This is intended to be a living document. We need sales’ help validating this messaging. Does it play well? We will gladly incorporate changes in future versions of this document in the spirit of continuous improvement. Thank you!
AUDIENCE: Primarily engineers and scientists, but increasingly, others who want to learn from data faster and are interested in a more scientific approach.
CRITICAL BUSINESS ISSUE: Time-consuming and costly experimentation.
GOAL: Realize improvements faster with repeatable methods.
TALK TRACK SECTIONS:
DEFINING THE PROBLEM
JMP'S SOLUTION TO THE PROBLEM
SALES CYCLE GUIDANCE
CAPABILITIES MAP
OPEN ALLCLOSE ALL
DEFINING THE PROBLEM:
Customer Issue Description
- Experimentation can be costly and time-consuming. To get better products and services to market more quickly and with fewer resources, companies need to become more efficient. To use research and government funds more efficiently and effectively, researchers and scientists need to be smart about experimentation
- Scientists, engineers and increasingly, other personas are responsible for conducting experiments to understand which parts of a process are working–and which are not. Unfortunately, error is ubiquitous, making it difficult to separate the signal from the noise to draw solid conclusions. Experimenters must test many factors to confidently identify the vital few that have a substantial impact often on many responses. Should a process or product need to change midstream, the experimenters must quickly redirect resources and all of this must happen quickly to meet shortening deadlines.
- Actively manipulating factors according to a prespecified design is the best way to gain useful, new understanding about the relationship between an input / factor and an output / response. Whenever there is more than one factor and many responses–that is, in almost all real-world situations–a design that changes just one factor at a time is inefficient. Design of experiments (DOE) is a practical and ubiquitous approach to properly uncover how factors jointly affect a response.
Business Impact of Not Addressing the Issue
- Profits are decreased when time and money are invested in irrelevant factors. Example: You do an experiment on Monday with a 50 percent yield. Tuesday’s experiment yields 55 percent, so you ask for capital to imlement this process. On Wednesday, after implementation, you obtain 50 percent yield. You then realize that the error of the experiment is 5 percent, and that you’ve spent precious capital based on the wrong conclusion.
- Important interactions are missed with conventional experiments that only change one factor at a time. Example: The effect of temperature may be significant, but may not manifest itself if the concentration, (another input), is different in the experiment.
- Time to market suffers when deadlines are extended because of unpredictable processes.
- Ability to compete effectively, be on time and within budget is reduced due to suboptimal learning.
External Validation of Not Addressing the Problem
- From the Atotech success story, “The big benefit is time savings. You have to be responsive in the electronics and semiconductor industry, which is quite fast-paced and competitive. JMP lets us do that.”
- From the Eastman Chemical success story, “Design of experiments is the bulk of what we do. We’re using it in R&D and manufacturing almost daily. We do a lot of traditional designs like fractional factorials, but we also rely heavily on the JMP Custom Design platform, because not everything is that simple.”
- The 2009 Brumbaugh Award from the American Society for Quality (ASQ) was awarded to Bradley Jones of the JMP division of SAS and Chris Nachtsheim of the University of Minnesota for split-plot designs. This award is given to the most significant single contribution to the development of industrial applications of quality control in the previous year.
- The 2012 Brumbaugh Award from ASQ was also awarded to Jones and Nachtsheim for their development of the new definitive screening designs.
- The American Statistical Association’s 2012 Statistics in Chemistry Award was given to Bradley Jones and Scott Allen of Novomer Inc for their work with definitive screening designs on a significant Department of Energy recognized and funded project for reducing carbon dioxide emissions.
Challenges/Barriers to Addressing the Issues
- Departmental silos. DOE works best when there is collaboration across groups, but many organizations lack the tools to share information easily.
- Lack of management support. Without commitment from managers, it’s difficult to implement a disciplined approach to experimentation and make that be part of the organizational culture.
- Status quo is fine. Some organizations are profiting and do not feel the pain of their inefficient, wasteful methods – yet.
- Designs that don’t fit. Many organizations force problems into designs that don’t fit well. They don’t know how easy it is to build custom designs, which would allow them to experiment more efficiently and gain maximum information.
- We tried that approach and that did not work for us. Some organizations have been unsuccessful in their previous attempts to implement DOE and it is usually not the approach that failed but some other reason for the failure such as too much noise in their system for a number of reasons such as raw material variation or analytical methods etc.
- Ignorance of new methods. Traditional approaches may leave out important factors, leading to costly follow-up experiments. New methods save time and money by identifying important factors the first time around.
