Everyone knows the stereotype of the lonely project management hero. Gaunt from too many nights in the saddle, he (or she) brings in the project when all have failed, before walking off into the sunrise in search of the next piece of trouble to shoot. In many companies in this rapidly growing industry, however, these figures are more corporate reality than corporate mythology.

They are also a danger to themselves and others, says Fergus O’Connell. The chairman and CEO of project management specialists ETP has trained nearly 5,000 project managers, many of them from the localization industry. His courses run in conjunction with LISA, are well attended. And yet, he says: “I’m in danger of sounding like a broken record. My message to people is still that they have to stop being magicians, taking impossible projects and making them happen. This style of work is unsustainable. Sooner or later you top out, and then the world falls apart.”

Hitting a moving target

The consequences of this can be serious. According to the 1997 LISA Industry Survey, 71% of publishers now generate 20% or more of sales from localized product versions, and 81% of publishers expect to be localizing into more than 10 languages in five years’ time, as opposed to only 48% at present. Getting it wrong can translate into serious money being lost.

As Kathleen Marvin, Director of Localization at graphics and page layout software publisher Adobe, says: “Economic realities are forcing both suppliers and publishers to change their working practices. The number of products at Adobe will double this year to 20, and each product is being localized into between 4 and 15 languages. This means we are having to move to greater outsourcing and we need products returned in a very high quality state. You can’t simply toss a bunch of files across to another continent, hope for the best and then scramble madly to fix them when they are returned. We all have to get away from gritting our teeth and making it happen by sheer willpower and working 24 x 7.”

“Strong project management is vital on both the services vendor and the client side”, she says, “since it is what keeps information flowing in both directions and manages the movement of deliverables. At its very best it should also facilitate problem-solving when difficulties arise.”

Investing in people

How, then, can good project management be achieved? By adhering to basic principles, says Fergus O’Connell. In ETP’s three-day structured project management courses, participants learn the methodologies needed to assess, run and monitor projects correctly and how to deal with interpersonal issues (see boxes 1 and 2 for a summary). One key aspect is the ability to say “no”, he continues: “The Duke of Wellington never took on any job that King George IV gave him without haggling about terms and negotiating a project that would work for him.”

For Gunnie Jacobsson, MD of Irish consultants Augur Ltd. this may not be enough. In her opinion, what is needed is “a measure of overall competence, a consistent system to assess such competence, and an industry-specific standard. It is necessary to recognize excellence and to ensure that such a system is accessible worldwide. This pays off in real terms, since more savings are to be gained through improved project management than through any other factor.”

Process, process, process

Well-trained people on their own, though, are still not enough in an industry facing a chronic skills shortage. Of necessity, many novices are being thrown in at in the deep end. What is more, both publishers and service suppliers are rapidly consolidating, shuffling teams and organizational structures. Process management must provide the framework to prevent failure here, says Kathleen Marvin. “By focusing on structure, planning and setting up the project to enable the vendor to get it right first time, you can do a lot with limited resources.” In more detail, she says, “publishers have to set up clear engineering guidelines and test plans which define quality, as well as delivering workable and documented tools. If at all possible, they also have to provide a solid user interface freeze—or at least an ‘interface frost’.” Equally, however, she says that service suppliers must become more professional. “We want people who have good technical skills but who also follow instructions. We simply cannot pick up the pieces any longer like we did in the past.”

ETP’s Ten Most Common Reasons Why Projects Fail The goal of the project isn’t defined clearly The goal is defined properly, but changes to it aren’t controlled The project isn’t planned properly The project isn’t led properly The project is planned properly but then isn’t resourced as planned The project is planned such that it has no contingency The expectations of project participants aren’t managed Progress against the plan isn’t monitored and controlled properly Project reporting is inadequate or non-existent When projects run into trouble, people believe that the problem can be solved simply, e.g. by working harder, extending the deadline or adding more resources

Communications are key

Virtual communications expert Dr. Jaclyn Kostner adds another point: “The old way of doing business was a contractual relationship, with instructions coming from the top down, and finger-pointing when things went wrong. In the new model, which is not unique to localization but also affects the IT industry as a whole, publishers and vendors will be forming teams based on personal relationships. Communication therefore needs to be carefully managed, and the right medium for the message must be used. One-way methods (such as phones and e-mail) can be used when no feedback is required, but sensitive issues or those requiring input should be handled via a two-way method (such as videoconferencing and Web-based or electronic meeting systems).”

However, according to many LISA Forum participants, such methods only make sense in the context of a trusting, long-term relationship, in which, says Kathleen Marvin: “The first project is a learning curve but everyone wants to make it work. The third project is when you know how it’s really going to be. To me, a good vendor is not one who doesn’t have problems. The real questions are what kind of problems do they have, how quickly they can solve them, and whether they reoccur.”

