Best Practice: Global Outsourcing

Some of these teams focus on software development, some on web globalization and e-business, some on content or software localization. However, the common challenge for all team members, as Janet Webberley (Development Manager for Xerox Global Services) expressed it, is to find ways to ensure that the team works as a team to deliver a result, rather than simply as a group of people working on the same program, but effectively in isolation.

And as Regine Herzog (Information Development Lead at Rockwell Automation) points out, underlying processes are often the secret to a team’s success: “My experience has shown that consistent, well-structured processes that emphasize high quality output can be considered as efficient a measure of a team's success as mere statistics.”

In this article, I will share how the following companies …

Alcatel-Lucent
Business Objects
EMC
HP
PerkinElmer
Rockwell Automation
SAS Institute
Schlumberger SEED Foundation
Sun
TIBCO
Xerox

… responded to the following question about global outsourcing metrics:

How is your global virtual team being measured to determine whether or not it’s successful?

These same companies also shared their views on the following four issues:

1. If you need a person with a certain skill set to join your global virtual team, how do you recruit that person?

2. Are you working on anything (it could be a product, a service, a piece of workflow, the way your team works together – anything) that should be considered a “reusable asset” for use by another global virtual team? What is it? How will you ensure that the next team benefits from what your team has learned?

3. What is the most difficult thing about working on a global virtual team for you? Would other members of the team agree with you? If not, why not? How could you overcome this challenge?

4. What is the one hi-tech tool that you can’t do without when you function as a member of a global virtual team?

If you would like to receive a copy of the Ask the Experts Guide that will include all of the data and analysis (along with background information on the teams) gathered from the 17 managers, directors and vice presidents who responded to these questions, just click Rebecca@lisa.org to send me an email.

And, now, onto what LISA Members say about the metrics they use to measure the success of their global virtual teams.

---

How is your global virtual team being measured to determine whether or not it’s successful?

Bill Price, Manager, CTO - CIS Localization Services, Alcatel-Lucent

Metrics are critical for us, since other groups within Alcatel-Lucent are not required to use our services. In addition, the localization function has always been somewhat of an oddity here, so we always face a requirement to demonstrate at short notice what we do and what value we add. That’s why we developed a balanced scorecard format for our metrics. We also maintain a TL9000 certification to help maintain our credibility.

There are three key metrics included in our balanced scorecard:

* Financials

Our organization is required to prove on an on-going basis that we reduce Alcatel-Lucent’s costs for localization and translations services. This involves constant scrutiny and adjustments of our internal processes, as well as the diligent tracking of cost savings as compared to general market rates.

* Customer Surveys

We do these at the end of a project, or on a quarterly basis for long projects. Our clients are the source content organizations within Alcatel-Lucent that fund localization projects and then deliver the documentation (or whatever) to the end customer outside of the company.

* On-time Delivery

This is a tricky one because there are so many people who just want to push a button and receive their translated document. And, of course, their perception is that they never get it as soon as they want it. Therefore, the only way for us to fairly measure on-time delivery, since there are many factors out of our control on the content creation side, is to ask our customers to focus on whether they received their finished product when we said they would!

I will point out, though, that we can meet impossible deadlines when required. For example, one of our customers (NTT Japan) performs general acceptance testing for hardware and software. In these cases, all documentation and other materials must be in NTT’s hands the day before their acceptance tests begin – whether or not the delivery date for these materials is realistic. Of course, we have a few tricks up our sleeve, including a set of our own tools that allow us to leverage updates in a way that no other provider can.

---

Kirsten Sutton, Director, Translation, Business Objects

External to our team, we are evaluated on:

(1) Timely delivery of translation, i.e., that we are not holding up releases, that we are not critical path to completion of any features/tasks, etc.

(2) Project cost/savings. Typical project-based metrics include external outsourcing costs, internal staffing costs, technology savings (TM reuse, etc.).

(3) Throughput that can be managed by our team, specifically that we are never in a position to turn away work due to resource constraints.

Internally, we also evaluate translation quality through using portions of the LISA QA Model, customer/partner feedback and internal linguistic review.

---

Marcia, Metz, Director, Globalization Program Office, EMC Corporation

Our team is measured on successful delivery, as well as on overall satisfaction. Successful delivery includes on-time delivery, as well as on-budget delivery. Satisfaction is measured on technical, management and overall product/language quality. Additionally, we are measured internally on (1) how efficient we are in terms of our own spend, and (2) if we are effectively utilizing certain locales to minimize cost and maximize efficiency, whether through time zone, proximity to vendors/partners, etc.

---

Peter Coleman, Program Manager, HP
Gregg Okunami, Program Manager, HP

Editor’s Note: Coleman and Okunami are members of an internal HP team that provides globalization consulting services to the rest of HP.

Coleman: We’re measured on quality and cost through tracking our volumes, throughput and customer satisfaction. With regards to the latter, we don’t proactively measure customer satisfaction through surveys. The real metric is whether someone pulls out of our centralized services infrastructure for obtaining globalization and localization services. This happens rarely.

Okunami: We do hold post-mortems for our projects with our internal HP teams. And, Peter is right, attrition is rare.

I would point out that our cost per word is a very important metric for our team. We’re driving HP teams (1) to centralize translation memory (TM) to gain more and more reuse, (2) to consciously ensure more consistency through terminology management and (3) to lower their costs overall.

