Вы находитесь на странице: 1из 9

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/267772729

AN ANALYSIS OF COST ESTIMATION AND PRICING IN MAKE-TO-ORDER


COMPANIES

Article

CITATIONS READS

2 58

2 authors, including:

Antônio Artur de Souza


Federal University of Minas Gerais
115 PUBLICATIONS   340 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Relationship Between Earnings Management and Accounting Frauds: A analysis of 1999-2015 of brazilians companies listed in BM&FBovespa View project

All content following this page was uploaded by Antônio Artur de Souza on 12 January 2016.

The user has requested enhancement of the downloaded file.


AN ANALYSIS OF COST ESTIMATION AND PRICING IN
MAKE-TO-ORDER COMPANIES

Antônio Artur de Souza, Ph. D.


Universidade Federal de Minas Gerais, Departamento de Ciências Administrativas - Rua Curitiba, 832,
30.170-120 - Belo Horizonte - MG, E-mail: artur@dedalus.cce.ufmg.br

Brian G. Kingsman, Ph. D.


Department of Management Science, The Management School, Lancaster University,
Lancaster LA1 4YX, England, E-mail: b.kingsman@lancaster.ac.uk

This article describes an analysis model of estimated and actual cost data applied
to (MTO) Make-to-Order Companies. It also describes an analysis performed with data
collected from a MTO company. The work described in this article illustrates the analysis
proposed to be performed by a decision support system called CEPSS, developed by the
authors. The main objective of this analysis is to investigate the relation between cost
variance and the confidence of the estimators when preparing the estimates (time
estimates). The analysis uses estimated cost data recorded according to a model developed
by the authors, which consists in recording three factors in association with the time
estimate: confidence factor (how sure the cost estimator is about the time estimate),
experience factor (how much experience the cost estimator has with the specific job), and
similarity factor (how similar the piece of work being dealt with is with others already
dealt with by the cost estimator). It was also an objective of this analysis to investigate if
the experience and similarity factors would help the confidence factor in the understanding
of possible cost variances. The recording of these three factors at the time of preparing the
estimates was found to be a useful mean of analysing and controlling cost variances in
MTO Companies.