Strategic travel at the beginning of the project can help get things rolling, she continues: “We have an on-site launch at the vendors for all major projects attended by a cross-disciplinary team. The meeting deals with all the administrative tasks and then we sit down with the people. This gives us an opportunity to get to know each other. When you have had a glass of wine together you feel like you can tell each other the truth when things hit the fan. On a personal level, people don't want to disappoint people.” Nevertheless, she warns, “even if key people don't get on, the job still has to get done”.

Changing the organization

Poor organization is another problem. Says Gunnie Jacobsson: “although everyone talks about the shortage of good people, there are in fact too many project managers on both sides doing the wrong things. You shouldn’t have project managers doing file management and acting as post offices, and you shouldn’t duplicate effort unnecessarily. Project management needs senior people who have control over the entire project, including the budget, rather than just implementing orders”.

Added to this is organizational inertia. Fergus O’Connell says that “when we did follow-ups on our course participants, we found that only one in ten could keep up the project management program we’d taught them back home. This is why we have expanded our focus to cover ‘Implementing Structure Project Management in Organizations’ (ISPMIO)”. In another Wild West metaphor, he says that “projects are like cattle drives. The project manager is the trail boss, and the individual tasks in a project are the cows. All of the participants have to meet up in Kansas City at the end of the drive.” In his method, projects are color-coded, with green meaning “under control”, yellow “problems”, and red “trouble”.

ETP’s ten steps to a successful project Planning Visualize what the goal is: set your eyes on the prize Make a list of jobs that need to be done There must be one leader (“trail boss”) Assign people to jobs Manage expectations, allow a margin for error, have a fallback position Visualization Use an appropriate leadership style Know what’s going on Tell other people what’s going on Repeat Steps 1) to 8) until Step 10 is reached The prize

In addition, organizations themselves are given ratings of one to five, with level one being “chaotic” and level five being “nirvana”. Most companies, according to Fergus, are at level two, “the dawn of civi-lization”, in which he puts the cost of poor project management at a massive 50% of total annual sales. Joe Ryan, Director and General Manager of Localization Services at Sykes/McQueen is more cautious, but can still point to considerable savings: “When we introduced our cross-functional project management program, PACE® (Product and Cycle-time Excellence”), we were able to reduce cycle time by 18% and share the resulting cost savings with clients. The time needed to solve problems also dropped and the delegation of authority to project managers increased management expertise within the company”.

However, such empowered project managers are unfortunately not the rule, especially in publishing companies. As a “non-core” activity, localization staff are still often at the bottom of the pecking order. In particular, this can complicate relations with powerful development departments, which are responsible for producing the monolingual base product and whose interests are often in conflict with those of localization. For example, developers generally want to cram in extra functionality up to the last moment, while localizers want the product frozen as early as possible. Also, programmers may only have domestic markets in mind, while localizers are pushing for globalized products to facilitate their work. All too often, the relative disparity in status forces localization project managers back into the role of “magicians” mentioned earlier.

Changing management

One common reason for this lack of clout is inadequate commitment by corporate management. Despite the importance to the bottom line of localized product versions, mind sets can be slow to change where executives have little direct experience of foreign markets. In addition, as many companies are still increasing their language range, they may not realize that ad hoc processes which work, if suboptimally, for one or two versions will spell a major problem for fifteen or twenty.

How to Help Projects Run Smoothly Planning Know what you want Define mutual expectations clearly at the outset Define engineering, testing and quality requirements clearly Base plans and budgets on these detailed specifications Monitor progress against them Allocate the necessary resources Plan for contingencies People and management Ensure the necessary expertise exists Build long-term, trust-based relationships Keep teams stable from project to project, or at least during individual projects Define clear communications channels and media Hold face-to-face kick-off meetings Ensure localization project managers and vendors have the backing of other departments and of senior management Tools and practices Make sure any tools provided are finished, in working order and documented Define terminology at the start of a project

“Localizers need to be proactive here and argue in terms that development and corporate management can understand”, says Alex McDonnell of Pervasive Software Ireland, who also gave a presentation at the LISA Forum - USA. They should also not try to play the hero. “Super project managers are a real threat to the industry because they hide the rocks the company will hit in the long term—bad planning, poor specifications, poor tools, etc. They make it look as if the management problem is not there, but it is. The complexity of online products and services is increasing, while the capability to deal with the resulting problems is decreasing. We need to solve the problems now.”

“We don’t have to reinvent the wheel to do this”, he continues. “The mainstream software sector and other industries have already done much of the work, with models such as supply chain management, process capability maturity, activity-based cost management, and the service quality gap model. However these do need to be applied in practice and managers must focus on the big picture. Anything else is simply an excuse for doing things badly”.

---

Deborah Fry
Fry & Bonthrone Partnerschaft

Rochusplatz 10
55252 Mainz-Kastel
Germany

Tel +49 6134 22504
Fax +49 6134 22860
E-mail deborah@mainz.netsurf.de