I had an opportunity to share ideas and experience about product cost management and product profitability with Eric Hiller. Eric responded to my post on PLM and Product Cost Management (PCM) and we have shared some great dialogue following that. Eric has some great experience in this area and I asked him if he would like to share it. I hope you enjoy it!
The following is a guest post from Eric Hiller:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Sure, cost is important… but I have a lot of other things to do, too.
I’ve been fascinated by the problem of product cost since 1996. I did my graduate work on the subject, I considered it when I was a young engineer at Ford Motor Company, I started a company that makes product that helps companies make their products more profitable, and I have been a consultant to executives in Fortune 500 companies on the subject of product cost management. I’ve learned a lot of lessons in those endeavors, but one of the most interesting is:
Product Profit is NOT as Important as One Might Think
I know that’s pretty shocking to say, especially for a guy who is fascinated by the subject of Product Profit. But it’s true… or at least it’s true in a relative sense versus other product attributes that the product development team must meet.
The statement above also seems surprising given that we have been taught for years that the purpose of business is to make money for the owners, and money for the owners comes from free cash flow, and free cash flow comes from profit, and profit comes when expenses are less than revenue, and in most manufacturing companies, Product Cost (roughly equivalent to Cost of Goods Sold for the finance types reading this post) is 70-90% of revenue.
Considering my rough Aristotelian logic in the last sentence, why wouldn’t we assume that product cost is, indeed, the most important thing to business? (I’m being a bit facetious in my rhetorical question to prove a point here.) The point is that there are many other things that go into making a successful business, even from the product development-only standpoint. These include:
- product performance
- time-to-market
- quality
- safety/compliance
- number of features of the product
- product profitability
- development cost
In product development, most teams have to balance all these goals. We can see this from Tech Clarity’s 2011 report on product development tradeoffs in Figure 1. When you ask the average executive in product development, they are concerned with a multitude of issues. Even the least important category below (Product Sustainability/Green Initiatives) has almost 50% interest from executives. (Note that I consider manufacturability to be a concern that is very closely aligned and/or drives product cost.)
Figure 1 – What Product Development Executives are Concerned About
But, which of these goals is that the MOST important goal. Michelle Boucher at Aberdeen Group gauged the sentiment of product development executives in November 2010 on this question. Here are the results that she found (Figure 2), when she asked 312 respondents to pick “The top business pressure driving their company to have better insight into decision-making during product development.”
Figure 2 – What is the MOST important concern to product development managers?
Source: Aberdeen Group
I particularly like this survey question, because it forces the respondents to pick the most important. There’s a couple of interesting things here. First, if we look at the results of this survey versus similar questions asked in the past, we will see that the focus on product cost and profitability is less than it was a year or two ago. Hopefully, this is indicative that the economy might just be starting to show little signs of recovery. Second, notice that when the survey forced the respondents to rank the importance of goals in an ordinal way, the focus of PD executives is still fairly balanced across the different goals that we discussed in the bullets above. In fact, if we combined “customer demand for lower product costs” and “need to lower development costs” together, the total focus on cost or profitability of the organization is about 27%, which makes it almost identical to the other four focuses.
This data matches my own experience in talking to hundreds of customers over the last eight years. In the highly unscientific and intuitive statistics cruncher in my brain, if I combined all the feedback (direct and implied) that I have gotten to the same question, I would say the ordinal ranking of the average product development organization, whether spoken or unspoken, is as follows:
- Safety/compliance (required to sell product at all!)
- Time-to-market
- Quality
- Product performance
- Number of features of the product
- Product profitability
- Development cost
In fact, when I asked the question directly to several product development executives, this is almost exactly the order they gave me, although often they will not list every one of these attributes in their answer.
So what does this mean?
Does it mean that product cost and profitability are unimportant things? No. It simply means that sometimes they may be less important than other goals. So how do we make progress in product development and delivery of corporate profitability, given that product cost is the least favorite subject of many design engineers?
I suggest the following simple framework:
- Set the required level needed for each of the other attributes, except for product cost / profitability
- Set a minimum level target for product profitability, but think of product profitability in your mind as the variable you are optimizing.
- Execute on meeting the goals for the non-product cost attributes actively first, while monitoring product profitability and evaluating each choice made about a non-cost attribute, asking “Will this decision make my product profitability go up or down?”
Those of us who come from a background of optimization probably remember that it is very hard to optimize on more than one variable at a time. Therefore, the typical way to make progress practically on a optimization problem with multiple goals (which is exactly what product development is), is to set some hard constraints (Attribute Targets) for all the variables but one, and then run the optimization to maximize or minimize that last variable of interest (in this case product profit or cost, respectively).
