Friday, July 25, 2014

High Tech People Management in Lean Times

To THINK BIG and to use our talent doesn't mean we don't have difficulties on the way. We will - we all do. How we view those problems determines how we end up. - Ben Carson

For several weeks now, I’ve been blogging about the coming Lean Times for the High Technology market. Even smartphones and tablets are peaking, while all other markets (with the exception of the clearly declining PC sector) are already essentially flat. High Tech’s escape from this emerging state of stagnation will depend in great part on revolutionary technologies currently being researched in its semiconductor sector – technologies intended to overcome the end of Moore’s Law in silicon and open up untapped applications and markets.

In the meantime, semiconductor companies will need to find a way to weather this drought themselves so as to be prepared to come charging out of the gate when the Next Big Thing presents itself. It will take some extraordinary efforts and preparations, including breaking away from some of the industry’s common methods, practices and operational procedures which had proved sufficient in the past.

Though certain old routines will need to be abandoned, there are other more fundamental practices which the chip industry will need to re-learn. One of the most vital lessons is a very old one indeed, and is effectively illustrated by a famous story out of the Old Testament.

Joseph and Pharaoh

Joseph was the son of a Shepherd in the very ancient Levant. Having demonstrated a knack for prophecy, Joseph became both a favored son of his father and an envied sibling of his brothers, who began to perceive him as a long term threat to their inheritance. Consequently, Joseph’s brothers kidnapped him and sold him as a slave to a merchant caravan on its way to Egypt.

Thru no fault of his own, Joseph got in hot water again in Egypt and was thrown in prison. While incarcerated, he made the acquaintance of one of Pharaoh’s previously favored servants and left a good impression on the man. The servant was eventually restored to Pharaoh’s favor and one day heard a disturbing story from Pharaoh himself.

Apparently Egypt’s potentate had been having recurring nightmares – ones where seven starving cows emerged from the Nile to eat seven fat cows and seven withered ears of corn devoured seven fat ones. The servant recalled the man he had done time with in the slammer and mentioned to Pharaoh the wisdom and insight that Joseph had demonstrated. Pharaoh’s dreams were unsettling enough that he took the unprecedented step of having the prisoner brought before him. Upon recounting these dreams to him, Joseph interpreted the nightmares as prophetic visions – that Egypt was about to enter a period with seven years of bounty, followed by seven years of drought and want.

Pharaoh was deeply impressed with the seer and his forecast, designating Joseph as his right hand man and assigning him the responsibility of preparing Egypt for the coming lean times. Joseph was so adroit in his planning and management of resources that not only did Egypt pass thru its long drought successfully, it became a critical source of grains and crops to its neighbors, thus vastly expanding Pharaoh’s power and influence abroad.

The moral of the story is rather straightforward: how you manage your resources in times of great stress will make the difference between your ship foundering in the storm or sailing through and emerging in a position of strength.

So how does this apply to the semiconductor industry? McKinsey penned a report that concisely captures the difficulties currently stressing the chip sector:

In an article entitled “What happens when chip-design complexity outpaces development productivity?”, McKinsey makes the following salient points:

-         3P (price, performance, power) competition continues to intensify across all chip markets.
-         New product development is assessed by complexity, TTM goals, budget constraints, value-add IP, reused IP, outsourced IP and embedded SW requirements.
-         More than ever, projects need to be predictable and dependably meet schedule.
-         Meeting TTM critically affects revenue and margins.
-         Chip firms will be continually pressured to perform more work with the same or less resources.
-          IP design reuse has already surpassed the 2/3 threshold.

With Moore’s Law having almost run its course, the ability of chip firms to add features and compete on value is giving way to competition increasingly based on price. As a result, any value-add that can be shoehorned into a chip design will have a short lifespan and therefore needs to reach the market quickly.


But where does that value come from? Some would say it comes from a company’s skill at integrating functionality; others would say it depends on the firm’s native IP. Neither is correct, however. In the final analysis, it comes from the chip firm’s PEOPLE. Their collective enthusiasm, experience, creativity, dedication, specialized expertise and energy, combined with all their sweat, strain and effort constitutes the sum total of the company’s value-add.


Unfortunately, this collection of resources also has to bear all the damaging side effects of stagnating end user markets. There are no extra funds to hire the additional personnel that are so desperately needed. Consequently, already overtaxed organizations are expected to continue delivering more of everything, more frequently and in a shorter time frame.


Difficult times separate good management from bad. The poor management teams will find their organizations producing less value-adding features and benefits, slipping schedules with greater frequency and with larger milestone misses, suffering more unforced errors and making the kinds of disastrous mistakes that cause chip companies to miss entire market cycles. These firms will also suffer more absenteeism, health problems amongst employees, internal friction and attrition. Such companies simply will not survive these lean times.



Picks, Shovels and Plum Lines


The most satisfactory definition of man from the scientific point of view is probably Man the Tool-maker. - Kenneth Oakley

Despite some protestations to the contrary, the great majority of semiconductor firms have implemented some sort of ERP (Enterprise Resource Planning) system without a whole lot of thought, choosing cost, expediency and ease of use over any deeper considerations. They pull together one or more tools which allow them to create a schedule that tracks tasks and major dependencies, permits sharing a Gantt chart with other team members, and helps the finance department to do some budget calculations and cost tracking. Once the estimates for scheduling and cost are ready and team members are expediently selected and assigned, executive management will command that the timeline be reduced by twenty percent and the team – after using some choice words to express their opinion of management – will suck it up and do their damnedest to build the product to schedule.

