1、 New Problems, New Solutions: Making Portfolio Management More Effective WORKING PAPER No: 9 Authors: Dr. Robert G. Cooper Dr. Scott J. Edgett Dr. Elko J. Kleinschmidt This article was published in Research Technology Management (Industrial Research Institute, Inc.) Volume 43, Number 2, 2000. A memb
2、er company of the Product Development Institute 34 Stone Church Road Suite 111 Ancaster, Ontario L9K 1P4 Canada (905) 304-8797 Fax: (905) 304-8799 2000 Product Development Institute New Problems, New Solutions Making Portfolio Management More Effective Copyright 2000 Product Development Institute 1
3、New Problems, New Solutions: Making Portfolio Management More Effective There are two ways for a business to succeed at new products: doing projects right, and doing the right projects. Most new product prescriptions focus on the first route for example on effective project management, using cross-f
4、unctional teams, and building in the voice of the customer. Portfolio management, the topic of this article, focuses on the second route, namely on doing the right projects. In spite of all the hype around the topic of portfolio management, and the myriad portfolio methods proposed, managers have id
5、entified major problems and have raised serious concerns about the effectiveness of portfolio techniques. This article reports the results of continuing research into portfolio management practices and results: it highlights some of the problems, and offers some tentative solutions solutions that ha
6、ve been witnessed in typical firms as they try to address the issue of picking the right projects (see box insert). Portfolio Management is Vital, But Flawed Portfolio management is fundamental to successful new product development. Portfolio management is about resource allocation - how your busine
7、ss spends its capital and people resources, and which development projects it invests in. Portfolio management is also about project selection - ensuring that you have a steady stream of big new product winners! And portfolio management is about strategy: it is one method by which you operationalize
8、 your businesss strategy. Recent years have witnessed a heightened interest in portfolio management, not only in the technical community, but in the CEOs office as well. According to our recent survey of IRI members, portfolio management has gained prominence for a number of reasons 3: Financial to
9、maximize return on R and thus too many projects “pass the hurdles” and are added to the active list. Management seems to have difficulty discriminating between the Go, Kill and Hold projects. 3. Making Go/Kill decisions in the absence of solid information: The up-front homework is often substandard
10、in projects, the result being that management is required to make significant investment decisions, often using very unreliable data. No wonder so many of their decisions are questionable! 4. Too many minor projects in the portfolio: There is an absence of major revenue generators and the kinds of p
11、rojects that will yield significant technical, market and financial breakthroughs. These four problems are clearly interlinked. For example, the inability to discriminate between projects invariably leads to a resource balancing problem. Insufficient resources on key projects in turn results in proj
12、ect teams short-cutting key activities. Cutting corners on projects results in poor information and difficulty in making sound Go/Kill decisions. Inadequate resources and poor A lack of Resources for New ProductsNo Portfolio Management ProcessNo NewProductProcessToo Many small, low value projectsToo
13、 many projects for the limited resources availablePoor project prioritization; failure to kill projectsPoor data on projectsPoor job done on projects weak market studies, poor launch, inadequate testingLow Impacton Sales, ProfitsPoor Cycle TimesHigh Failure RatesFigure 2 A lack of resources, no port
14、folio management, and no new product process (left) are the root cause of many problems, which feed on one another, resulting in a downward spiral of negative effects and results.New Problems, New Solutions Making Portfolio Management More Effective Copyright 2000 Product Development Institute 4 inf
15、ormation invariably leads to a tendency to do short-term, quick and simple projects. And so the portfolio problems continue, feeding one another in an endless downward spiral (Figure 2). Consider each of these four challenges in more detail: Problem #1: Too many projects, not enough resources. Pipel
16、ine gridlock plagues many businesss portfolios: There are simply too many projects and not enough resources to do them well. This is a universal complaint within product development groups everywhere. The demand for more new products than ever coupled with corporate restructuring has helped to creat
17、e this resource crunch: One frustrated new product project leader at her companys technology conference exclaimed: “I dont deliberately set out to do a bad job. Yet, when you look at the job that the project leaders around here do, its almost as though our goal is mediocrity. But thats not true . we
18、re good project leaders, but were being set up for failure. There simply isnt enough time and not enough people or the right people to do the job wed like to do!” She went on to explain to senior management how insufficient resources and budget cuts coupled with too many projects were seriously comp
19、romising the way key projects were being executed. She was right! The point is: the resource commitment must be aligned with the businesss new product objectives, strategy and processes for positive results 5. The lack of resources is part of the problem. The other side is the failure to allocate re
20、sources effectively. Here portfolio tools and methods are partly at fault, along with a lack of will on the part of senior management to cut back the number of active projects to say “no” to some worthwhile initiatives. The fact is that most project selection and portfolio management methods do a po
21、or job of resource balancing. Projects are evaluated, Go decisions are made, but resource implications are often not factored in. Example: One of the most popular methods for evaluating projects and making Go/Kill decisions is the use of financial models, such as NPV 2,7. More advanced versions intr
22、oduce probabilities and uncertainties into the financial calculation. Management is presented with the NPV of the projects, along with probability distribution curves. These same models, while so elegant in their handling of financial estimates (revenues, costs, profits) are notably lacking in their
23、 handling of the resource constraint problem: resource availability is rarely part of the financial calculation. The majority of project selection techniques are quite weak when it comes to making Go/Kill decisions or choosing the portfolio in the light of constrained resources. There is really no w
24、ay to check that the required resources are available when using most of these selection tools. Indeed these selection tools consider individual projects one-at-at-time and on their own merits, with little regard for the impact that one project has on the next. Worse yet, people resources are assign
25、ed to projects, but only later is it discovered that the same resources are committed to multiple projects, and that some people are committed 150% of their time. New Problems, New Solutions Making Portfolio Management More Effective Copyright 2000 Product Development Institute 5 In one major bevera
