Global ROI Part 3 – Impactful Action

In Part 1 and Part 2 of this article series, we looked at the cost to decide, and the zen of inaction to go forward and build your plans for global growth. In part 3 we will dive into some effective models and deconstruct the windmill that ROI cases can be.

Winning cases include a great story, supported by believable evidence, arriving at a fortuitous moment, to a receptive audience. They also include a clear definition of the Returns and how they can be measured.

  • You must have a crisply defined “R” – keep it simple and clearly tie it to your organization’s core goals.
  • You need access to the data to support the “R”, both current and planned.
  • Your model must contain a timeline, ideally with the timing of investments and returns and not simply activities and milestones.

It may be hard to envision the ROI model and all of its components at first, but most importantly you need to Establish a Time Window, and within it – all direct costs, all indirect costs, and all overhead costs (e.g. ~$22k a year for an employee of a Global 500 Enterprise to have a desk in a building in San Jose), mitigated by capital labor credit and expected capitalizable costs. Timing can also have a huge impact on the direct costs. A good timeline sequence will optimize on effort and costs as well as revenue.

What’s in a Good ROI Calculation?

Perhaps a better question to start with is ‘Should I try to calculate ROI for this?’ ‘Will I have the components needed for an in-depth ROI?’ It’s important to note what assumptions you are making and why. Mapping the data requirements as early as possible to limit the number of basic assumptions that will be made will streamline the work and strengthen the end result.

Keep in mind Occam’s Razor, “When presented with competing hypotheses to solve a problem, one should select the solution with the fewest assumptions.” So, if ROI is not appropriate, given the number of assumptions, then maybe a simpler hypothesis can be tested. (Like the hypothesis that MT works as well as human translation in terms of some truly measurable business activity.)

In any type of case, it is important to hit both the top and bottom lines, or build a solely cost avoidance style case. Since budgeted costs will usually be spent regardless, there are rarely any actual savings to an overall organization. This can create the perception that cost avoidance cases are basically a ‘pack of lies’ since the money will still likely be spent elsewhere. Causing increased revenue and growth is a much better basis to work from.

Going global for the first time is very different than adding markets. Both typically require the services of language service providers (LSPs). The best optimizations can occur when you choose vendors whose business model aligns closely to how you want to interact on the business level. Costs and quality are easier to balance and change the closer your mutual business objectives are. Data regarding this activity will be the most accurate when working with your provider, and assumptions in this area are too impactful to make without great data.

Typically, the translation costs are quite tiny compared to the costs of an organization’s methodology. For product companies these costs can be attributed to the Cost of Goods Sold (COGS), but for other organizations such as platform companies, these become Operating Expense (OpEx) dollars. When money is COGS, it should be coming back as goods are bought, but if you are a platform or information-based organization you are going to increase your Operating Expense. Capital Expenses, such as servers and software licenses can be amortized over a period of time. Resource effort can, in some cases, also be capitalized. Enabling toe costs of effort to be recognized over time – effectively making labor costs cheaper and the results an asset that is tracked in the accounting books.

In an ROI model, with your clearly defined Return, next clearly define the Investment and investment timeline. Your model will contain things such as:

  • The cost to decide
  • The cost to implement
  • Sequencing optimizations impact on implementation timeline
  • Current cost to operate
  • Projected cost to operate
  • Financial accounting on timeline and all mitigations including:

Licensing, headcount, facility requirements, OpEx, CapEx, and potentially COGS

  • Cost to produce current/future
  • Costs to start, run/maintain, stop for things like assembled products
  • Market expectations
  • Timeline
  • Methods
  • Projected Impact(s) – clear path to the ‘R’ (Example 250% uplift to conversions/MAUs )

Content Powers Transactions

In my previous role I owned a lot of content, and was able to obtain fantastic data. I was lucky and had direct revenue and cost data, access to all supplier contracts, and the components and construction of each property in each locale that the company maintained. Regulated industries may incur higher costs in introduction and qualification of products, but usually more than make up for this in the value of the data they are required to maintain.

Billion-dollar market influencers produced direct effects in models and a lot of dramatic up and to the left graphs. Worthless. Like a paper dollar was, without some content to enable a change – to power the transaction required for it to have value.

I could see some clear places where the organization had not adapted quickly enough. I built a case for global growth, and then extracted 3 simple markets with high, untapped returns to start. Global properties with a content language mismatch causing a significant lack of traction and loss of revenue.

The content was accurate, but available in 4 languages not likely to be understood by 70% of the inhabitants of the top country on the list. I built a 6-page presentation to gain support for solving the problem. Then I incurred 2 quarters worth of ‘costs to decide’ on the market change, which overlapped 7 quarters of local government lobbying to reinterpret our value to the country and our standing to do business there. In this case, 2 years were spent gaining permission to localize an deploy the offering to the target market.

The regulators needed a clear implementation plan, and examples of the fully-localized experience for approval. This timeline allowed for other optimizations that reduced the effort of this market entry by half, and allowed us to produce the materials needed for regulatory approval in record time and on the new low-cost basis. A regulatory timeline delay enabled completion of fully-automated quality testing to be completed across all languages, further reducing costs and enabling velocity.

Merchant onboarding was rolled out first.

It’s Flat!, The Earth is Flat!

The product organization came 30 days after the initial deployment with ‘dire news – revenue for the country is mostly flat.’ Unfortunately, this was not the measurement nor the Return goal for this point in the timeline. I had failed to keep my stakeholders on the same page in the unfolding story.

Although the product team had done some work, the results were seen by marketing and analytics. Site hits went up >1200%, beating forecast. Merchant conversions went up >330%, also beating forecast. Awareness had risen greatly, and more merchants were quickly onboarding, paving the way for consumers to buy their goods, and then meet the monetization goals of the product team.

What the expected Return was, and how it would be recognized were not the same at each point in time. Scrutinizing the results of each phase enabled adjustments and optimization of the next phase.  Limiting the initial scope of what was rolled out enabled it to be more easily accomplished, clearly tracked, and if successful, further enabled the next phase.

After 2 months of positive results, the next phase in the timeline began and consumer onboarding was released in the same language and country. In-country revenue lagged forecast slightly, yet global transactions with the country soared. By the end of quarter it became clear we were on-track to exceed all of the Growth, Engagement, and Monetization goals.

Enabling merchants first meant that the very different R for customers – finding merchants to buy from – would be fulfilled before full monetization could occur. It also greatly reduced the initial effort and costs for the work. Limiting the initial scope also meant it could get done early in the year, meaning more revenue could be attributed to the country earlier, and over a longer period of time. In this case, users in all other enabled countries could transact with the newly onboarded merchants, driving better than projected uplifts in transactions from other markets. By releasing native-language content, in the right sequence, people discovered and understood each other and began transacting.


Often, from a finance perspective, localization functions spend inconsistently, and have difficulty in identifying the financial metrics and levers available in their business model. Leaders who address and resolve this, greatly impact the global reach of their organizations.

There are a lot of ways to do this, as each environment is different and contains its own challenges. In some, the true costs to decide can be daunting, enough to warrant the move to not decide. Great data, insight, timing, and a great story can power global growth, and these work most successfully in concert with each other. Growing globally may not be easy, but it is always worth the journey.

I hope this ROI series left you better equipped to grow your global business in 3 easy steps. Crisp up your goals, collect the data, and tell a great, global story.

-Ben (