JMP'S SOLUTION TO THE PROBLEM:
Our Approach and Value Proposition
While methodical experimentation is the best approach to gain a new and useful understanding of your products and processes, changing one factor at a time is a waste of precious time and resources. DOE is the correct alternative. JMP offers world-class capabilities for DOE and analysis in a form you can easily use. Whether or not you are a statistician, JMP–with its combination of interactive graphics and robust analysis–makes statistical experimentation accessible. With JMP you can:
- Explore multifactor opportunities that exist in nearly all real-world situations.
- Understand complex problems with both classic and custom designs.
- Tailor your design to answer specific questions without wasting precious resources.
- Optimize and co-optimize all of your responses.
- Effectively communicate your findings with colleagues.
- Specify your inputs and outputs – as well as the number of runs that you can afford – and let JMP provide the plan.
- Easily see the pattern of response, identify active factors and optimize responses.
- Enable the change management process with data-driven experimentation.
Differentiating Capabilities
- R&D/intellectual property. SAS co-founder and JMP chief architect John Sall is committed to developing this product and has enlisted the help of premier thought leaders to improve the product. The brains behind this tool are recognized in the industry and world-class.
- Unique design capabilities. Custom designs are proprietary to JMP. If you need to create an experimental design that takes into account specific parameters such as time, budget and other considerations, the Custom Designer constructs a design to address your specific problem (using an optimal design), so you don’t have to force your problem to fit a prescribed textbook design. Its flexibility allows you to specify numeric, categorical and mixture factors. Custom, classical, split-plot and other designs – JMP has them all.
- Continuous innovation. As of JMP 11, JMP has the first commercial, quality-assured implementation of the new award-winning definitive screening design. This new class of design for quantitative factors reveals more information in fewer runs because it has three levels (versus the traditional two-level screening designs), often allowing elucidation of polynomial terms which is a breakthrough for screening designs!
- Subsequent analysis. With JMP you can model, simulate and optimize your experiments.
- Personalization and extensibility. Buying software that cannot grow beyond your initial needs is shortsighted and expensive. JMP includes a diverse set of basic and advanced tools that let you customize the software to address unique problems that change as your organization evolves. With a greater level of customization and automation than other software, JMP is designed to make the user productive.
JMP 15
- Group Orthogonal Supersaturated Design is a new platform for creating and analyzing supersaturated designs (more factors than runs), having groups of correlated effects. Should confounding persist after the initial analysis, the factors involved are identified and an appropriate augmentation scheme (requiring only a few runs) is advised.
- Instantaneous creation of a globally optimal design when runs are multiple of four and all factors are two-level continuous or categorical.
JMP Pro 15
- Custom Designer provides modeling scripts for Generalized Regression and Fit Mixed.
- Other JMP Pro 15 enhancements can be leveraged as part of the experiment workflow.
JMP Live
If they need to share information in an interactive way with others that scales (not in Powerpoint, Excel, emails, etc.), make JMP reports available to non-JMP users, publish reports that are created automatically, have a repository for project data and reports, deploy dashboards, or create a shared space to foster collaboration, JMP Live may be an option.
SALES CYCLE GUIDANCE
Prospecting Criteria
JMP should appeal to any organization seeking to learn more from data faster. Traditional industries include high tech, pharma (big due to QbD), chemical. New opportunities are in marketing, retail, financial services. DOE users:
- Organizations using competitor software who may be won over by JMP’s new and award-winning implementation of definitive screening designs.
- Partners, consultancies and academic institutions using JMP who consult with or collaborate with industry and may be instrumental in a commercial sale (e.g., Ornek and Ramsey presentation from Discovery Summit 2012 and Ramsey work with Lonza featured in Bioprocess International).
- Organizations using JMP and willing to help us expand usage to more scientists and engineers.
- Scientists and researchers in other industries including government agencies who may have a need for new methods like definitive screening designs.
Non-DOE users:
- Organizations who are not taking advantage of DOE, but who seek opportunities to save time and money through repeatable means.
- Scientists and researchers in other industries including government agencies who may have a need for newer methods like definitive screening designs.
Conversation Starters
- Which functional group are you in? How does your group fit into your organization?
- What is your primary goal: speed of learning, ensuring that you aren’t missing important relationships, other?
- What makes you successful in your organization?
- How fast does your company innovate?
- Do you have a good handle on how key factors relate to specific outcomes?
- Find out how important DOE is to a prospect. Do you do designs?
- Are you maximizing the amount of information gained with your experimentation process?
- Do you have to run more experiments to clear up questions that weren’t addressed in the first round? If yes, what type of designs and which software do you use? § If using classical designs, are you getting suboptimal results when you force a problem into a design that doesn’t fit?
- Are your problems forced into textbook designs?
- Do you have a culture of collaboration?
- What are you using for predictive modeling? (JMP is a comprehensive tool and offers so much more than Design Expert like custom designs, definitive screening designs and the interactive visual paradigm that is JMP.)
