We’ve heard it countless times. ROI is the “holy grail” of the localization industry.
Recently my friend Jeff Beatty, Senior Head of Localization at Mozilla, made an astute observation with regard to measuring localization ROI:
“ROI is an inaccurate and false metric for localization, feature development, and release engineering, the list goes on because it’s based on often false assumptions and prophetic predictions about the market. Translation is a revenue enabler not a revenue generator. Because we try to fit it into the generator concept, we get stuck with the false ROI metric. The enablement concept can clear the way for us to focus more on opportunity, outcomes, purpose, and process and less on moving units (words).”
Jeff knows a thing or two about this topic — he has seen Firefox go into 90 languages, after all. And, after countless conversations with peers working in localization roles across software companies of all types, I would tend to agree. There isn’t a direct and perfect calculation to help determine return on investment in localization.
But if you can’t figure out localization ROI, what can you measure instead?
Metrics You Can Use For Localization Initiatives
Here are some things to consider when you’re asking yourself how to measure the impact (not the ROI) of localization at your company:
- How much supported language revenue does localization enable? How much revenue does your company obtain from the countries that speak the languages your team supports? I refer to this as “supported language revenue.” It’s a revenue enablement metric I look at pretty often. By tracking this metric, you can look at your investment in localization as a percentage of the revenue generated from the languages you support. However, you cannot attribute all of that revenue to localization, because localization isn’t the only team involved in generating it. If you try to compute what percentage is attributable to localization, you’re just as unlikely to find answers as you are if you try to figure out what percentage of revenue the team who built the website contributed, or the team who created your invoices. And surely your demand generation and sales had something to do with making the sale? What about product marketing?
- What is the cost of localization as a % of supported language revenue? Another metric you can look at is the cost of localization compared to the total amount of revenue it supports. This will give you a percentage value, and ideally over time as your localization program scales, you’ll see this number shrink. That said, there is no formula for figuring out what the “right” investment or percentage is. It will change over time, and depending on what you localize versus do natively, what your go-to-market strategy is, how much content is involved, and so on.
- What are localization costs as a % of the cost of acquiring a customer? In SaaS companies, customer acquisition cost (CAC) is a widespread metric. However, figuring out customer acquisition cost for international customers can be tricky, because often many costs of serving international customers rely on central teams and are not completely regionalized. However, if you know the CAC for all supported language markets, then you can easily calculate your localization costs as a percentage of this. The reason this isn’t usually done is that product localization is usually a big part of the total localization spend. However, if you can segment out what you spend on product localization versus sales and marketing, for example, you can figure this metric out too.
- What are localization costs as a % of the R&D budget? One way that many software companies look at localization is simply as a percentage of their overall research and development budget. This is nice and simple. However, if localization also supports other functions outside of product, this isn’t a perfectly clean metric unless you can map them squarely to product versus other functions.
When you go through all of these and figure out these metrics, it’s nice to have some proof of the work that localization does and its impact on associated costs and revenue. But it still isn’t a true ROI equation, which would clearly say that you put X dollars in and you get Y dollars out. Everyone can track X, sure, but no one can ever say for sure exactly how many dollars came from the localization investment alone, versus additional investments your company has made in serving those very same customers in a given locale.
Translation agencies are often fairly naive in this regard. They often think their work has outsized importance in unlocking a company’s revenue in a new market. I’ve even heard some go so far as to claim that their customer has achieved a certain amount of international revenue thanks to their efforts. It sounds something like this: “They spent $200K with us on localization services, and in their last earnings call they said they earned $100 million in international revenue. That’s clear ROI.”
Hold on. There is far more that goes into international expansion than just localizing something. Launching an office in another country can be an incredibly resource-intensive, detailed and complex process. Building teams and creating channels in other languages doesn’t just take energy from localization, but from the hiring managers, human resources partners, payroll specialists, recruiting teams, and many others — and that’s all investments they have made before the company’s other non-English employees have even set foot in the door for their first day of work!
Then, when it comes to actually serving customers in those languages, training them, supporting them, and renewing them, localization isn’t something you can just directly link to ROI. It’s really just table stakes. It’s what you have to do in order to operate and scale there. While it’s true you can’t do business without it in many places, you also can’t do business without, say, a product, a sales team, a website, a marketing team, and so on.
