TEACHING NOTE
Genentech—Capacity Planning
A – Description
In November of 2004, David Ebersman, senior vice president of product operations at Genentech,
was preparing to meet with his team to discuss whether they should recommend that the company
begin construction of a $600M US cell culture production plant (“CCP3”) to produce Avastin, a new
cancer drug, and other related products. Ebersman’s recommendation decision was complicated by
long construction lead times and by uncertainty about future demand for Avastin and yields in the
production process. A number of FDA trials involving Avastin were underway, and the outcomes of
those trials would not be known for some time. In addition, Genentech would have to estimate the
size of the patient population that would choose Avastin if the FDA approved its use for additional
types of cancers. In addition, yields related to existing production facilities made processes
calculations regarding available production capacity uncertain as well. Finally, if Genentech decided
to proceed with construction, it would have to decide where to build the new plant—close to the
company’s current operations in Northern California, or elsewhere, possibly abroad.
B – Teaching Objectives
This case has been taught in a first year required course on Technology and Operations
Management, and in a second-year elective MBA course on Operations Strategy. The specific
teaching objectives of the case have included:
1. To provide a context to discuss capacity expansion where capacity cannot be added in small
increments, but rather must be added in large “chunks.”
2. To expose students to the realities of planning for capacity expansion with long lead times
and high levels of uncertainty.
3. To expose students to the details of operational planning based on estimates of future
demand. For example, this case can be used to introduce the concept of confidence intervals
around demand forecasts.
,606-111 Teaching Note—Genentech—Capacity Planning
4. To give the students an appreciation for the tradeoffs required in making location decisions.
Some of these considerations include natural disaster risk, employee willingness to relocate,
and protection of intellectual property.
5. Finally, this case gives students an introduction to some of the operations issues that are
distinctive to biotech production and the FDA regulatory processes.
C – Materials and Assignment Questions
Materials
Genentech – Capacity Planning (HBS No. 606-052)
Genentech – Capacity Planning Base Case Spreadsheet Model (on course platform)
For the assignment questions, assume the following:
a) Each of the two contract manufacturers can devote two 10,000 liter tank lines to Genentech
production, and Genentech hopes they will achieve yields similar to those at Genentech’s
own plants. These tanks will be fully utilized in the production of Rituxan and Herceptin.
b) Industry experts make demand forecasts for drugs like Avastin, but a number of sources of
uncertainty—yet to be determined dosage amounts and treatment regimens, unexpected
problems in the FDA’s approval process, unexpected success or failure of a competitor’s
product, and unexpectedly large or small consumer uptake—means that real demand will
depart from the experts’ forecasts. For purposes of our case discussion, assume that future
demand is distributed normally, centered on the experts’ forecasts, and with variation such
that one standard deviation is about 25 percent of the expected demand. For example, if
th
expected demand is 100 kg per year, then 85 percentile demand (about one standard
deviation above the expected demand) would be 125 kg per year.
Questions
1. What is your evaluation of Genentech’s production capacity requirements given expected
demand in 2010 and 2015 for Avastin and Genentech’s other products as per Exhibit 3? Does
your evaluation change if Genentech wants to cover the 85th-percentile level of demand? (See
spreadsheet posted on course platform)
2. Assuming Genentech decides to proceed with CCP3, what size production lines (tank sizes)
would you recommend? Why? What criteria should Ebersman use in selecting a location?
Why? Should Ebersman move forward with CCP3 now? (If not, when?)
3. What recommendations would you make to Ebersman regarding the process he and his team
should use in deciding how best to meet the demands for Avastin?
4. A contract manufacturing firm has had an unexpected reduction in demand for a drug it
produces. It is now offering to devote four 10,000 liter lines to the production of Avastin at a
price similar to Genentech’s existing contract manufacturing agreements. How should
Ebersman respond?
2
,Teaching Note—Genentech—Capacity Planning 606-111
D – Case Analysis
We begin this section by considering the sources of variability that affect Genentech’s forecast of
Avastin demand. Next we outline some of the key factors that need to be considered in Genentech’s
projections of required capacity. Third, assuming that Genentech decides to proceed with
construction of a new facility, we outline some of the important considerations in addressing the
location decision. Finally, the question of committing to additional contract manufacturing capacity is
addressed.
Demand Variability
The considerations that are relevant to Genentech’s capacity expansion decision can be organized
into two categories – issues that affect demand for Avastin, and issues that affect Genentech’s ability
to supply Avastin to meet that market demand. We will address each in turn.