Let me be brutally candid. This process has been and always will be a stupid way to run an enterprise. In even the best of companies, products that are planned in this manner will either be late, missing significant features, or more often both.

Quite a few chip companies use MS Project as the centerpiece for their product planning and development efforts. To be fair, MS Project is a good tool for scheduling. However, it is often found wanting as a full project management tool, as developing a schedule is not a particularly difficult thing to do. The missing components of a project plan are often handled with a mix of other tools – MS Excel, Sharepoint and sometimes a custom utility or two for handling costs, collaboration and other issues related to group projects.

A few established software companies have seen an opportunity in this hodgepodge selection of project planning tools. Solutions such as Oracle Primavera and HP PPM integrate such functions and add a few bells and whistles of their own in an attempt to make enterprise project planning software a one-stop-shop experience.

Nevertheless, despite the herculean exertions of experienced, talented and specialized team members, an EE Times study from 2012 indicates that 60% of projects fail to meet their revised schedules. This is because none of the current popular solutions – whether home grown by the firm or provided by some established enterprise software vendor – addresses product/project planning and management in its entirety. Some solutions focus on identifying and resolving bottlenecks, others the financial reporting part, but none of the current approaches are truly holistic and complete.


A moment's insight is sometimes worth a life's experience. - Thomas Fuller

In High Tech, crisis begets creativity. So it is with the worsening semiconductor sector struggle with product development & project management – a number of fledgling firms are developing and bringing to market software solutions that capture a much larger dataset and help chip firms manage their resources and development efforts far more intelligently and comprehensively. Just as server and network virtualization have arisen as software products that save companies significant money and help them manage their current networking and datacenter resources much more effectively and efficiently, so has a new “Talent Management ” sector of the software industry emerged to maximize the potential of the most vital resources a chip company has – its people.

Just like many of the advanced software tools developed for the enterprise today, the most sophisticated of the new talent management software offerings are built on a foundational database. The database captures previous project histories in comprehensive detail. In particular, the database reflects what it really took to get a given chip firm’s projects successfully completed in the past, with details on people, task completion times and materials. Consequently, the business model for Talent Management tools companies that provide such databases is by necessity bifurcated, as the first stage of the engagement is perforce implemented as a service.

Once the database is populated with information, the tools can generate what is referred to as an ‘unconstrained demand’ for a new project. Stated differently, the talent management software identifies what a given program’s composition, requirements, tasks and schedules are in detail, including design, simulation, verification, layout, test insertion, fab turnaround, assembly, test and everything in between. The building of the database plays a particularly critical role here in that it helps to precisely identify individuals with the right skills and matches them to a required time commitment.  In the end, this leads to a more reliable cost structure, stated in familiar terms of resources, time and materials. An example of such a flow combining services and software is offered below in Figure 1.

Figure 1 – combined SAAS/tools flow for Talent Management (image courtesy of eDevelopment Corp.)

Effective talent management software solutions need to make it easy for the user to identify the entire range of obstacles to successful and timely project completion, such as sources of friction, gaps in abilities, overstress, resource bottlenecks, dependencies and other organizational issues for product development. This leads to an organizational understanding that helps project execution go smoother, faster, more efficiently & effectively, resulting in higher success rates.

There are secondary benefits as well. Defining and executing projects successfully also helps with retention (ensuring star employees don’t burn out and people in general don’t reach their breaking point), productivity (people are more efficient when they use their talents effectively, find work rewarding and have a chance to recover their strength) and recruitment (a firm will know exactly what sorts of specific skills and talents it needs to find for which tasks and to fill very clearly defined gaps.) The end result is optimization of design cycle time, cost and on-time delivery of new, value-added products.

Because of the completeness of this approach, well-designed talent management tools actually capture the total value of a given chip company – not only its current structure and personnel, but its institutional history. Such a quantification serves to not only encapsulate the fullness of a semiconductor firm’s value proposition as  a function of its people, but also reveals defects in the organization or empty spots in the required talent pool. As such, the software serves as a total solution for enterprise management, including recruitment, performance management, training & development and compensation management.

Darwinian Struggle

Accounting and Finance departments normally resist such tool investments – after all, they’re more expensive than the patchwork of software most chip companies currently use. Going forward for the next several years, though, managing one’s resources well will become a decisive competitive factor. Enterprises that hope to survive the Lean times and be well prepared when a healthy economic climate returns cannot simply focus on their balance sheet and watch fixed costs to control quarterly financial results. As prices broadly erode in semiconductor markets over the next several years, any chip company resorting to RIFs to ‘square’ their balance sheet will be cutting into muscle and bone at this point, with a consequent deterioration in operational effectiveness  – they will, in other words, lose even more competitive ground.

The trick to long term success will be to grow in a lean market. This means using one’s organizational resources extremely well – with high productivity, high efficiency, and most importantly, achieving effective output while doing more with less. To survive and even grow in Lean times, a company must manage its resources skillfully to more swiftly develop new products of superior value. That means having the right tools to do it. Some vendors offer weak tools and others very strong ones. Chip companies will need to choose wisely, as it will make all the difference.