top of page

Top 5 challenges to cost estimating

Projects are a long-lasting feature of every industry. Whether they are multi-million dollar efforts or a neighborhood lemonade stand, the cost may not be immediately apparent. Creating an accurate, comprehensive, replicable and auditable, traceable, credible, and timely cost estimate is an important task. These estimates form a basis for analyzing alternatives, design tradeoffs, and make-or-buy decisions. They help our clients make better decisions faster.


However, generating quality estimates is not an easy feat. Cost estimators and analysts face many challenges that compromise the accuracy and quality of their estimates. Keep these 5 challenges in mind when building your next estimate:


CHALLENGE #1: Quantifying cost impacts

An engineer may need to reduce the cost of a component by 20% through design modifications. How do we quantify the impact of each design parameter for every alternative to meet the requirement?


Suppose we have a cost estimating relationship (CER) that is sensitive to the weight of the component, but we need to quantify the cost impact of using various materials of differing strength. Strength is our key design parameter, so the weight CER is not sensitive to alternatives based on strength. Estimators need to determine if cost is sensitive to that parameter and to determine how it responds.


Estimators must also analyze cost impacts for new products or approaches that are unknown or intangible. Quantifying the impacts of new regulations, human lives, or even time as a dollar value can be subjective but necessary to compare alternatives and make trades. You may be considering a move that will save you $300 in rent, but your commute will take an additional 30 minutes each way. How much is that extra hour of your time worth each day?


CHALLENGE #2: Resource constraints

Cost estimates must be timely! Time constrains many aspects of estimating, including data collection and validation, data quality, and consistency.

Analysts always want more data, but good data takes time.


CHALLENGE #3: Quality of available data

Resource constraints challenge the quality and quantity of data that estimators can obtain. If there is not sufficient time available, estimators may use secondary data sources, manipulated from the original source. Secondary data, especially that lack thorough documentation, have limited usefulness.


Additionally, the data estimators want for a credible, defensible estimate may not exist. If data do exist, analysts want more of it. Is there enough data? It is "good" data? Data of sufficient quality and availability strengthen the defensibility and provide statistical fidelity for estimates.


When you purchase a new car, you don't just buy the first car you see. You shop around for the best deal. You gather data – dealership sales, manufacturer promotions, consumer reports, safety ratings. You inform your decision with a variety of data types that provide context from a variety of sources.


Trying to get complete schedule, technical, performance, operational, budget, actual costs, and any other data an estimator could dream of from these sources is CHALLENGE #4:


CHALLENGE #4: Large number of organizations involved

Collecting data involves many different data sources and organizations of which estimators must be aware. Estimators need to communicate and coordinate with all of these sources. To be useful, the data must be accessible, so estimators and analysts may need to obtain proper non-disclosure agreements (NDAs) or to meet other security requirements for each organization. This process poses time constraints (CHALLENGE #2). If data are inaccessible, the quality of the estimate suffers.


CHALLENGE #5: Consistency

Once analysts collect all these data from the different sources through many time periods, they must make it consistent and comparable. Sounds simple enough – we made it through CHALLENGE #4, didn't we?


The challenge here is that these organizations are inherently different and processes, procedures, and structures have changed over time, even the value of the dollar has changed!

If you were to estimate a new phone model, you would start with a wide range of historical phone data. As an estimator, you adjust the costs for inflation. You verify that the labor costs for manufacturing each phone contain the same items. Are they unburdened? Do they include fringe? Is fee included? Over time the technical parameters of phones have changed. One phone has 64 GB of storage; while, another has only 260 MB of storage. The units must be consistent to be comparable – 64 GB versus 0.260 GB. The components of each phone are also different and so are the components of your new phone model. Analysts need to make the data comparable – all phones were not created equal. You can't assume that data for a flip phone from 10 years ago is comparable to that for a 2015 smartphone.


Organizations are investing time and money in building relational databases, containing a variety of data types from multiple sources and historical data in a consistent format, to alleviate these challenges. By reducing this burden, analysts have more time to work with high-quality data and focus on CHALLENGE #1.

Comentarios


bottom of page