The first step in forecasting demand for Avastin is determining the number of cancer patients who
have a type of cancer for which treatment with Avastin is suitable. (“Indication” is the term used in
the industry for an approved use.) Four pieces of information are required for this:
1. The indications for which Avastin is expected to be approved by the U.S. Food and Drug
Administration (FDA),1
2. The expected date of completion of successful trials;
3. The number of people who have each type of cancer for which treatment with Avastin is
indicated;
4. Genentech’s expected penetration or “market share” of patients with cancers for which
treatment with Avastin is indicated.
In November of 2004, Avastin had been approved for use in the treatment of metastatic colorectal
cancer—colorectal cancer that has spread beyond the patient’s colon. This is a late-stage cancer that
is most often terminal, and outcomes in cancer trials such as this one are typically measured in
number of months added to a patient’s life. In the case of Avastin, and indeed in the case of most
cancer drugs, the first indications for which the approval is received from the USFDA (United States
Food and Drug Administration) are end-stage cancers. Approval for less severe indications, if
received at all, typically follows approval for the more severe indications. This is largely a result of
the fact that outcomes in terminal cases are observed relatively quickly—often within a few months.
Success in less-severe indications is often measured by percent of patients for whom the cancer has
gone into remission, an outcome that requires much more time to observe. Regardless of the severity
of the indication for which approval is being sought, FDA trials occur in three main phases, followed
by a fourth post-trial phase:
Phase I. Researchers test a new drug or treatment in a small group of healthy volunteers (20–
80) for the first time to evaluate its safety, determine a safe dosage range, and identify side effects.
(This is often referred to as the stage that determines toxicity and safety.)
1 Although Avastin is intended to be sold elsewhere, its largest and most profitable potential market is the U.S. market. For
this reason, the analysis in this case considers only the U.S. approval process and U.S. drug consumption. Most other countries
have much smaller markets and accept the standards followed by the FDA. Thus once approved for patient use in the U.S., the
remaining steps to gain approval for drug sales in other countries are mainly a matter of organizing and presenting the
available data, along with a business plan including pricing, to the appropriate local government agencies.
3
, 606-111 Teaching Note—Genentech—Capacity Planning
Phase II. Researchers study the drug or treatment with a larger group of people (100–300) to
determine if it is effective with those suffering from the disease and to further evaluate its safety.
(This is often referred to as the stage that determines the efficacy and the effective dosage.)
Phase III. Researchers study the drug or treatment with a large group of people (1,000–3,000)
to confirm its effectiveness, monitor side effects, compare it to commonly used treatments, or no
treatments at all, and collect information that will allow the drug or treatment to be used safely. The
results of Phase III trials typically are designed so that if successful, they provide enough data for a
company to seek approval from the FDA.
Phase IV. Researchers do post-marketing launch studies to gather additional information
regarding the drug’s risks, benefits, and optimal use. (This can be thought of as the fine-tuning or
refinement stage.)
Case Exhibit 10 lists, as of the date of the case, the most significant Avastin indications under trial
in the FDA approval process together with information about each trial including the phase of the
trial. Additional information about each trial (including number of potential patients) is found in the
case text on page 9 and in case Exhibit 11. Estimates of market penetration given FDA approval are
presented in case Exhibit 11.
Industry analysts routinely provide their estimates of the probability of a drug successfully
advancing to subsequent phases. These probabilities may be used to estimate the probability of the
drug ultimately gaining approval from the FDA. In the case of a drug like Avastin being developed
by a company like Genentech, analysts have estimated a 50% probability of advancement to the
subsequent phase given successful advancement through the current phase. (Footnote “c” in case
Exhibit 11 provides these estimates of the conditional probabilities of successful advancement to
subsequent stages.)
In the case, the numbers of US patients afflicted with the relevant types of cancers are presented in
two places, on page 9 and in case Exhibit 11. Also found in case Exhibit 11 are analyst estimates of
the drug’s likely market penetration for each type of cancer.
This information on potential demand is presented in Table TN-A to show that case facts may be
combined to create a reasonable estimate of the size of Avastin’s market. Students are required to
make reasonable assumptions where case facts are ambiguous, for instance where a drug is at “Stage
2/3.” (Our experience with MBA students is that the final result, that Genentech needs to proceed
with construction within the next year or two, is not sensitive to reasonable differences in these
assumptions.)
With the information students gather from the case, they will be able to build a 10-year demand
model for Avastin. For the required course in TOM at HBS, it was concluded that the case would be
most effective for its intended use if all students had a spreadsheet model with much of these data
provided to them as part of their assignment, but to which they needed to apply certain assumptions
in order to complete the analysis. The content of this spreadsheet model, as it relates to product
demand, is shown in Exhibit TN-1A.
