On Costing and Pricing, Design, Engineering and Documentation

Selling Time and Other Poor Metrics

This business does not believe in selling time. Time is a poor metric as far as the buyer is concerned, it is a metric over which they have no control, and there is no motivation for the seller to reduce the time taken for a task. Take builders for example they have little interest in time saving methods, as using such reduces the time they can charge for, and requires them to look for another project: and such projects may not be available. On the other hand a builder is concerned about time when they buy the labour of subcontractors: so its ok to reduce someone else’s time but not their own. Selling time can therefore result in conflict between the parties concerned.

Most other incentive payment schemes are also problematic for the buyers, typically employers. With incentive payment schemes workers are paid more if they make more units. From the business viewpoint however there is little point making more product if there is little sale of the product.. Basically require a flexible workforce that can change its role, and adapt to current needs: for example swap from production to sales, to design, to production. Alternatively make to order rather than make to stock. The problem with make to order is that do not have product readily available off-the-shelf, and therefore people have to wait.

A good metric should be under the control of both buyer and supplier. The buyer should be able to stop the project and still have gained something of worth. Whilst the supplier shouldn’t be committed to supplying the impossible. Finding an acceptable metric is difficult.

Minimum Weight Does not Result in Minimum Cost

Even the costing of buildings and structures is flawed: costings are typically based on dollars per tonne ($/t) or dollars per cubic metre($/m^3). Whilst these rates may seem reasonable they are flawed with respect to optimisation, as they tend to push for reducing the weight or volume of material in a structure. The minimum weight structure however is not the minimum cost structure, as minimum weight tends to favour more complex fabrications and thus increased labour cost. Additionally when it comes to factory automation, more robust components are required to resist damage by mechanical handling, and therefore use more material. However the automated production system can fabricate significantly faster than manual labour, and in turn significantly reduce the unit cost of production. Repetitive manufacture of buildings therefore optimises costs from a different perspective than one-off custom design of buildings. Custom design and construction of a building can be several times more expensive than off-the-shelf repetitively manufactured buildings. Both cost of product and process need to be optimsed not just the material content of the product.

Price is a matter of Experiment not Science

Another issue regarding price, is that price is the exchange value of the goods and/or service, it is not the cost of production plus profit. If the exchange value is less than the cost of production then no further production will take place. But that one and only item may then become a rarity and its price may then increase. Determining a suitable price is a matter of experimentation. For example a business makes a product releases it to the market at some price and it sells out in a short time, the demand is high and they now see an opportunity to increase production capacity: but to do that they need to invest in larger production facilities. So they increase the price, to get the money they need to expand, doing so slows down demand. So a question arises: at what rate should they supply the market? For example increase production capacity and satisfy the demand in next 2 years, or stay with current capacity and satisfy demand over next 10 years? Will demand grow, will alternatives become available? It all gets very complicated, and its all very uncertain, lots of estimating attempting to predict the future. What ever decisions and actions the business makes its choices have to be funded by the revenue generated.

Service and Labour Costs

A starting point for services is to take a look at labour rates, in Australia we can look at industrial awards and market rates. Most engineering design involves drawing and calculations which are within the capabilities of 2 year qualified engineering associates or as I prefer to call them associate technologists. So a useful starting point is the now obsolete Draught Persons, Planners and Technical officers award. Being obsolete is irrelevant, price is indeterminate and relative.

The only three weekly rates of concern are:

  1. [C10] Trade [$640.40] relativity to C10: 100%
  2. [C7] Technician [$701.00] relativity to C10: 115%
  3. [C3] Engineering Associate/ Associate Technologist [$815.70] relativity to C10: 145%

The rates given are minimum rates, and seem to ignore the defined relativities. For comparison current federal minimum is [$656.90 ], from payscale.com average salary for a structural engineer is $70,565 , for 52 weeks, gives weekly rate of $1357/week. For an associate degree, structural drafter $45,032/year or $866/week, such position however does not give opportunity to put the full education to work. Civil engineering drafter, $53,750/year or $1033/week would typically be involved with storm water drainage and roads design, and more completely putting their education to work. For advanced diploma mechanical engineering, $77,300/year or $1486/week which more likely reflects putting the full education to work. Work done by an inexperienced engineer, can typically be done by an engineering associate. These are average market rates, so in rough terms expect a wage between $1000 and $1500 for an engineering associate. It would therefore appear that there is no real value to the 4 year degree in engineering unless actually going to put that education to work beyond the level of engineering associate. For small jobs it is inefficient to employ both an engineer and drafter, Australia’s civil/structural engineers typically seem to object to producing their own drawings, therefore to get drawings and calculations done efficiently by one person would need the services of an engineering associate. Anycase the values taken from payscale.com are averages, therefore some people get paid significantly more, and others get paid less, therefore the obsolete industrial award is still a useful reference for a minimum labour cost for relative costings and pricing exercises.