Área: Gestão Econômica

Keywords: cost estimation, pricing, cost analysis

1. Cost Estimation and Pricing in MTO Companies


MTO Companies prepare cost estimates prior to bidding to determine the price to
be quoted, aiming at being the lowest bidder. The cost estimates must be low enough to
obtain work while still being at a price, within which the product may be manufactured. If
the cost estimates are too high then the price is set too high and no work is obtained, and if
cost estimates are too low then there is risk of losses. After bidding, the cost estimates are
generally used to monitor costs (cost control) during the manufacturing of the product.
For many MTO Companies bidding is everything. The more customer requests (or
enquiries) to which a company replies with bids, the more chances it has of getting orders.
Successful estimating and subsequent bidding require speed, accuracy and consistency.
Customers (or Contractors) when requesting bids often submit only rough design drawings,
from which the cost estimators have to develop estimates for the different expected costs.
The setting of a price is a very difficult task. Bad pricing decisions may lead a
company to bankruptcy, so they have to be made very carefully. The setting of a price is
strongly based on the cost estimates, but several factors have also to be taken into
consideration. The mechanistic approaches used to estimate the cost and set the price have
to be complemented by the cost estimators' judgement.
In MTO Companies the main operational phases in the manufacture of a product
are: 1) estimating, 2) pricing/bidding, 3) planning, 4) scheduling, and 5) cost control. These
operational phases are cyclic. This cyclic flow of information and operations does not really
start with the preparation of the cost estimates. In fact, when preparing the cost estimates
for a new product the company is using information gathered from past estimates and their
respective cost analysis. However, it is very difficult for MTO Companies to use past
information effectively. Usually there are several cost estimators in one MTO Company,
making it difficult for the estimates to be prepared homogeneously.
The cost estimation process in MTO Companies is in general very complex, due to
the nature of their operations. These companies make one-of-a-kind or special-order
products. In these companies the product unit costs are calculated by dividing total
manufacturing costs for each job order by the number of good units produced for that order
(Needles et al., 1994). The costing system in these companies requires the manufacturing
costs to be collected and assigned to specific jobs, rather than to set time periods.
In many MTO companies cost estimates are frequently obtained through what
might be called off-the-cuff or rule-of-thumb-methods. Sometimes these methods use
average historical costs. MTO companies almost always have initial estimates adjusted
according to the company's characteristics and plans, as well as according to the market
conditions and economic and technological trends.
Sometimes, cost estimates are prepared from incomplete drawings and without
complete specifications (Rich, 1993). Incomplete drawings and/or specifications often lead
to distortions in the cost estimation. For instance, surface finishing can be very expensive
and may be forgotten during the estimation stage. Very frequently MTO companies have to
prepare estimates quickly, risking making big errors. Tolerance requirements, for instance,
are a great difficulty for manufacturing MTO companies, because the estimation of the
precise times required become very time consuming and open to errors.
Historically, manufacturing companies (including MTO companies) have had high
variable costs because the production process requires many materials and extensive labour.
But as more and more companies automate their production facilities, variable costs are
substituted by fixed costs. These companies are becoming more capital intensive (Ricketts
& Gray, 1988). This makes the cost estimation process more complex, because it is very
difficult to estimate the indirect semi-direct and indirect costs. Activity-Based costing
systems are being used by many MTO companies in order to overcome this difficulty. All
MTO companies studied during the research used the cost-plus method for pricing, using
activity based costing to determine the cost used as the basis for setting the selling price.
Management uses this information together with their knowledge of the product, of the
production system and of the market, in order to set competitive prices. The cost-based
price in these companies is adjusted upwards or downwards when management takes into
consideration the sales demand and other factors which are very important for pricing
decisions.

2. The Data Collection


The data that was collected for the analysis represents the estimated and the actual
costs of products. The part of the cost data that was used for the analysis is related to the
times required to perform the different activities of the manufacturing process. The
company from which the cost data was collected is a medium size British metal-
manufacturing company. This company produces machines and equipments to several
industries, but most of its products are contracted by the nuclear industry.
The cost estimates are prepared when the customer submits an enquiry and asks
for a price proposal. If the customer accepts the price and contracts the company to
manufacture the product, then the product is manufactured and the actual cost is known.
The estimated cost may be revised after the customer accepts the price, when the
production planning is prepared and/or when changes are made to the specifications during
the manufacturing process. In the former case it happens because the production
procedures/manufacturing methods considered by the estimators are revised with more time
and details, resulting in some cases in modifications in the estimated cost. Changes occur
during the manufacturing process mainly when the customer decides that the initial
specifications were not correct. Figure 1 contains a diagram representing the life of a
contract, showing the phases in the chain of estimated and actual costs.

3. The Difficulties Related to Analysing the Variances Between Estimated and Actual
Costs
Due to the strategic importance of cost data to the companies it is very difficult to
convince a company to allow its cost data to be accessed by an outsider. Fortunately partial
access to cost data, enough to perform the analysis, has been gained in one company.
The Process of collecting the data was very difficult because the company does not
record them in an "appropriate" way. It was found that in the rush to complete one order
the chance to keep proper records of the costs (incurred to complete an order) is almost
always lost. Another reason for this lack of data, pointed out by the cost estimators, is the
effort by the shop floor managers to minimise overall costs spending as little time as
possible in recording the changes to the order and the actual times. Due to the lack of
integration between the different departments, they are not aware of the importance of
having the data properly recorded for future cost analysis.
One great difficulty in gathering adequate cost data was that the time the orders
take to be completed is quite long. The great majority of the orders take at least 6 months
to be completed from the time the enquiry is received. During this period the information
about the order passes through different departments, and by the time the order is
completed it is almost impossible to find the information required. To overcome this
problem it was necessary to contact and talk to people in the different departments. In spite
of having the support from the "bosses", it was very difficult to find people in the different
departments willing to help in collecting the data. After many interviews and broken
promises by many people it was possible to arrange with four cost estimators to obtain the
estimates and the actual costs.
Even with the co-operation of these four cost estimators and of three clerks, it was
possible to collect adequate cost data (estimated and actuals) for only four orders. It proved
to be a very difficult task. The estimates are recorded with one identification number for
each enquiry, and the actuals are recorded with another identification number (for the
contract). This makes relating the estimates with the actuals a very complex process.
The company does not record the actual costs in for each of the components, like
the estimates. Because of that it was not possible to make statistical analysis to find out
possible patterns of estimator's biases. None of the companies studied does record the
actual cost data in a proper way to make this type of analysis.
Customer Enquiry