- Which tools do you currently use?
- Who are the primary users of the current tools?
Sales Cycle Tips
- Wherever prospects are feeling the pressure, DOE is a great tool because it offers strategic competitive advantage and fuels innovation.
- Individuals doing DOE will be recognized. Since you’re only as good as your last failure, you want to be successful and be viewed as an achiever.
- DOE can enhance your patent portfolio. Whenever you look at the literature and try to patent something, lawyers often can’t support the application. Designs can show why your idea is unique and helps make it patent-worthy. Patents are easily obtained when you can identify exceptions through the use of DOE.
- Where does gut feel fit in the equation? Gut feel is needed for the initial conception of the thought, but you have to be disciplined to test the hypothesis. Good companies prove the idea behind the gut feel. Basic principles get you so far, but the companies that excel (and the ones that get the most patents) are the ones that are data-driven and doing DOE.
- Many times, DOE results will be counterintuitive. Those special nuggets of insight often lead to a better comprehension of why the exception exists and to a deeper understanding of your business.
- Good companies (such as Intel, Dow, etc) provide additional training since they understand you need more than what you got in college. A serious investment in training is needed during the first few years of employment to bring workers up to speed statistically. JMP training and partner training are valuable.
- What about the cliché that correlation is not causation? Root cause is often assumed to be one thing, when it’s actually something else or a combination of factors. DOE allows you to examine the sensitivities of the factors (which is important for yield, quality and costs) and of changes in the factors to achieve optimization.
- If the account is a current JMP customer, determine if there are opportunities to expand usage, especially to chemists who often think “statistics is for people who don’t know science.” Chemical engineers tend to have more statistical training than chemists, but there may be opportunities to expand the use of JMP in organizations that already have it.
- If the prospect is using a competitor’s product, the capability for DOE may win them over to try JMP.
- If the prospect isn’t using any DOE software and there are no JMP users in the organization who can influence the prospect, it may be a longer haul, but a sale is still certainly possible.
- Probe for more information about the organization’s culture. The culture can indicate the potential for a single JMP sale versus many. We should always encourage our customers to have a users group to spread the use of JMP (as well as DOE) in the organization. DOE enhances collaboration.
- For classically trained DOE users who are new to or uncomfortable with custom designs, reassure them that classical designs are a subset of custom designs. One day, they may need a custom design. They can also use all the other platforms in JMP to communicate data visually, so they can present results objectively.
- Definitive screening designs are still so new that people may be unaware of them. Simply informing them about a better approach for dealing with many factors may be enough to get them to try JMP (especially if using competitor’s software).
CAPABILITIES MAP
Complete DOE Workbench
Capabilities Description:
- Custom designs help you make the best use of your experimental budget. You don’t have to force your real-world scenario to fit a textbook design.
- Classical designs and more: you have all the designs you expect, as well as other options.
Diagnostic Questions:
- Do you use experimental design as a means to efficiently and effectively gather information about important outcomes?
- Do you have all the design types you need?
- Do you have the best DOE capabilities to speed continuous learning and improvement?
- Are you able to easily compare designs to determine which best meet your needs?
So What? Value/Impact:
- If not you aren’t using designs efficiently, you may be wasting time and money.
- With DOE, you could be creating more value faster and become more confident in your experimental process.
- JMP lets you assess the best designs for your goals and constraints.
Not just more DOE, the right DOE
Capabilities Description:
- Custom designs let you tailor the design to your problem and not the other way around.
- Definitive screening designs augment JMP’s already rich DOE menu. Definitive screening designs may reveal things you didn’t know you didn’t know.
- Covering Arrays – important for complex system testing where the outcome is deterministic.
- A-Optimal designs, useful if the optimal design needs to focus more heavily on certain effects rather than others. (There isn't a natural way for D- or I-optimality to do this.)
Diagnostic Questions:
- Are you missing important information and potentially incurring unnecessary costs: time, money, materials, missed opportunities, etc.?
- Are you using the same design process you’ve always used? How do you assess its effectiveness? What if there were more efficient designs to help you create more value and be more competitive?
- Are your products and processes growing in complexity? Do you have more factors affecting your outcomes of interest?
- Is failure testing for complex systems expensive and time consuming?
So What? Value/Impact:
- Missed information may mean missed opportunities. If you don’t have a handle on the complexity of the influences on your outcomes of interest, you may be using resources inefficiently.
- With increasingly complex products and processes, as well as a growing number of factors to test, understanding the intricacies of all the factors separately and collectively as they relate to your responses of interest is vital.
- Covering arrays design in JMP Pro set world records for fewest number of runs required for a given problem. Fewer runs mean lower costs.
- A-optimal designs let you more easily focus on the effects you most care about.