Effort Proportional to Complexity not Size

When it comes to structural design, the design effort is proportional to the complexity of the structure not to the size. So as long as the structural form remains the same, the engineering effort remains the same, no matter how small or large the stucture. Consulting architects and engineers, typically like to sell time or charge a fee which is a percentage of the capaital value of the physical works. Since larger buildings use more materials, they cost more and engineers fees are greater, and this tends to reflect the greater risks associated with the larger structures. However for smaller structures such fees may not cover the cost of the design, and therefore time based fees are used, the problem then is that the design fees can exceed the cost of the building. It is therefore important that small buildings are designed once an made many times, with the cost of design being distributed across sales of multiple units. This however raises another problem, in that time based fees do not cover the high risks involved of putting a large number of units into the built environment each year: say 50 or more units per year. The pricing approach for design, offered by consultants, therefore is neither appropriate for small one-off structures, nor for repetitively manufactured structures.

Assuming the following breakdown of labour for a small structure using manual methods starting from scratch:

  1. Drafting [40 hours]
  2. Calculations [40 hours]

Then using a 40 hour week, {I am aware that some awards mention 38 hour/week}, then the required work effort is two weeks for an associate technologist at cost of 2 x $815.70 = $1631. It would appear that by employing a technician to produce the drawings and associate technologist for the calculations that the fee could be reduced to: $701+$815.70=$1516.70, but this raises other issues. Can the two people complete the work in 1 week, rather than 2 weeks, and what value does the time saving have? Basically the two people cannot produce the work faster, as the drafting and calculations cannot be completed in parallel, the work has to be done in series, therefore one person has to wait for the other, and therefore needs additional work to do: that is the two people require two projects to work on in parallel. As a simple example assume the work is broken into 20 hour blocks: so that:

  1. Drafting stick diagrams [20 hours]
  2. Calculations over all framing [20 hours]
  3. Drafting of connections and other details [20 hours]
  4. Calculations for connections [20 hours]

So that at any point in the sequence, one person has to find 20 hours of work on another project. It is actually more complex than that, as the iteration between drawing and calculations can be occurring every few minutes whilst design is being carried out. Drawing is required to get dimensions for the structural calculations, and structural calculations are required to size parts placed on the drawings. As to whether drawings or calculations occur first is a chicken and egg problem. So whilst design is taking place and do not have an abundance of alternative projects to work on, a single person doing all the work is more efficient. However if there is no actual structural design, then drawing and calculations can occur in parallel, in such instance the calculations are being carried out to assess suitability of some common standardised structural system. In such situation the proactive design of the structural system has value, and the time saved by using such system has value. In other words reducing time does not reduce price, as getting things faster has high value.

For example assuming through the use of technology the 40 hours of calculation can be reduced to 5 minutes, this is effectively what manufacturers have achieved with their point-of-sale configurator software. Buying such software off-the-shelf has a cost, developing such software has a cost, a cost which needs to be incorporated into the sales price of the structural product it is used to sell. So time may have been reduced but the costs haven’t been reduced. Further more in terms of consultants still providing such design service, the market price didn’t suddenly drop in half to only cover the cost of the remaining 40 hours of drafting. The price will only drop, when the automation technology becomes the routine method adopted by all, and each new entrant into the market, calculates a cost based on time. That is to say a new entrant estimates that task takes 5 minutes, therefore at labour rate of $0.34/minute, the cost is $1.70 so how can the competition get away with charging $815.70 ? Then again they may not know what the competition is charging. Either way the new entrant to the industry offers a service at significantly lower price and transforms the market.

If something can be designed once and then manufactured many times, then it is possible to place more design effort into the efficiency of the production system and adopt more automated manufacture. Where once there was a profitable market for custom design and construction, emerges a profitable market in repetitive manufacture, with design and development costs distributed across sales of increasing number of units. Now whilst there may be limited market for off-the-shelf product, the manufactured product provides a base from which alternatives can be easily introduced making the product better suited to the needs of others. So the cost of design is recovered and then invested in the design of alternative product. So there is a continuous process of adapting the product to better match the needs of consumers: at the very minimum the manufacturer may cycle around a few standard product forms, which meet the current trends in the market.