Cost Estimation Process

Estimated Cost

Production Planning

Revision in the Estimated Changes in the Specification


Cost of the Product

Manufacturing

Actual Cost

Figure 1 - From the Cost Estimation and Pricing Process to the Actual Cost

4. Preparation of the Sample


The sample consisted of estimated and actual costs for two of the activities. Table
1 contains the list of manufacturing activities of the company. These two activities were
selected because they were considered by the cost estimators interviewed as two of the
most important ones in terms of difficulty in preparing the estimates. It was not possible to
collect data for more activities because that would have required too much time from the
estimators. In fact, the whole process of collecting the data, although for only four orders,
took more than two years (1993-1995).
During this period of two years much data was collected but was not adequate for
the analysis because it was incomplete or incompatible. Much data was incomplete because
it was either possible to collect only the estimates or only the actuals. Much data was
incompatible because:
1) After the preparation of the estimates, during the production planning stage (see Figure
1), changes were made to the specifications and/or design of the product. In many cases
there were actual costs recorded for activities not even considered during the cost
estimation stage.
2) During the manufacturing stage there were changes made to the product specification
and/or design. It was found that in many cases the changes were not recorded in any
way. It was only possible to see in the records that there were changes, but it was not
possible to trace the changes to the activities.
3) Parts of the manufacturing work were subcontracted.
4) The delivery time was changed after the manufacturing work started.
NATURE OF ACTIVITIES ACTIVITIES NAMES
Facility sustaining activities Management, Plant maintenance, Building maintenance, Heat
and Power, Communications, Sales, Accounts, Training,
Consumable Items
Product-sustaining activities Mechanical Design, Electrical Design, Process Engineering,
Product specifications, Engineering
Change notices, Product enhancements
Batch- activities Planning, Transport, Purchasing, Inspection & QA,
Supervision, Production Control, Stores
Unit-level activities Projects, Fabrication, Polishing, Turning, Milling, Boreing,
Fitting, Electrical Work, Testing, Sub-contracting
Table 1 - ABC Hierarchical Model of the Manufacturing Activities of the Studied Company

The preparation of the sample started with the design of a spreadsheet form to be
used by the estimators to record the estimates and the three factors: confidence factor,
similarity factor and experience factor (see De Souza & Kingsman, 1997 for details about
these factors). Figure 2 presents details about how the information about the three factors
are recorded in this spreadsheet form.
The design of the form was supervised by the chief cost estimator, and by the
director of the Mechanical Department, the latter being the departmental head responsible
for the estimates for the activities Assembly and Fabrication. The next step was to explain
the objectives of the analysis and of the data collection to some of the estimators, and then
explain the spreadsheet form to the two cost estimators who eventually agreed to co-
operating with the research. The estimates for the two activities started then to be recorded
by the two estimators directly in the spreadsheet.

