Investigating and adopting a PLM solution

Deciding on a PLM solution is no different from looking at other business productivity applications, and can be as simple or as detailed as you want to make it.

Whether you spend a few hours or several months, you'll probably want to consider the following topics:

1. Define your Project Objectives:

Understand your project's strategic objectives. That is, answer the basic question: Why do you need a PLM solution?

Your goal is to...	PLM can help by...	Measure PLM impact by...
Increase product quality	Ensuring affected users consistently review product changes; automatically notifying users about new & revised documentation; tracking that obsolete data is promptly removed from production areas; checking that as-built products match approved design and process documentation	Monitoring change-related defects; periodically auditing production floor documentation; performing periodic audits of as-built product
Manage increased product complexity	Encouraging on-going collaboration and review of product design and process files; providing consistent structure to your bill of material management; enforcing product change review using defined workflow	Categorizing reasons for changes (principally those related to failure to meet design requirements)
Reduce product unit costs	Encouraging part re-use by easier searches, and thereby increasing purchase volume	Tracking reductions in new part numbers that are issued for each new product
Reduce time-to-market	Reducing number of items designed through re-use; reducing design iterations via concurrent review; speeding the design release process though automated workflow and notifications	Measuring average project development time; monitoring defect frequency immediately after new product launch
Reduce change-related administrative overhead	Eliminating manual change processing, and carrying physical change packages from one approver to the next; automating materials declarations for environmental compliance	Demonstrating change in staff throughput, such as number of changes processed per unit time; eliminating manual processes such as copying, faxing, shipping, "walking changes through system"
Reduce production rework and inventory scrap	Speeding changes through the review and approval cycle; automatically calculating cost impact of changes based on item-by-item disposition expense	Monitoring rework and scrap expenses; tracking total cost impact of changes
Improve customer satisfaction and enhance loyalty	Executing quickly, consistently and predictably	Asking customers and sales staff for feedback
Plug gaps in current business processes	Encouraging business process designers and users to seek more efficient methods, and capture these methods in well-defined PLM attributes and workflow	Assessing employee satisfaction; eliminating tedious activities
Enhance cooperation with your supply chain partners	Encouraging early review of requirements; exchanging design, procurement, production, and service information	Tracking rework charges and non-recurring tooling and setup costs

2. Assess your organization's existing processes for design, data management & change:

Examine your organization's existing design, data management & change processes. Your alternatives are to:

• Do nothing: You may decide that your processes are sufficient for current business needs; you don't anticipate any worthwhile efficiencies by modifying or automating your existing manual processes.
• Adopt PLM without changing your process: This is a reasonable path when you're under significant resource (time, money, staff) constraints, but believe that immediate automation of key areas can provide immediate benefits or can control an on-going problem (such as chronic bill of materials errors or slow change processing). You'll usually be able to tweak a few of your most challenging process pressure points under the flag of "system configuration".
• Re-engineer processes without PLM: Another possibility when you don't have the time or budget to do everything, but believe that your processes are too flawed to automate efficiently. This should make a subsequent PLM project much easier to implement.
• Re-engineer processes as part of a PLM project: The most challenging approach. You'll need to spend considerable time negotiating cultural changes, specifying PLM business rules and workflows, and trying to reconcile the demands of users who like things just as they are with those users demanding radical change. This approach requires a particularly strong champion with enough political clout to get everyone lined up behind the effort. But it can offer a radical transformation in your organization's efficiency, and the risks may well be worth the rewards.

3. Set realistic Goals:

After reviewing the strategic objectives and assessing your organization's needs, set realistic goals for your PLM project

• Plan for an incremental & benefits-driven implementation; avoid the "big bang" approach if possible, and capture the quick payback before tackling tougher problems with harder-to-justify ROI
• Establish simple, easy-to-track success metrics and make sure that they're directly related to your original objectives

4. Identify potential expenses and forecast actual costs related to purchase, installation and operation:

According to AMR Research, in 2004 software license fees represented about 38% of total PLM industry revenue.

Typically, most of the implementation, consulting and training costs are incurred at the beginning of a PLM project, so you may find that these expenses consume more than half of your initial project budget.