A simple conceptual graphic of this framework is shown below. In the green ovals we show lines of equal product profit. We’d like to climb the hill to maximum profit, but there are real world constraints on the level of our other product attributes (e.g. time-to-market). This leaves us a tan shaded region where all our non-cost targets are met. In my suggested framework, the team first works to get into the shaded region and then starts making choices to meet (and optimize) the last target of product profitability. The axes represent two choices the team is making, but we know that the team will really be making thousands of choices (it’s just hard to represent that in 2D).
Figure 3 – Graphical Representation of meeting product development targets as constraints, while optimizing product cost and profit
This approach may sound like common sense to some people, and if so, great! However, don’t let the point pass you by without internalizing it a bit. One of the blinding flashes of the obvious from David Allen’s book Getting Things Done is that the human brain just can’t let go of something important until it is written down or handled. So, one of his first rules is that you have to “write it down.” Similarly, instead of letting Product Profitability be that nagging voice that keeps distracting everyone, companies can simply “write it down,’ too. That is, they should set a maximum product cost target and keep checking it in the rear view mirror, while keeping the immediate goals of Safety/compliance, Time-to-market, Quality, and Product performance, etc. in the front windshield view.
This may sound like a nuance, but I still often get the question “How do I balance my product cost targets with all the other targets I have in product development?” Years ago, I was perplexed what to tell the questioner other than “Make product cost your top priority.” That was not a realistic solution to the real concern. Eventually, after thinking through the problem and observing the culture of product development, I came up with the simple framework we are discussing. When I have explained this framework to product development teams at past clients and customers, it seems to make people feel much more at ease and helps people regain a sense of control.
Obviously, one could change the framework to make any product development attribute or target the ‘to be optimized’ variable, just as easily. But given the immediacy of the challenges of program timing, quality, and performance, it seems that making product cost the optimization variable works best for most product development teams. Sometimes, we just have to learn by experience.
Eric
(p.s. I still think Product Cost Management is really important!)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
So that is what Eric has to say, I hope you found it interesting. Are we really too busy to care about cost? Do we just believe that profitability is more about top-line issues by having the right product? Let us know how you feel about it.
Eric, great post. I was surprised to learn that product costs figured as low as they do according to both surveys cited in your post. With regards to better insight for decision-making during development, what role do you think access to the right data has in both improving decision making and addressing external business pressures?
Vic,
Thank you for your interest and response. As a recovering engineer
myself, I realize that cost geeks like me don’t have the sames mindset
as the average engineer. There are just a lot of ‘urgent’ priorities
on the mind of the dev team… and to be honest, maybe a lot of them are
more *fun* to work on than profit is.
I think your question could be the subject of a whole new blog post, so
thanks for the idea. But, to muse here a bit, I think that having the
right data is like the old GI Joe motto: “Now you know, and knowing is
half the battle.” I’ll interpret your question in the broader sense
of ‘data’ being the top level data, such as “what will this design
cost”. As such, I see at least 2 ways that people end up in a bad
place.
1. They don’t have complete data or have the wrong data at the time of the decision
2. They cannot synthesize the data together to make a good decision
I believe that (1) is the big problem most organizations face. Most
managers and engineers are very sharp people whose brains are quite
adept at ‘auto-optimizing’ to create the highest value product. If I
had to shoot from the hip, I would say (2) is the problem only 10% or
less of the time.
So, then we ask, what does “access” to the data mean? Does it mean:
a. The data is there, but not available to the decision maker easily OR
b. The data just doesn’t exist (needs to be calculated, measured, etc.)
With respect to product cost info, my experience is that both are
problems. Assuming we had the best PLM / ERP / SCM systems in the world
and all users were 100% compliant with recording all their knowledge in
these systems and everyone knew exactly how to query the system to find
the data, that would solve (a)… or go a long way towards it.
However, I think (b) is actually the problem more often, i.e. the cost
information for this design, or supplier, or routing in a manufacturing
plant just does not exist in the organization.
E
Good discussion guys. I think PLM provides a great way to communicate data, and I have long felt that PLM is the right place for costs. The issue (I think I am agreeing with you here Eric) is that there is a difference in the ability to share data and having data that is worthwhile to share.
– For sharing information (and collaborating on it, and making decisions on it) – PLM is the right place to focus on cost. It is early in the product lifecycle where costs can be impacted, and it has the right product-centric data to put cost in context (even manufacturing cost if PLM expands to digital manufacturing).
– Having worthwhile cost data – PLM is not designed to develop real costs. It has a lot of the right elements (you need BOMs, for example) and some PLM systems have developed some supplier information. But, developing a cost is not just sharing information – it is an analysis and optimization function. PLM, to date, does not have the proper analysis tools for cost. For purchased items, ERP integration (for past costs) and supplier collaboration will help. But for new costs, what does PLM have to offer? In most cases, not much.
To me this is a no-brainer. PLM is the right place for cost, and just like other attributes that are derived from items, BOMs and designs (like environmental compliance) it belongs as part of the integrated product model. Maybe I view this too simply?