4. Discussion of the Results


As mentioned before, it was only possible to collect data for four orders. The
estimates for these four orders were recorded directly in a spreadsheet prepared specifically
for the research. Estimates were prepared for dozens of enquiries, but unfortunately it was
only possible to get proper (adequate) data for four actual orders (orders that were
completed). On average, in the company studied only one in ten of the enquiries becomes an
order. The long time it takes for an order to be completed and the actual costs recorded also
contributed to the difficulty of collecting data for more orders.
Yet another difficulty in collecting this sort of data is the fact that for many orders
the customers ask for significant changes after contracting the job, and the difference in
price is agreed after negotiations without a formal cost estimation procedure. All of these
factors made it impossible to obtain any more data in the time available and the effort the
estimators were willing to provide to the exercise.
The initial estimates for the four orders are not fully presented in this article due to
its length (readers interested in this data should contact the authors). The summarised
estimated and the actual cost data and the modifications to the estimations made by the
production planning are presented in Tables 2 to 5. The additional information about each
of these four orders, such as how much time was spent in preparing the estimates, was
collected from interviews with cost estimators.

Confidence: from 0 (very high) to 10 (very low/bad)


Similarity: a - in size Experience: a - lot of experience
b - operational (function) b - some experience
c - manufacturing process c - little experience
d - raw materials d - no experience
e - completely new
Figure 2 - Legend used to record the information about the three factors

5.1. Order One


The first of the orders, numbered E22545, was a small one. It was prepared very
quickly because it was considered by the enquiry evaluator that the company had very good
chances of winning the order. Indeed they got the order, and it was a good one in the sense
that the overall actual costs ended up being significantly smaller than the estimates, as can
be seen in Table 2.
Assembly (hours) Fabrication(hours)
1) Estimates - Initial 79.5 2
2) Estimates - Changes by the Planning Dept. -33.5 +3
3) Final estimate 46 5
4) Actual Cost 33.5 8.25
Difference (4-3) -12.5 3.25
Variance on estimate (4-3) 27.17% 65.00%
Table 2 - Cost Summary for Order E22545

The initial estimates for the activity Assembly were significantly reduced by the
Planning Department after getting the order, when they had time to evaluate in more detail
how the job should be done. The actual cost for assembling ended up being 27.17% less
than the amended estimates. It is possible to see from the initial estimates that the time
estimator had a low confidence when estimating the time required for a component that
accounted for 75.47% of the total time initially estimated for assembly.
It is also possible to see that the estimator was not very experienced with two (of a
total of 20) of the components, and that 4 of the components were not similar to previous
products in terms of "function" (see the legend for the similarity factors presented in Figure
2). This is also likely to have contributed to the big cost variance of 27.17% (from the
estimated 46 hours to the actual 33.5 hours) for the activity Assembly. Based on that
variance one can say from that the estimators overestimated in order to compensate for their
lack of confidence in estimating the cost of some components, and also for not having much
experience with other components.
The variance in the cost for the activity fabrication was very big, but was not
financially relevant because the number of hours was very small. The confidence of the
estimators was not high, and this is likely to be the reason for the variance. For one of the
two components that required fabrication the estimator did not have much experience, and
the component was not similar to previous ones in terms of function.
It was learned from the interviews that in cases when the time required for an
activity is very small and the estimator is not confident, that the estimator's general practice
is not to bother to overestimate because he/she knows that the variance is not going to be
relevant. This was the case with the activity Fabrication presented in Table 2.
5.2. Order Two
This order, numbered E22177, had its cost estimated normally, but the confidence
of the estimators was considerably low. The product ordered was very different, as a whole,
from any other manufactured before. This can be noted from the experience factors.
The cost variance (Table 3) was very big for activity Assembly (63.77%), and very
small (3%) for activity Fabrication. The confidence factors were not very good for either
activity. As a matter of fact, it was worse for Fabrication. The overestimation for Assembly
can easily be related to the low confidence of the estimator. Considering that it was the
same estimator who prepared both the estimates for Assembly and for Fabrication, one can
infer that the estimator is more accurate in estimating the times for fabrication.

