Designing on Steroids—Using Insanely Robust Information Plans to Improve the Efficiency and Productivity of Offshore Outsource Writing
Many companies are driving information development departments to lower their cost structure by using outsource vendors, offshore employees, or outsource offshore workers. The financial goal behind this initiative is to lower headcount run rate, reduce project cost, and cut unit costs. These drivers are being added to the on-going historical challenges of maintaining schedules under increasing levels of feature and product churn while maintaining, or improving, quality metrics. Better, faster, cheaper!
More than a few who have ventured into the complexities of the global resource pool have experienced increased levels of rework, increased editing, more project management, and higher levels of risk. One way to mitigate the risk and minimize the rework is to produce extremely robust Information Plans during design so that the handoffs are smoother, more effective, and more efficient. Like a track and field relay race, getting off to a good start and clean handoffs are two key areas that can put outcomes in jeopardy.
Why are so many companies putting information development projects (and departments) in jeopardy with high-risk demands? One factor is stock market investor expectations for corporate growth. Investors insist that share prices increase over time. When the economy is weak, excessive supply and insufficient demand strangulate corporate revenue, making financikal and market share growth unlikely drivers to increase the top line of the balance sheet. Without a revenue driver, cutting expenses at the bottom of the balance sheet is the only viable strategy left to maintain profit and drive up share prices. Since loaded (salary plus benefits) labor costs are the biggest expense for most organizations, layoffs and offshore outsourcing have become common strategies to increase profit and share price.
For information development departments, these bottom-line, expense-driven initiatives are particularly painful. Bottom line costs are the easiest to manage and measure. Information development is often considered one of the “overhead” items in organizations and becomes a target for cost reduction. Even alignment with engineering is not a safety net as software and hardware development, IT, and call centers have all become targets for outsourcing. Many information development organizations are now under siege to find ways to add increasing levels of value while cutting headcount run rate and unit costs.
Information managers have several strategies to address this kind of disruptive innovation. Reactive managers may wait until the storm hits their department, hoping against hope that they can escape. Proactive managers anticipate the storm and develop strategies for dealing with cost challenges before they are forced into achieving unrealistic, aggressive targets. Managing a proactive change process can be easier when continuous improvement models like Six Sigma are employed.
Six Sigma is a process innovation model that helps map the change process and provides a series of tools under which the effectiveness of change can be measured in more objective ways. The DMAIC model of Six Sigma is a structured approach to improve an existing process. Other models within Six Sigma are designed for new processes. In the Six Sigma DMAIC model, the letters stand for the five phases of process change: Define what is important; Measure what you defined as important; Analyze the process for root cause problems and areas of improvement; Improve the process to accommodate the change; and finally, Control the process to ensure consistency and sustainability.
One telecommunications company used the Six Sigma model to pilot two offshore outsource models. The following is an overview of that process. The goal was to see if there was a way to use offshore outsourcing to reduce cost while maintaining process, schedule, and quality standards. The DMAIC model was used as a framework.
The Definition phase involved brainstorming possible solutions to meet the market driver of lowered cost, which included reducing scope, reducing content, reducing cycle time, reducing labor cost, or reducing quality. Reducing labor cost was the only option that realized sufficiently great numbers to meet the anticipated financial need.
To reduce labor cost, four possible development cost reduction strategies were investigated: outsourcing the entire department, outsourcing an entire product, outsourcing individual documents, and outsourcing one part or phase of the development cycle across many documents.
Outsourcing the entire department was deemed an excessively high risk. Moreover, conversations with people who had tried this approach lowered its rank to the bottom of the list. Outsourcing an entire product was identified as one of the pilot programs. It mirrored the work being done in engineering and at the time seemed a good candidate for success. Outsourcing individual documents was strongly considered, and this approach did end up flavoring the second pilot, but quality control was a strong risk as a single solution. The fourth approach, outsourcing part of the development cycle, was the model chosen for the second pilot.
By outsourcing part of the development cycle, the entire process and quality measures could be controlled by the in-house staff. The handoff point between Design and Implementation was selected as the separation point between in-house and outsource. This horizontal splitting of the work by development phase allowed the higher value work and intellectual capital to remain within the company. By owning the voice of the customer, analysis, and design, the necessary elements for success were more controllable, reducing the overall risk for success.
Adopting this approach revealed a serious impediment to successfully using this model-writers were not performing the full analysis and design work on existing projects. To properly hand off work that was designed in-house to offshore outsource companies, much more robust Information Plans were required. In fact, it quickly became clear that insanely robust design documents were going to be necessary to ensure success.
One key lesson learned from this phase-splitting approach was that maintaining the intellectual capital and voice of the customer was essential to long-term success. Both
company writers located offshore and offshore vendors were essentially clueless about meeting the needs of the customer, what the customer wanted, or even who the customer was.