4
Genentech—Capacity Planning
A – Description
In November of 2004, David Ebersman, senior vice president of product operations at Genentech,
was preparing to meet with his team to discuss whether they should recommend that the company
begin construction of a $600M US cell culture production plant (“CCP3”) to produce Avastin, a new
cancer drug, and other related products. Ebersman’s recommendation decision was complicated by
long construction lead times and by uncertainty about future demand for Avastin and yields in the
production process. A number of FDA trials involving Avastin were underway, and the outcomes of
those trials would not be known for some time. In addition, Genentech would have to estimate the
size of the patient population that would choose Avastin if the FDA approved its use for additional
types of cancers. In addition, yields related to existing production facilities made processes
calculations regarding available production capacity uncertain as well. Finally, if Genentech decided
to proceed with construction, it would have to decide where to build the new plant—close to the
company’s current operations in Northern California, or elsewhere, possibly abroad.
B – Teaching Objectives
This case has been taught in a first year required course on Technology and Operations
Management, and in a second-year elective MBA course on Operations Strategy. The specific
teaching objectives of the case have included:
1. To provide a context to discuss capacity expansion where capacity cannot be added in small
increments, but rather must be added in large “chunks.”
2. To expose students to the realities of planning for capacity expansion with long lead times
and high levels of uncertainty.
3. To expose students to the details of operational planning based on estimates of future
demand. For example, this case can be used to introduce the concept of confidence intervals
around demand forecasts.
,606-111 Teaching Note—Genentech—Capacity Planning
4. To give the students an appreciation for the tradeoffs required in making location decisions.
Some of these considerations include natural disaster risk, employee willingness to relocate,
and protection of intellectual property.
5. Finally, this case gives students an introduction to some of the operations issues that are
distinctive to biotech production and the FDA regulatory processes.
C – Materials and Assignment Questions
Materials
Genentech – Capacity Planning (HBS No. 606-052)
Genentech – Capacity Planning Base Case Spreadsheet Model (on course platform)
For the assignment questions, assume the following:
a) Each of the two contract manufacturers can devote two 10,000 liter tank lines to Genentech
production, and Genentech hopes they will achieve yields similar to those at Genentech’s
own plants. These tanks will be fully utilized in the production of Rituxan and Herceptin.
b) Industry experts make demand forecasts for drugs like Avastin, but a number of sources of
uncertainty—yet to be determined dosage amounts and treatment regimens, unexpected
problems in the FDA’s approval process, unexpected success or failure of a competitor’s
product, and unexpectedly large or small consumer uptake—means that real demand will
depart from the experts’ forecasts. For purposes of our case discussion, assume that future
demand is distributed normally, centered on the experts’ forecasts, and with variation such
that one standard deviation is about 25 percent of the expected demand. For example, if
th
expected demand is 100 kg per year, then 85 percentile demand (about one standard
deviation above the expected demand) would be 125 kg per year.
Questions
1. What is your evaluation of Genentech’s production capacity requirements given expected
demand in 2010 and 2015 for Avastin and Genentech’s other products as per Exhibit 3? Does
your evaluation change if Genentech wants to cover the 85th-percentile level of demand? (See
spreadsheet posted on course platform)
2. Assuming Genentech decides to proceed with CCP3, what size production lines (tank sizes)
would you recommend? Why? What criteria should Ebersman use in selecting a location?
Why? Should Ebersman move forward with CCP3 now? (If not, when?)
3. What recommendations would you make to Ebersman regarding the process he and his team
should use in deciding how best to meet the demands for Avastin?
4. A contract manufacturing firm has had an unexpected reduction in demand for a drug it
produces. It is now offering to devote four 10,000 liter lines to the production of Avastin at a
price similar to Genentech’s existing contract manufacturing agreements. How should
Ebersman respond?
2
,Teaching Note—Genentech—Capacity Planning 606-111
D – Case Analysis
We begin this section by considering the sources of variability that affect Genentech’s forecast of
Avastin demand. Next we outline some of the key factors that need to be considered in Genentech’s
projections of required capacity. Third, assuming that Genentech decides to proceed with
construction of a new facility, we outline some of the important considerations in addressing the
location decision. Finally, the question of committing to additional contract manufacturing capacity is
addressed.
Demand Variability
The considerations that are relevant to Genentech’s capacity expansion decision can be organized
into two categories – issues that affect demand for Avastin, and issues that affect Genentech’s ability
to supply Avastin to meet that market demand. We will address each in turn.