Eric, great post. I was surprised to learn that product costs figured as low as they do according to both surveys cited in your post. With regards to better insight for decision-making during development, what role do you think access to the right data has in both improving decision making and addressing external business pressures?
Vic,
Thank you for your interest and response. As a recovering engineer
myself, I realize that cost geeks like me don’t have the sames mindset
as the average engineer. There are just a lot of ‘urgent’ priorities
on the mind of the dev team… and to be honest, maybe a lot of them are
more *fun* to work on than profit is.
I think your question could be the subject of a whole new blog post, so
thanks for the idea. But, to muse here a bit, I think that having the
right data is like the old GI Joe motto: “Now you know, and knowing is
half the battle.” I’ll interpret your question in the broader sense
of ‘data’ being the top level data, such as “what will this design
cost”. As such, I see at least 2 ways that people end up in a bad
place.
1. They don’t have complete data or have the wrong data at the time of the decision
2. They cannot synthesize the data together to make a good decision
I believe that (1) is the big problem most organizations face. Most
managers and engineers are very sharp people whose brains are quite
adept at ‘auto-optimizing’ to create the highest value product. If I
had to shoot from the hip, I would say (2) is the problem only 10% or
less of the time.
So, then we ask, what does “access” to the data mean? Does it mean:
a. The data is there, but not available to the decision maker easily OR
b. The data just doesn’t exist (needs to be calculated, measured, etc.)
With respect to product cost info, my experience is that both are
problems. Assuming we had the best PLM / ERP / SCM systems in the world
and all users were 100% compliant with recording all their knowledge in
these systems and everyone knew exactly how to query the system to find
the data, that would solve (a)… or go a long way towards it.
However, I think (b) is actually the problem more often, i.e. the cost
information for this design, or supplier, or routing in a manufacturing
plant just does not exist in the organization.
E
Good discussion guys. I think PLM provides a great way to communicate data, and I have long felt that PLM is the right place for costs. The issue (I think I am agreeing with you here Eric) is that there is a difference in the ability to share data and having data that is worthwhile to share.
– For sharing information (and collaborating on it, and making decisions on it) – PLM is the right place to focus on cost. It is early in the product lifecycle where costs can be impacted, and it has the right product-centric data to put cost in context (even manufacturing cost if PLM expands to digital manufacturing).
– Having worthwhile cost data – PLM is not designed to develop real costs. It has a lot of the right elements (you need BOMs, for example) and some PLM systems have developed some supplier information. But, developing a cost is not just sharing information – it is an analysis and optimization function. PLM, to date, does not have the proper analysis tools for cost. For purchased items, ERP integration (for past costs) and supplier collaboration will help. But for new costs, what does PLM have to offer? In most cases, not much.
To me this is a no-brainer. PLM is the right place for cost, and just like other attributes that are derived from items, BOMs and designs (like environmental compliance) it belongs as part of the integrated product model. Maybe I view this too simply?
A topic I’ve sort of given a lot of thought to lately, not in an academic or expert way as you have, but just in the course of my client work and their increased focus. If you polled my industry (pharma/biotech), the most dominant issue would be development cost, so it’s interesting that more broadly its not perceived to be as big a concern. Interestingly enough, despite the fact that I technically consult upon growth strategy (topline growth), all I hear about all day are strategies to minimize development costs. In an industry where revenue (number of people using drug x drug price) is intrinsically tied to the development process (number of people using drug is directly limited by the kinds of people you enroll in clinical trials, which are the biggest dev. cost driver), clients are constantly making the trade off of bigger, more expensive development programs to access a larger consumer pool, or smaller cheap development programs targeted at a smaller patient pool – essentially a profitability tradeoff.So, when we run our optimization process, which we call target product profile development, it quickly become a two-pronged process: optimizing on maximizing revenue and a separate process to minimize development costs, then the clients pick the lesser of the two evils (i.e., do I shoot for profile A that gets me 3B in revenue with 1B in dev costs and a 10% probability of success or profile B that gets me 600m in revenue with 200 M in dev costs with 75% probability of success…). Rightly or wrongly, risk-adjusted NPVs and productivity indices (expected NPV / expected dev costs) have become key drivers of decision making in pharma / biotech, as a way of representing the immense risk, reward, dev/time, and dev costs. On a separate note, I did a project 6 months ago for a big Japanese company helping them with optimization of their R&D portfolio, and obvi product costs played a big role. Not on a CoGS basis, but understanding the risk/reward tradeoff on a portfolio basis of making design changes / considerations to a single asset, and understanding how to balance risk/reward inherent to individual products across the portfolio as a means to hitting long-term revenue targets. We’ve done similar projects with individual franchises at a number of pharma companies, and I’m always fascinated by the different approaches to decision making. On a portfolio level, product costs bring up a different question – not how to minimize product costs, but rather who to choose which costs to incur and what goes into that decision.