The offshore people were, however, extremely motivated to meet clearly defined requirements. This cultural shift meant that a more “contract like” mentality was required to ensure a successful outcome. When defined in excruciating detail, requirements were nearly always met, and in many cases, exceeded. When the requirements were not defined with legal, vice-grip specificity, the results were consistently below managers’ expectations.
During the first pilot, when an entire project was given to an outsource company, the initial results were very positive. However, as the project wore on, there were continual and escalating issues with quality, scope of content, and format and style, which evaporated much of the cost savings. Since the project was entirely handled by the outsource vendor, inside visibility of issues became increasingly more difficult to see.
The worst aspect of the total project outsource model was the near complete “outsourcing” of the intellectual capital and expertise. In-house knowledge about the products whose documentation was entirely outsourced became a serious problem for the department. Also, morale for some information developers was affected because the outsourced products were in areas that they had planned on leveraging in their careers and now that development avenue of product expertise was gone. This model was declared a non-option well before the end of the pilot.
The phase-splitting approach across documents yielded greater benefit once the detailed design definition requirement was understood. After excruciating amounts of detail were added to the statements of work, Information Plans, and other design documents, the offshore outsource writers were able to perform as expected. This kind of detail required extreme specificity on what to do, how to do it, who will do it, and when they will do it. Every handoff was defined as a performance specification with detailed, specific, and measurable acceptance criteria. Single points of contact were established and weekly progress reports against the plan were required. More detailed process controls had to be established to identify any process defects. Frequent content inspections were necessary and more substantial editing was required. Finally, a formal, cascading escalation process was created, including quarterly reviews by the senior management.
The robust design documents included a detailed feature-to-document matrix, with complete mapping of source materials and SMEs. Backup and escalation criteria with event triggers and escalating owner name were required. Roles and responsibilities needed to be defined in greater detail and with more clarity than had been the previous practice. SMEs had to be informed of the review and interview changes caused by time zone differences, cultural differences, and a radically different kind of initiative than they had expected. Extremely detailed risk and mitigation was required to identify potential blockers with pre-defined mitigation strategies. Risk assessment and mitigation were beefed up to predict every possible problem. Each risk was associated with an agreed-to mitigation with measurable trigger criteria. SMEs, product managers, project managers, and documentation managers were involved in the more robust agreement.
Levels of service were added to define the kind of support to which everyone agreed. The seven most common process defects identified as a level of service included feature churn, scope churn, product stability, SME changes, testing changes, engineering changes after handoff, and deliverable changes. Measurable defect (churn) levels for each process area were identified and mitigated with a pre-defined strategy.
The assessment of work was broken down to the topic level, with linkage between each feature. As funding changed on a feature basis, the associated topics and resulting offshore statement of work could also be changed.
Each work assignment became a “kit” from which offshore information developers could build the deliverables.
Ongoing challenges stressed the resource managers’ time. Excessive amounts of effort were required to ensure that all resources were on the same page. Project managers needed to “see” process evidentials to ensure that everyone was following the process.
At the end of the two pilots, the essential goals were met: headcount run rate was lowered, total project cost was reduced, unit cost was cut, and overall quality was maintained. By leveraging the high-value-add work with badged employees, no one was laid off from the company due to offshore outsourcing. This statistic was due in part to an expanding work portfolio.
On the negative side, resource management was a nightmare. At an outsource rate of about 20 percent, the existing resource managers were totally maxed out. If further outsourcing would be required, more resource managers would need to be hired. The same would apply to project managers.
Another lesson learned was the importance of process maturity. The documentation department was recently assessed at Level 3 process maturity. The successful outcomes achieved would not have been possible had the organization been just one level lower. Knowing, following, and believing in a workable process was critical to the outcome. Process controls were essential to ensure a successful outcome.
What was not factored into the financial equation was the SME. The old SME behaviors of “when I get around to it” and “it’s in the spec” went out the door with the offshore work. SMEs were forced to spend more time answering questions and handling interactions in a more timely (and measured) manner.
One other interesting finding related to the differences between the two pilots. One pilot used new offshore company employees and the other used an offshore vendor who specialized in training and documentation. During the initial months of the project, the badged employees fell behind the vendor in ramping up to speed. However, after four months, the outsource vendor had reached a plateau while the badged employees were continuing to improve their performance. After six months, the badged employees were outperforming the vendor’s employees.
The key lesson learned from these pilot programs was to make sure that insanely robust Information Plans are used for all offshore, outsource work. Assume nothing. Measure everything. Inspection is more effective than expectation. Finally, include all costs when assessing savings-the total financial benefit may be smaller than you think. Low cost centers certainly have a place in a resource portfolio. Make sure you know everything about those places before putting your customers at risk.
About the Author