The first step in forecasting demand for Avastin is determining the number of cancer patients who
have a type of cancer for which treatment with Avastin is suitable. (“Indication” is the term used in
the industry for an approved use.) Four pieces of information are required for this:
1. The indications for which Avastin is expected to be approved by the U.S. Food and Drug
Administration (FDA),1
2. The expected date of completion of successful trials;
3. The number of people who have each type of cancer for which treatment with Avastin is
indicated;
4. Genentech’s expected penetration or “market share” of patients with cancers for which
treatment with Avastin is indicated.
In November of 2004, Avastin had been approved for use in the treatment of metastatic colorectal
cancer—colorectal cancer that has spread beyond the patient’s colon. This is a late-stage cancer that
is most often terminal, and outcomes in cancer trials such as this one are typically measured in
number of months added to a patient’s life. In the case of Avastin, and indeed in the case of most
cancer drugs, the first indications for which the approval is received from the USFDA (United States
Food and Drug Administration) are end-stage cancers. Approval for less severe indications, if
received at all, typically follows approval for the more severe indications. This is largely a result of
the fact that outcomes in terminal cases are observed relatively quickly—often within a few months.
Success in less-severe indications is often measured by percent of patients for whom the cancer has
gone into remission, an outcome that requires much more time to observe. Regardless of the severity
of the indication for which approval is being sought, FDA trials occur in three main phases, followed
by a fourth post-trial phase:
Phase I. Researchers test a new drug or treatment in a small group of healthy volunteers (20–
80) for the first time to evaluate its safety, determine a safe dosage range, and identify side effects.
(This is often referred to as the stage that determines toxicity and safety.)
1 Although Avastin is intended to be sold elsewhere, its largest and most profitable potential market is the U.S. market. For
this reason, the analysis in this case considers only the U.S. approval process and U.S. drug consumption. Most other countries
have much smaller markets and accept the standards followed by the FDA. Thus once approved for patient use in the U.S., the
remaining steps to gain approval for drug sales in other countries are mainly a matter of organizing and presenting the
available data, along with a business plan including pricing, to the appropriate local government agencies.
3
, 606-111 Teaching Note—Genentech—Capacity Planning
Phase II. Researchers study the drug or treatment with a larger group of people (100–300) to
determine if it is effective with those suffering from the disease and to further evaluate its safety.
(This is often referred to as the stage that determines the efficacy and the effective dosage.)
Phase III. Researchers study the drug or treatment with a large group of people (1,000–3,000)
to confirm its effectiveness, monitor side effects, compare it to commonly used treatments, or no
treatments at all, and collect information that will allow the drug or treatment to be used safely. The
results of Phase III trials typically are designed so that if successful, they provide enough data for a
company to seek approval from the FDA.
Phase IV. Researchers do post-marketing launch studies to gather additional information
regarding the drug’s risks, benefits, and optimal use. (This can be thought of as the fine-tuning or
refinement stage.)
Case Exhibit 10 lists, as of the date of the case, the most significant Avastin indications under trial
in the FDA approval process together with information about each trial including the phase of the
trial. Additional information about each trial (including number of potential patients) is found in the
case text on page 9 and in case Exhibit 11. Estimates of market penetration given FDA approval are
presented in case Exhibit 11.
Industry analysts routinely provide their estimates of the probability of a drug successfully
advancing to subsequent phases. These probabilities may be used to estimate the probability of the
drug ultimately gaining approval from the FDA. In the case of a drug like Avastin being developed
by a company like Genentech, analysts have estimated a 50% probability of advancement to the
subsequent phase given successful advancement through the current phase. (Footnote “c” in case
Exhibit 11 provides these estimates of the conditional probabilities of successful advancement to
subsequent stages.)
In the case, the numbers of US patients afflicted with the relevant types of cancers are presented in
two places, on page 9 and in case Exhibit 11. Also found in case Exhibit 11 are analyst estimates of
the drug’s likely market penetration for each type of cancer.
This information on potential demand is presented in Table TN-A to show that case facts may be
combined to create a reasonable estimate of the size of Avastin’s market. Students are required to
make reasonable assumptions where case facts are ambiguous, for instance where a drug is at “Stage
2/3.” (Our experience with MBA students is that the final result, that Genentech needs to proceed
with construction within the next year or two, is not sensitive to reasonable differences in these
assumptions.)
With the information students gather from the case, they will be able to build a 10-year demand
model for Avastin. For the required course in TOM at HBS, it was concluded that the case would be
most effective for its intended use if all students had a spreadsheet model with much of these data
provided to them as part of their assignment, but to which they needed to apply certain assumptions
in order to complete the analysis. The content of this spreadsheet model, as it relates to product
demand, is shown in Exhibit TN-1A.
4
