the US market as an example).