Assembly (hours) Fabrication (hours)


1) Estimates 221.5 446.25
2) Actual Cost 80.25 457.75
Difference (2-1) -141.25 11.5
Variance 63.77% 3%
Table 3 - Cost Summary for Order E22177

5.3. Order Three


The estimates for order 3 (number E22683) were prepared very quickly, and
because of that the estimates were not very accurate. It's possible to see from the summary
presented in Table 4 that the estimators overestimated in order to compensate for their lack
of confidence in the estimates. The estimates for the components that required significant
amounts of time had lower confidence factors because the estimators did not have the time
to evaluate in detail the required time.
For the activity Assembly, 69.70% of the initial estimates were prepared with a
confidence factor 3 or 4 (the confidence factor used varies from 0 -very high- to 10 -very
low/bad). 45.15% of the initial estimates for Fabrication were also prepared with a
confidence factor 3 or 4. It is possible to see that once again the restriction of time made the
estimators to overestimate to compensate for their lack of confidence.

Assembly Fabrication
1) Estimates-Initial 1420 1198
2) Estimates-Planning +84 -524
3) Estimates-Total 1504 674
4) Actuals 1045.75 477
Difference (3-1) -458.25 -197
Variance 30.47% 29.23%
Table 4 - Cost Summary for Order E22683

5.4. Order Four


The last of the orders, numbered E22355, was an exceptionally large order for the
company. The estimates were prepared very tightly, because there was strong competition
for the order and the company needed the work. The product was as a whole quite similar
to any produced earlier, as could be noted from the similarity factors attributed to the
estimates.
The estimates in the case of this order proved to be quite accurate, particularly
considering the size of the order. It was possible to see from the estimates that for the vast
majority of the components the estimators had a lot of experience. 63.4% of the total
Assembly estimates had confidence factor 1 attributed to them, signifying that the
estimators were very confident in most of the estimates. Moreover, the remaining estimates
for the activity Assembly also had high confidence factors attributed to them. The variance
for activity Assembly of 11.85% (Table 5) was considered very small by the cost estimators.
The activity Fabrication had a very small variance of 3.99% (Table 5), and reflects
the high confidence factors attributed to the estimates. Only one (of a total of more than one
hundred) component had a confidence factor 3 attributed to its estimate.

Assembly Fabrication
1) Estimates-Initial 1256.5 3075.25
2) Estimates-Additions 2243.5 0
3) Estimates-Total 3500 3075.25
4) Actuals 3914.75 3198
Difference (4-3) 414.75 122.75
Variance 11.85% 3.99%
Table 5 - Cost Summary for Order E22355

6. Conclusion
The analysis presented in this article made it possible to see the usefulness of the
factors recorded with the estimates. There is support for the hypothesis that the variance
between the estimated and actual costs is related to the lack of confidence by the estimators.
The results discussed in this article were presented to some cost estimators and
they found them very interesting. They said that this type of information would help them to
identify the activities for which estimates varied most, and then they could dedicate more
time to prepare the estimates for these activities. They also said that with this information
the estimators could identify their weakness (in terms of the activities that presented more
variances) and revise their estimation procedures. In other words, they said that with this
sort of information the estimators who presented significant variance could revise their
estimation procedures and then follow their progress with the information generated with
new analyses.
The analysis presented in this article, in spite of the difficulties with the collection
of cost data, was useful to show that a decision support system could be used to provide the
cost estimators with information that would allow them to perform the estimation more
accurately.

7. Bibliography
DE SOUZA, A.A. & KINGSMAN, B.G. (1997) “CEPSS - A Decision Support Tool For Cost
Estimation And Pricing In Make-To-Order Companies”, 21º ENANPAD, Rio de
Janeiro, Anais em CD-ROM.
NEEDLES, B. E., ANDERSON, H. R. & CALDWELL, J. C. (1994). Financial & Managerial
Accounting. Houghton Mifflin Company, USA.
RICH, J. E. (1993). Government and Contractors' Estimates: Needs and Intents. In: AACE
Transactions, pp. K.3.1-K.3.3.
RICKETTS, D. & GRAY, J. (1988). Managerial Accounting. Boston, Houghton Mifflin.

View publication stats

Вам также может понравиться