Very well explained and interesting, Ankita.
The world of discreet mechanical manufacturing (cars, blenders,
iphones, etc.) is obviously very different from pharma, but at a certain level
costs are costs… it’s just that the percentage of each bucket of cost is different.
In manufacturing, R&D is 3-5% of Rev and COGS is
70-90%. In your industry, COGS is so low nd all cost is from
Marketing (SG&A) and R&D. Had a look at Merck
COGS (TTM) ~36%
SG&A ~29% (I assume mostly in ads and mkting?)
R&D ~23%
Merck’s been “struggling” to get by with a 10% Operating Margin, which is down from the mid 2000’s where it was in the 20’s!
In the discrete manufacturing world 10% EBIT is a happy healthy business. There are a few companies that shoot the moon (GE, Apple, Danaher), but not many — especially in non-electronic products.
Another thought that pops in my mind.
* In pharma, it seems that before you go to market, you both know the product will work, AND you seem pretty confident customers will buy it.
* In discrete manufacturing, you know the product will work, but I perceive there is much more uncertainty on how the market will develop.
I wonder if there is data on this?
Eric
A topic I’ve sort of given a lot of thought to lately, not in an academic or expert way as you have, but just in the course of my client work and their increased focus. If you polled my industry (pharma/biotech), the most dominant issue would be development cost, so it’s interesting that more broadly its not perceived to be as big a concern. Interestingly enough, despite the fact that I technically consult upon growth strategy (topline growth), all I hear about all day are strategies to minimize development costs. In an industry where revenue (number of people using drug x drug price) is intrinsically tied to the development process (number of people using drug is directly limited by the kinds of people you enroll in clinical trials, which are the biggest dev. cost driver), clients are constantly making the trade off of bigger, more expensive development programs to access a larger consumer pool, or smaller cheap development programs targeted at a smaller patient pool – essentially a profitability tradeoff.So, when we run our optimization process, which we call target product profile development, it quickly become a two-pronged process: optimizing on maximizing revenue and a separate process to minimize development costs, then the clients pick the lesser of the two evils (i.e., do I shoot for profile A that gets me 3B in revenue with 1B in dev costs and a 10% probability of success or profile B that gets me 600m in revenue with 200 M in dev costs with 75% probability of success…). Rightly or wrongly, risk-adjusted NPVs and productivity indices (expected NPV / expected dev costs) have become key drivers of decision making in pharma / biotech, as a way of representing the immense risk, reward, dev/time, and dev costs. On a separate note, I did a project 6 months ago for a big Japanese company helping them with optimization of their R&D portfolio, and obvi product costs played a big role. Not on a CoGS basis, but understanding the risk/reward tradeoff on a portfolio basis of making design changes / considerations to a single asset, and understanding how to balance risk/reward inherent to individual products across the portfolio as a means to hitting long-term revenue targets. We’ve done similar projects with individual franchises at a number of pharma companies, and I’m always fascinated by the different approaches to decision making. On a portfolio level, product costs bring up a different question – not how to minimize product costs, but rather who to choose which costs to incur and what goes into that decision.
Very well explained and interesting, Ankita.
The world of discreet mechanical manufacturing (cars, blenders,
iphones, etc.) is obviously very different from pharma, but at a certain level
costs are costs… it’s just that the percentage of each bucket of cost is different.
In manufacturing, R&D is 3-5% of Rev and COGS is
70-90%. In your industry, COGS is so low nd all cost is from
Marketing (SG&A) and R&D. Had a look at Merck
COGS (TTM) ~36%
SG&A ~29% (I assume mostly in ads and mkting?)
R&D ~23%
Merck’s been “struggling” to get by with a 10% Operating Margin, which is down from the mid 2000’s where it was in the 20’s!
In the discrete manufacturing world 10% EBIT is a happy healthy business. There are a few companies that shoot the moon (GE, Apple, Danaher), but not many — especially in non-electronic products.
Another thought that pops in my mind.
* In pharma, it seems that before you go to market, you both know the product will work, AND you seem pretty confident customers will buy it.
* In discrete manufacturing, you know the product will work, but I perceive there is much more uncertainty on how the market will develop.
I wonder if there is data on this?
Eric
Eric, great post. I don’t think listing the preference order made it seem that PCM was less important. To me, it is that human side of where a group feels “comfortable,” or where they feel like they can succeed and positively impact PCM. With that said, I also have a question in how a new company vs. an established company views quality? Some products do have to stand the test of time…
Amy
Eric, great post. I don’t think listing the preference order made it seem that PCM was less important. To me, it is that human side of where a group feels “comfortable,” or where they feel like they can succeed and positively impact PCM. With that said, I also have a question in how a new company vs. an established company views quality? Some products do have to stand the test of time…
Amy