Your readers should also know that our cost per word includes our TM cost and the actual vendor cost to translate words. In addition, we calculate these costs a bit differently, depending on whether or not the content is maintained in one of our CMS systems or processed outside of CMS. For example, our costs are measured from the time a piece of content goes into the CMS until it is returned to us. For non-CMS content, we measure the time from when we receive the files until we turn them over to another team.

---

Hans Keil, Director, E-Business Channel Development, PerkinElmer

The number one success factor for our web sites are lead / direct revenue generation. We also pay attention to the number and quality of qualified visitors (visitors who identify themselves through registrations or form completions) whom we can confirm as customers (as a B2B site we receive a significant number of non-customer visits). Since we are still in the earlier stages of building some of our sites, we also track how much content is actually going up on each site.

Overall, we need to prove, on an ongoing basis, that there is incremental revenue being generated from a particular site – whether in the form of e-commerce or leads – to continue to fund it. We haven’t been asked the following question yet, but we will be soon, i.e., should we translate our product catalog into German, and if so, what will the return be?

The U.S. site has an e-commerce component, so we focus there a lot more on analytics, shopping carts, bounce rates and abandonment rates. The vast majority of researchers and technicians set English as the default on their browsers, but the Germans stand out, and the Chinese visitor rate is definitely climbing.

---

Heike Caldwell, Translation Manager, Rockwell Automation
Regine Herzog, Information Development Lead, Rockwell Automation

Caldwell: By using our internal Localization Tracking System (LTS), we can create statistics for turnaround times and expenses. This helps me to identify improvement possibilities. We have also implemented Quality Metrics to measure the quality of the language content (closely following the LISA QA Model).

However, these metrics are not intended for use by either our internal or external translation teams. We know what works and what does not, and where the weak points are. The statistics are required by management to justify certain expenses, e.g., proofreading or a change of vendor.

I'd always recommend metrics for time and quality because they are the only way to justify costs.

Herzog: Localization projects are usually part of larger release projects, so they are planned and tracked centrally by our PMO (Program Management Office). I create reports on a regular basis to monitor schedules, along with our internal efforts and external costs. At the end of a large project, we always hold a post-mortem, which can also lead to a lessons learned meeting or a document, if necessary.

I've found that consistent, well-structured processes that emphasize high quality output are considered as efficient a measure of a team's success as mere statistics.

---

Trine Lai, Senior Localization Project Manager, European Localization Center, SAS Institute

SAS maintains systems to track errors and questions during the translation and localization processes. We have our own tracking system for Europe to filter out the issues and questions that we can solve ourselves without sending them back to the U.S. development teams. We ensure timely delivery of high quality localized software, but our role also involves acting as an effective advocate between the developers and translators on our worldwide teams. We meet once a year with all of our translators and testers worldwide and invite SAS developers and other colleagues to join us.

---

Eya Tkachenko, Translations Coordinator, Schlumberger SEED Foundation

Unfortunately, we have no metrics system in place to measure our overall quality. We do produce monthly reports where each member pens in the highlights and the lowlights in his/her work. We also judge our success by reviewing the project schedule created prior to a project launch. Lots of dates in "red" indicating delays provide us with an idea of how successful or unsuccessful we are in terms of time.

I recently attended Lean Six Sigma training where I learned about various tools and instruments that help team members evaluate their performance and create a culture of continuous improvement within a company. In case not all of your readers are aware, Lean Six Sigma is a methodology for process improvement that uses quality and statistical tools to reduce defects, variation and waste. There are different types of analysis that could help us track and measure our team's performance here at SEED in terms of time, money, quality, etc. However, to determine what analysis or methodology will work for all of us, the entire virtual team has to get together and decide on a tool.

---

Melissa Biggs (and her team), Sr. Globalization Business Manager, Sun Microsystems

We measure our deliverables in tangible ways, e.g. our translation volumes in terms of words/releases. We measure the success of our overall organization through decreasing costs, on-time (or shorter) deliveries and the ability to provide more deliverables and value add to field partners within the company. We also evaluate ourselves in “softer” ways, e.g., whether or not we are gaining better visibility through increased voice of customer data.

---

Fiona Tan, Vice President of Engineering, TIBCO Software

Currently, our distributed development teams are being measured by their ability to meet delivery dates. We do not have very quantitative measures in place; however, we have used some in the past for a different partner. In that case, we measured their ability to release a product by the date promised with all functionality as specified. We also looked at the number of defects reported against the product after it was released and graded them accordingly.

---

Katia Azevedo, Language Services Manager, Xerox Global Services
Janet Webberley, Development Manager, Xerox Global Services

Azevedo: We establish targets, action owners and due dates, as well as desired outcome for projects during global team kick-off meetings. We subsequently monitor our progress on a regular basis through live meetings and make adjustments as needed to conform to any changes (foreseen or otherwise) that may impact the projects. In terms of tangible metrics, we do utilize QCD’s (quality, cost and delivery) as key drivers to help measure a team’s overall effectiveness.

Webberley: Our standard KPIs (key performance indicators) include cost/budget, quality and time/schedule for the team's deliverables. However, there are also soft aspects of success, such as cooperative working and team spirit, which we also take into account. A successful team is one that works as a team to deliver a result, rather than simply as a group of people working on the same program, but effectively in isolation. These soft aspects are, of course, much harder to measure so ultimately, the effectiveness of the team is reflected in a project's KPIs.