At the end of the day, the fundamental end-product that people want from consulting design services is decent documentation specifying and proving a physical object which is to be made. Such documentation, provides a specification for an end-product that can be made many times. The documentation supplied by consultants is typically only permitted to be used to make something once. In such situations the service provided is typically more important than the documentation, and the documentation is a side effect. However, it is clear that cannot be certain get any real service from consultants, nor can it be clear that they actually tackle the clients problem. When it comes to small projects, more often than not the consultant just goes through a ritual, performing calculations with no concern for the results. For example building designer specifies beam by engineer, the engineer produces calculations and specifies a beam size with no consideration to whether there is space to fit the beam. In such situation the engineer has not solved the problem, just followed a ritual of calculation, they have not provided a service of any value. Further more can take into consideration that the calculations are also highly repetitive, for example time after time they specify the same 250PFC floor beam, they never consider using say a lighter cold-formed steel C25024. The latter would require more complex calculations: raising issue of calculation time versus material savings. For example, there is little point paying $1000 for extra calculations to save $1000 worth of concrete. It is better to have the added resistance of the concrete than to have calculations saying do not need. The calculations have to provide benefit: if the design is repetitive, so that saving $1000 worth of concrete on 50 projects a year, then getting benefit.

National standards and performance based design allow multiple prescriptive solutions to be documented, so that as with the timber framing code (AS1684), the need for technical specialists such as professional engineers can be eliminated, and thus enabling more people to implement and gain the benefit of a given technology. Further more a designer is not limited to the type of projects which turn up on the consultants desk. As a consultant, a designer may be stuck designing boxes. As a designer, rather than a consultant, a person can design what ever they want, it doesn’t require anyone to implement. Further more people who buy the documentation of the design do not have to implement. That is a designer can communicate and share ideas, for a fee or gratis.

Digital Products

That is where the next issue of pricing comes into play. With digital products the replication and distribution of the end-product has very little cost. Therefore after recovering the original costs of design and development behind the structural product specified in digital documents, the document cost can approach zero. Despite the use of digital technologies, national standards and codes of practice are not made available at zero price: given that such documents are typically produced by volunteers the fees for such digital products seems significantly higher than the distribution costs would indicate. So if have a captive market created by legislation, can hold the community to ransom: at least until the volunteers produce alternative documents and community accepts the de facto standards.

Any case if a service is productised, then the content delivered by that service is clearly defined, its price may initially be based on the original time based cost of labour required for the task, but there after the time for the task can be reduced and increasing levels of automation can be adopted. The cost can be reduced whilst the market price remains the same. In such case however it still remains a service.

Replace the service however by documentation, then the cost of production can be recovered from multiple sales of the document. So assuming the cost of production is $1631, then with sale of 2 copies the unit cost can be reduced to $815.50, with sale of 10 copies to $163.10, and 100 copies down to $16.31, and with sales of 1000 copies the unit cost would be $1.63. Ultimately the document could be given away but that would not cover the costs of distribution. Distribution is not just a matter of making the document available, but informing people that it is available. The price of such document can therefore be compared against the price of technical books on amazon.com.

Such informing about existence of document, is an important issue if copyright is important. If someone gets the document and makes the document available free of charge, then it is not a major issue unless it is easier to find than the paid version. If it is easier to find, then the owners of the copyright, can find the illegal distribution and can generally shut such supply down.

So move from supply of services to retail distribution of documents. A retailer expects to make an income by selling objects previously made by others: in this case those objects are digital documents. Assuming it does take two weeks to fully design and document a structural product, then in 52 weeks can produce 26 documents. Therefore after a year of production have 26 documents available for sale, and effort could shift from design and documentation, to simply improving the presentation of the documents, and otherwise selling the documents. So assuming sell one copy of each document each week, then would have to sell at a minimum fee of $31.37 to make the same weekly wage. With sale of 26 documents in a week, then would need to average sales of approximately 4 items per day every day. Which may or may not be viable.

The price of individual items now becomes dependent on the demand for that item not the cost of production. Those items with high demand would have the lowest price, whilst those items with low demand would have the highest price, even though the cost of producing each document is the same. So for example high demand documents could be sold at $20 each, whilst low demand documents could be sold at prices around $500. All that matters is that, at the end of the day, the product mix, and the total sales, generates adequate income for the retailer to provide a shop.