In summary, localization is one of many things that helps enable a company to unlock revenue in other countries and languages. It’s hardly the only thing, but it’s a mission-critical piece of most go-to-market strategies.
So, if you’re not going to measure localization ROI, what can you measure instead to show the value of localization? As I wrote in a separate post regarding a one-word definition of localization, it’s the customer experience. CX is the value of localization that really matters to any business, and that you can explain to anyone only to see heads nodding in agreement. Talk about ROI and you’re quickly going to lose people, because it doesn’t make logical sense to talk about localization that way.
As such, one thing you can do is to partner with others at your company to carry out a customer experience survey. Ask the same questions every time, and ask about the end-to-end experience in another language. Measure the improvements in each area and overall. Work with the teams that own those aspects of the CX in another language. Measure it for each language and each area of the experience, and see where the gaps are so you can improve.
True, it’s not going to answer a localization ROI question for you, but I don’t think the ROI question was ever the right question to ask. Asking about the customer experience in another language will at least help you understand the customer’s perceived value when they work with your company in another language. And, if you run a localization team, you’ll be able to better support your internal stakeholders with proactive initiatives to help improve that non-English CX.
CX, not ROI, is a more natural fit where localization can truly show value, because it’s obvious to most people that localization is an important driver of the CX in another language. CX is also a good place for localization to be highly visible and integrated with other teams working toward the same goal. At any customer-centric company, ROI on the other hand is merely a formula that looks at dollars in versus dollars out. The customer experience is what actually has the ultimate power to increase the dollars out side of the equation. Improve the customer experience and they’ll see more value and be willing to renew, or even pay more for it. Don’t we want localization to be associated with the value that truly enables revenue and loyalty, instead of just ROI?
Well, I certainly do. Maybe it’s time we all stop holding out hope for a mythical object with magical powers. I’m giving up on the fairy tale of localization ROI and focusing on the customer experience instead. Care to join me?
Natalie there is no doubt the ROI rabbit seems to elude the LSP greyhounds in every race. But I am not sure that means we should throw it out. That just means we haven’t figured it out. Even your own suggestion, that l10n be wrapped up in the CX conversation runs into the same issue. CX was a thing for a long time that struggled to gain traction until Fred Reichheld and the Bain folks established NPS as a metric that could be used to connect CX to revenue. There is only one metric that matters at the end of the day, and it always starts with a $. All other metrics have to connect to that. Call it ROI, ROA, ROE, RO E(xperience), what have you.
I am not giving up on ROI, not do I think it’s the wrong thing to pursue. You are 1000% correct that l10n is not a straight line contributor to international revenue, but neither is TV a straight line contributor to sales, or awareness. Many things are not straight lines, they require correlation analysis, A/B testing, and various and sundry other means of connecting them to $$$. One significant retailer I know has a quite robust model that can tel you exactly what l10n of their product descriptions contributes to revenue lift. And that’s they key, you have to be able to show the incremental lift.
Maybe I am chasing the fountain of youth, but since everything else in marketing is increasingly being measured and modeled there is no reason l10n can’t or shouldn’t be. If you would like to join me and Ponce De Leon on our quest, always happy to chat.
Jaime, you make some great points. I think there is nothing wrong with continuing to find ways to measure localization ROI. My general sense however is that we need to move away from “dollars in, dollars out” type equations and toward something that ties more into CX. But I’m happy to be proven wrong too! I think every time anyone can say “this positive thing happened thanks to L10N” it’s great, but want to just ensure it’s truly something that happened thanks to L10N and can be separated out from all other things that also happened. That is usually the hard part. 😉
Well, in the end “dollars in, dollars out” is what the decision-makers actually want to know, so we’d better try and answer that as close as we can!
It’s definitely doable for digital products with digital distribution (apps, games) but we need to use scientific methods, isolating the localisation factor (as much as possible) and carrying out A / B tests. There are a few ways to do that, some of them require cooperation from the app distribution platforms, others don’t. As for surveys in general, they are a poor predictor of actual buying behaviour (and it’s the buying behaviour that matters to the decision-makers). I’m rather voting for experiments and observations, probably aided by expert interviews with product managers on the client side, which won’t be super-accurate but would generally validate the A / B test findings.