26、ge company, there were constant complaints that major bottlenecks were encountered in new product projects in the package development department. A demand analysis was undertaken on a project-by-project basis; only then was it discovered how heavily committed certain players were. Each project team
27、member was assigned to projects (number of person-days each month). When the packaging departments time commitments were totaled up across all active projects, it turned out that this three-person group had been committed about 100 person-days each month. Figure it out: thats a 160% commitment. No w
28、onder there were logjams in the process! The results of too many projects in the pipeline are serious. Here are some of the negative effects weve observed: 1. Time to market starts to suffer, as projects end up in a queue waiting for people and resources to become available. A senior technology mana
29、ger in one Xerox division, concerned about project timelines, undertook a quick survey. He picked a day at random, and sent an e-mail to every project leader in his division: “How much work got done on your project today?”. The shocking news: more than three-quarters of the projects had no work done
30、 on them at all! Subsequent follow-up revealed that a minority had legitimate reasons for inaction waiting for equipment to be delivered, or waiting for tests to be completed. But the great majority were simply in a queue, waiting for people to get around to doing something on them. His best guess w
31、as that he could have halved time-to-market for most projects simply by having fewer active projects underway, and thereby avoiding queues 2. People are spread very thinly across projects. With so many “balls in the air”, people start to cut corners and execute in haste. Key activities may be left o
32、ut in the interest of being expedient and saving time. And quality of execution starts to suffer. The end result is higher failure rates and an inability to achieve the full potential of would-be winners. One major chemical company undertook an audit of its new product practices and performance acro
33、ss its many businesses. One common conclusion, regardless of business unit, revealed a lack of good market knowledge and customer input in the typical new product project. A task force was set up to study why. Their conclusions: marketing people were so thinly spread across so many new product proje
34、cts that they barely had time to oversee the launch of new products, let alone even think about doing market studies and solid market research. 3. Quality of information on projects is also deficient. When the project team lacks the time to do a decent market study or a solid technical assessment, o
35、ften management is forced to make continued investment decisions in the absence of solid information. And so projects are approved that should be killed. The portfolio suffers. 4. Finally, with people spread so thinly across projects, and in addition, trying to cope with their “real jobs” too, stres
36、s levels go up and morale suffers. And the team concept starts to break down 6. New Problems, New Solutions Making Portfolio Management More Effective Copyright 2000 Product Development Institute 6 Problem #2: Project selection methods fail to discriminate between projects Most project selection too
37、ls - for example scoring models and financial tools - consider the project against some hurdle or “minimum acceptable value”. In the case of NPV, for example, the NPV is calculated using a risk adjusted cost-of-capital. If the NPV is positive, the acceptable hurdle rate is achieved, and the project
38、is deemed a Pass. The trouble is, lots of projects pass the hurdles. What these methods really fail to do is provide for a forced ranking of projects against each other. Projects are rated against objective criteria, but are rarely force-ranked against each other. So there is little discrimination b
39、etween projects: they are all Gos! An international banking organization had established a well-oiled new product process, complete with rigorous Go/Kill decision points built in. These Go/Kill decisions were based in part on a scoring model and also on traditional profitability criteria. The proble
40、m is that many projects “passed” the hurdles at the gates, and so kept getting added to the active project list. As the list got longer and longer, the resources became spread thinner and thinner! The gating method looked at projects, each on their own merits, but failed to distinguish the top prior
41、ity ones from the rest. Forced-ranking of projects means making tough decisions: The result of this forced-ranking exercise is a prioritized list of projects, with the best ones at the top. Projects are listed until the business runs out of resources. Below that point, projects are put on hold or ki
42、lled outright. But all too often these tough decisions are not made: as one executive put it, “No one likes to drown puppies in our business!” This lack of discrimination among projects - where the best rise to the top of the list -is in part due to weaknesses in the particular selection tools used:
43、 NPV was designed for one-off decisions - for example, the decision to buy a new piece of equipment. But NPV was never meant for portfolio decisions, where multiple projects compete for the same resources. And ranking projects according to their NPVs does not yield the right portfolio either the met
44、hod ignores resource constraints. Finally, NPV calculations are always suspect in the early stages of a new product project. As one senior manager remarked: “What number do you want to hear? The project team always delivers the right number to get their project approved!” Scoring models are valuable
45、 decision aids for evaluating projects. But they too tend to rate projects against absolute criteria, rather than against each other. Admittedly, one might consider ranking projects according to their Project Scores. But again the issues of resource constraints and “bang for buck” are ignored: One m
46、ajor financial institution developed a scoring model to rate projects. Four fairly typical criteria were used: strategic fit and importance; market attractiveness; competitive advantage; and magnitude of the profit opportunity. Projects were scored on these criteria by senior New Problems, New Solut
47、ions Making Portfolio Management More Effective Copyright 2000 Product Development Institute 7 management on zero-to-ten scales; and the scores were added to yield a Program Attractiveness Score. Projects falling below a certain minimum score were discarded, and the remaining ones were rank-ordered
48、according to the Attractiveness Score. A review of the resulting three-page prioritized list of projects revealed an artifact of the ranking scheme. All the big-hit projects were on page 1 at the top of the list, and the small ones on page 3. But a closer review of the projects showed that many of t
49、hese big-hit projects also consumed large resources, while some of the projects on pages 2 and 3 of the list, although having lower scores, were also relatively inexpensive to do. The scoring model had missed the notion of efficient allocation of resources. A final complaint about scoring models is that often they fail to discriminate well. They tend to yield middle-of-the road scores 60 out of 100 which makes it difficult to spot the stars from the dogs. This is especially true when a large number of scoring criteria are used: high