Now in the age of software, documents may be considered of little value. However, someone has to operate software and currently that software is used to produce documents for submission to others, who use the documents to grant approval. For efficiency, the documents should and will ultimately be scrapped. If a structure is modelled and designed by computer, then the computer model should be handed over for interogation, checking and the granting of approval, not paper or digital documents. However the design and approval process will still involve reviewing information displayed on the screen. Documents from other sources will therefore provide important bench marks for the learning and reviewing process. Furthermore before software can be customised to assist with design in a more automated manner, human legible documentation of the design process and examples are required. Whilst software can determine that something doesn’t work it doesn’t provide guidance as to what needs to be modified to make it work. A qualitative understanding of what is going on behind the scenes is required to efficiently make modifications to get a proposal to work.

Besides reference documents and bench marks for developing software and learning design, standard designs still have important value. Pushing numbers through software takes time, if that has been done already, it is more productive to make use of the documentation and make the specified object multiple times, than to keep running software.

There is thus need for fixed form product, which locks all the design characteristics, and variable form product which allows some of the characteristics to vary as design parameters. In both cases however the cost of producing the document and/or software is to be distributed across multiple sales to reduce the unit cost of smaller scale buildings and structures. Rather than the cost of the design being distributed across sales of multiple buildings by a manufacturer, the cost is distributed across the sales of multiple documents to private individuals. That is private individuals are granted license to build at least once upon purchase of the licensed documents. This changes the market, as private individuals and small builders are currently locked into buying from manufacturers unless they are willing to pay high fees for design. Generally they are not so willling, as the objective of building themselves is to save money, and they consider that they can save money once they have costed the materials and subtract that from the prices offered by manufacturers. As a DIY builder they remove the costs of labour and the profit that is embedded in the manufacturers price: in some instances the saving is enough that even if they pay high fees for design, they still achieve an over all saving.

Standardised documented designs also cut the costs for manufacturers, it also provides them with alternative and more interesting designs that they may not otherwise pursue. In other words sales of a design document, indicates interest in an idea, an idea that a manufacturer may not wish to pursue.

Designers publishing documents puts ideas out there in the public space, and design consultants can pick up those ideas and implement them. The documents do not need to be limited to complete design of structures, they could be concerned with information and limiting constraints on structures. For example can wait for some cold-formed steel shed manufacturer to come along and ask to design a 21m span shed for them, or can design such structure before they turn up and ask, more importantly could identify what the limits are for cold-formed steel sheds. Knowing the limits for cold-formed steel portal frame sheds, can then consider what is required to increase the heights and spans from the use of cold-formed steel: such as doubling up sections or using flat bar reinforcement.

From consideration of a single end-product to design of an entire range of products to the design of an industry. The potential that can be achieved from the knowledge contained in the documents now starts to have far more value than the cost of the medium through which the document is communicated and distributed. A $20 document could make thousands of dollars for someone who has the resources to make and distribute the end-product specified by the document. The idea hidden and dormant is worthless, unleashed and public the idea has value.

Since knowledge has value, the price of the documents can escalate, as it is the knowledge that is being sold not the document. However have to be able to read the document and make use of the knowledge for it to be of any value. That is where the design of the document takes over from the design of the object it specifies: improving presentation and the communication achieved. Further the documents can be complemented by study and training programmes to assist people to understand the knowledge and put it to use.

There are two negative price perceptions to address. The public on low incomes, wanting to do DIY to save money who compare the cost of documentation of design against their own wages: to them the fees are extortionate for a bundle of papers which they cannot understand. Then there is the small group of engineers who think they are gods gift to humanity, and desire higher wages and consider low fees to be undercutting their actual wages which they consider too low.

For the DIY, there has to be repetitive demand to distribute the cost of engineering across sales of multiple documents (prescriptive solutions): therefore cannot have custom design if want low fees. For that small elitist group of engineers, most of those complaining and whinging, are far from gods gift to humanity, and for the most part fail to provide an efficient and competent service. So either provide a proper service and increase fees accordingly, or understand the market and design an end-product that can be provided at a price the market can support.

The prices shown on this web site are based on many assumptions, and whilst feedback from actual sales will provide satistical data to forecast future demand, that demand will still be based on an assumption that past demand can predict future demand: which it cannot. Statistics need to be complemented by qualitative technological forecasting, and even then the future is still unknown.

Anycase, this post has provided enough information for people to make their own rough estimate of costs based on selling time, though the cost ignores the operational costs of running the business, the biggest contribution to the final price is expected demand. That costing also ignores that a designer can sell their knowledge and expertise for any price they like: assuming they can get the price they ask for.



  1. [16/10/2015] Original
  2. [21/10/2015] Minor edits