Measuring Localization Velocity

If you work in software-as-a-service (SaaS) or any other area of high-growth tech that requires fast-moving, ongoing localization, chances are you’ve come to realize that most of the models out there for localization don’t mesh with your business reality. The underlying cadence in any hyper-growth business is… different. The company’s heartbeat is faster. The blood pressure of the organization is higher. It’s thrilling, but intense.

But it’s not just the pace that is a challenge.

It’s the frequent and rapid changes in direction.

It’s like sprinting a marathon, on a course that unfolds while you run. With lots of unexpected twists and turns.

What Localization Looks Like at High-Growth TECH Companies

In high-growth organizations, the demands on localization teams are also different in several ways:

  • Speed of delivery is the highest priority
  • Quality is a baseline expectation
  • Volumes are nearly impossible to forecast
  • Stakeholders change continually

There are many things that this reality forces us to re-think and reconfigure about traditional localization systems. But the biggest one, I’ll argue, is the core way we measure production. If speed is the highest priority, and offers the greatest value, how can a localization team measure this?

Localition Throughput < Capacity < Velocity

Traditional localization systems like to measure production in terms of throughput. Past tense. “How much did we localize?” The problem with throughput on its own is that it’s merely a look back at how far you traveled. At a high-growth company, no one really has the time to look back. Everyone is gripping the wheel tightly and keeping their gaze intensely forward. From their localization team, they want reassurance that the vehicle can keep driving with minimal disruption.

On my own team at HubSpot, we still measure throughput of course. But that isn’t sufficient for our company’s high-growth, fast-paced context. In recent years, we also began to measure capacity. “How much could we localize?” To measure this, we looked at the baseline of how much our vendors are routinely able to deliver without sacrificing quality, on top of how much our internal team can deliver. This gave us a capacity range for each language, to help us with production and headcount planning.

But that still wasn’t enough. We realized that we not only need to measure just the numbers and the capacity they represent, but a metric that could measure our velocity as well. Liken these differences to those between simple arithmetic (throughput), more advanced geometry (capacity), and physics (velocity).

What does velocity measurement look like? It’s a top-line metric that looks at localization production in terms of words per time increment, such as hours, minutes, or seconds. It’s a way to measure the total health and progress of a localization program. Fortunately, the metric itself is very simple to understand.

Example: words per hour.

To figure out the velocity, take the total number of working hours in the time period you want to measure (day, week, month, quarter). Divide your total capacity by the number of working hours. This tells you your total potential velocity. Then, to figure out the velocity you realized in the same time period, divide your throughput by the number of working hours. This tells you your total actual velocity achieved, the speed at which your localization system truly traveled.

Using Localization Velocity to Improve Relationships

If you work in a fast-paced company, and your stakeholders are demanding speed, turn-around time metrics are probably just not very helpful for you to map to their reality, which is more “real-time focused” than “past-focused.” Metrics like turn-around time and throughput, which are past-focused, might be too simplistic to reflect the complexity of your ongoing, continuous projects and reality of the localization system you are building.

Localization velocity (which we’ll initially be measuring in words per hour) is a new metric for our team. I’m sharing early thoughts about it, because I think it could be helpful for other teams at high-growth companies that are facing similar challenges in terms of figuring out what metrics will really make the most impact. It’s important for us to all help our internal stakeholders understand the value of what we actually do, in terms that they actually care about. That, in turn, helps build stronger relationships.

Speed vs. Quality: A False Trade-off?

One of the concerns many localization traditionalists might have about focusing on a velocity metric is that it could encourage teams to sacrifice quality. That might be true in some contexts. But at a high-growth company where quality is always a baseline expectation, I don’t think that is actually a valid risk to worry about. Actually, I worry more about the localization industry’s longstanding focus on “price per word,” which only represents a small portion of the full cost to any buy-side company, than I do about a focus on speed. When we focus merely on the number of words delivered (throughput) with a certain amount of budget, we are minimizing, or even erasing, the strategic value of localization.

At high-growth companies, investing money wisely is, of course, always a concern. But the question is not, “Are we squeezing out the maximum number of words from this investment?” The question is instead, “Are we investing for high growth at scale?”

Bottom line, if you’re working at a high-growth company, and especially if you have a digital delivery model (SaaS, e-commerce, and so on), the traditional localization playbook and associated metrics might not be a great fit for your business. Many of those were invented decades ago. They mapped to a different reality, that of waterfall development and staged releases. We live in very different times!

In today’s reality, where the line between what constitutes digital marketing and software development gets blurrier each day, where nearly everything can be done online, and where the location of users matters less and less, and where on-demand access, speed, and uptime are everything, localization teams just can’t keep recycling yesteryear’s metrics, processes, and playbooks. Many of them just no longer work.

The New Normal Requires New Metrics

The so-called “new normal” — society’s collective accelerated leap to geo-agnostic, remote-friendly practices in order to advance the digitization of how we live, work, and learn — really does require us to change the way we do things. Even those of us who have long advocated for the very things that are happening right now, to enable further business globalization. 😉 Many of us in localization have always seen the world as a bit flatter and remote-friendly than others, due to the nature of our own globally distributed work. However, the pace at which others are moving toward this as a reality now puts even more pressure on us to step up our game and come up with even better standards and practices than we used before.

I believe that someday, one industry standard metric for localization teams at buy-side organizations could be the number of words per hour, per minute, or even per second. It would reflect how automated their processes are, as well as how adeptly they are leveraging the best technology has to offer. It won’t tell the full story, obviously. And it might not be the best metric to apply to all types of content. But when looking at an overall localization system, and the real-time nature of whether teams are capable of delivering content quickly to customers, I do believe it will be important. It might even help us understand the phase of maturity of various localization team initiatives, and what benchmarks are common, say, 2 years into building out a new program, 5 years in, and so on.

I’ll be sharing more on other metrics and concepts that I believe are essential for running localization teams in this new reality in future posts. My hope is that other companies building localization systems at high-growth companies, especially in the #newnormal, can benefit from what we’re learning, and share what they’re learning too.