In practice, the most cost-effective methods for minimizing project expenses:

• Emphasize easy to use software   Ensure that your installation, configuration and training costs are as low as possible by choosing the simplest, most user-friendly software that meets your needs. Most reputable vendors will let you try before you buy.
• PLM will be a big win, so don't over-buy  Minimize configuration consulting and custom development work, principally through avoiding enterprise integration projects that yield only minor productivity improvements.

Actual Costs:

There are many PLM solutions available. Some vendors will require an extensive hardware, software and services investment; others are more sensitive to a limited budget, and will try to work with existing hardware, low-cost (or free) databases, and minimize the use of consultants.

This list suggests some possible expenses to explore:

• Staff time to research & evaluate alternatives
• Software solution costs
  • per-user or concurrent-use client licenses
  • per-CPU (or other) application license
  • database license
  • special administrator licenses
  • annual maintenance, support, upgrades
• Hardware costs
  • workstations
  • application and/or database servers
  • backup devices
• Special file viewers
• Downstream ERP file import or integration
• Network & bandwidth costs
• Deployment costs
  • Consulting services for process reengineering and application configuration
  • Legacy data import and migration
  • Pilot (proof of concept)
  • Production deployment

5. Review how software licensing can impact your budget and operations:

Before contacting vendors, consider the operational and financial aspects of software licensing models carefully. Some vendors offer more than one licensing approach. Decide what will work best for you and make your list accordingly.

Traditional perpetual license, with annual maintenance

If you have the budget, this path is usually the simplest to manage as well as cheapest over the long run because you have a definite control over your costs. The alternative license methods all represent an on-going cost, and may bundle financing costs into their monthly charges.

Perpetual licenses, as well as subscription licenses, let you tailor the configuration to your exact needs and even customize the product and/or its interfaces. You have complete control over performance, storage costs, maintenance and service interruptions, and - most important - your critical product data. Furthermore, if you're perfectly satisfied with the system that you have, the maintenance can usually be stopped.

Subscription or lease

This is typically like a magazine subscription where you can buy for any number of users, and can stop at the end of the subscription period. It's an inexpensive way to get into PLM, since there are no large upfront software costs, except (if needed) a database server license. You still get the operational benefits of a perpetual license: high performance, typically a superior user interface, in-house control of your critical product data, ability to tightly integrate your other systems. Subscription licenses are sometimes preferred because they can be dropped if, for instance, you only required a large number of licenses for a temporary project, or you're concerned about unaccounted-for licenses installed on obsolete PCs. Some subscription models allow you to convert to perpetual licenses after a certain period ("lease to own").

Internet ASP-based services

A PLM application service provider (ASP) hosts the entire PLM system off-site, using a data center shared among all of its customers. A PLM ASP perhaps offers the most financially attractive short-term PLM solution but comes with significant longer-term risks. The per-user cost is initially reasonable if you aren't particularly concerned with the total cost over 5 or 10 years.

The principal benefit of a PLM ASP is that the vendor handles the hardware and software management tasks: acquisition, installation, maintenance, and upgrades. Furthermore, the ASP defines a rather specific set of features that simplifies your configuration choices. Finally, if you don't want to bring your legacy data into the system, and can quickly make your configuration choices, you can have access to your PLM system in a matter of hours.

As you'll see, we have reservations about a PLM ASP solution; a hosted application may be perfect for incidental business processes (think "travel reservations" or "HR benefits"), but it's a much tougher decision when applied to mission-critical functions and irreplaceable proprietary data.

If limited budget or internal IT resources suggest a PLM ASP, you'll want to weigh these issues:

• Your company's mission-critical data is not on site. This has two effects:
  • How much do you know about the security, backup process, operations skill, disaster recovery procedures, and financial stability of your PLM ASP, and
  • What happens to your productivity if your Internet connection or PLM ASP goes down for a day?
• A PLM ASP solution typically implements a "one size fits most" approach, and is much more limited in its configurability; make sure your business processes won't be constrained.
• Usability will be limited by the browser-based user interface, and web technology simply isn't as flexible and as powerful as native applications. Can you imagine CAD or even a spreadsheet hosted on the web?
• Performance is determined by your Internet connection's available bit rate and utilization, your vendor's available bit rate and utilization, the load on datacenter computers, and the systems selected for the datacenter. If you aren't happy with the system performance, upgrading will require more analysis and negotiating. There's no guarantee that you'll get the performance you need if your requirements exceed your vendor's abilities.
• There may be additional charges for exceeding contract limits for data storage, processing time, or connection bandwidth.
• Upgrades and service downtimes are scheduled for the vendor's convenience, not yours. Except during project "crunch time", this should not be a significant issue if your office runs inside the typical workweek, especially if you select a PLM ASP in your time zone.
• In most cases, all of the PLM ASP's clients share a common code base, so you can't delay or skip an upgrade. Whether you're in the middle of a time-critical project or simply satisfied with your software, you'll still be getting your new upgrade (and possibly re-training, data conversion, and new bugs) with everyone else.
• If you're not happy with your PLM ASP, check your service contract and practice your diplomatic skills before asking for any help in moving your data to a competing PLM system. 

A PLM application service provider may be the only reasonable alternative for your company, but you'll want to anticipate all of the ways the arrangement could fail, and make appropriate contingency plans.

6. Establish a budget and define expected results:

Establish a budget and figure the return on investment (ROI)

• Control costs at each stage of the project; savings accrue at each step as you simplify your requirements
• If you are having trouble with justifying ROI
  • Scale back project scope
  • Look for simpler, more affordable solutions
  • Lease rather than buy

If you choose to concentrate on the "low-hanging fruit", ensure that this short-term focus supports a flexible, well-rounded PLM Solution and is not a dead end.

7. Determine the selection criteria that you'll use :

There's good news: Almost any broad-based PLM solution that can support your small or medium business (SMB) will be better than the manual process you have now. To simplify your selection process, consider these as necessary (and perhaps sufficient) points to start your examination.

Focus on your needs, not the supplier's capabilities

Make sure that you align your PLM supplier's capabilities to your goals. With all of the choices available, it's easy to get distracted. But there's not much return on investment (ROI) in buying a CAD file management system if you're mostly looking to accelerate your change process or reduce your bill of materials errors.

Protect yourself from fatal commitments  

Your organization will be living with your decision for awhile. A low upfront cost may be essential to get the project funded, but it's the on-going cost of day-to-day operation, user productivity, training, administration, and support and maintenance that will determine success in the long run. 

However, don't worry that your PLM decision is forever. Once your data is in any modern PLM database, shifting to another PLM system ought to be easy provided that:

• your PLM license vendor uses a well-known database server (SQL Server, Oracle, DB2, etc.); or 
• your PLM ASP provides a written guarantee that all data they manage for you can be extracted easily, and will demonstrate this capability whenever you request it. 

Generic user interfaces are, well, generic

Many PLM systems offer both browser-based Java clients and OS-specific (typically Microsoft Windows) clients. Where you have a choice, align the client software with the types of users you'll be supporting. 

For example, compared to a dedicated Windows client, web browsers still have relatively slow and weak user interfaces. Enabling complex PLM UIs in a web browser requires heavy-duty Java and custom plug-ins, which burden the web client while compromising the run-anywhere thin-client advantage. 

Most users will prefer the higher performance, information-rich OS-optimized interface if they have the choice. 

However, web clients are appropriate for casual or guest users (such as supply chain partners) who just need to browse data, for less-popular operating systems (e.g., Mac OS X), and for mobile users who must look up data or approve a change while out of the office.

PLM is great, but don't get in over your head

Surprisingly, for smaller companies, CAD and ERP integrations are often "nice to have", not essential. Why? Because there's a lot of ROI available in just data and file management, bill of materials construction, automated change control, and workflow/email notification. In smaller companies, integrating PLM with CAD/ERP may only represent 10% of total PLM financial benefits, yet cost 40% or more of the total project budget. Projects that require custom integrations can be time-consuming and can delay project payback by months (or worse). Quick PLM implementations rely on simple data exports for loading the ERP BOM tables; your CAD and ERP processes will be no worse than they are now, and you'll see the PLM benefits much sooner. Only after you've extracted all other PLM benefits should you then examine system integrations.

Courtesy - www.Product-Lifecycle-Managment.com