Contingency when proposing IT service projects--the supplier’s viewpoint
This paper considers some of the supplier's challenges when assessing project contingency for service projects that have a large component of human resources for their completion. The interaction between contingency from the supplier's perspective and contingency from the customer's perspective is considered. A poor understanding of contingency from either party can result in project management difficulties during both the sales phase and the project execution phase. In the extreme this can ultimately cause one or other party to write-off significant project costs.
The importance of “understanding” is critical—understanding of the types of contingency, understanding where the contingency is held and who is responsible for it, and understanding how the contingency has been calculated and continues to be recalculated. This does not mean that a complex method must be developed; rather it just means clarity of the method is required.
To understand how contingency fits into a holistic view of the project, consider the relationship between contingency and estimate, risk, cost and profit.
Estimate—This is the expected cost of the resource(s) needed to undertake some work. This means that the work needs to be defined—the scope of the project must be clear. In fact, understanding of project scope can vary significantly, from “we want to do something like this,” to “here are the very detailed specifications and constraints.” For the supplier, the estimate is the estimate of the work they are bidding on. The customer's project may be broader in scope—there may be other systems that need modification. The importance of understanding how the baseline estimate was undertaken is absolutely critical to contingency planning, as seen later. This paper does not delve into how to do baseline estimating—there are many good references for this (Winters, 1999).
Risk—This refers to an uncertain event or condition that may occur, and if it does occur, will have an impact on the project. For this paper's discussion of contingency, risks are divided into two categories:
1. The risk that the base estimate for a specific piece of work is inaccurate.
2. The risk that some specific event or condition will occur that will adversely affect the base estimate, or introduce additional work activities.
The extent to which risk analysis has been undertaken influences the amount of contingency. In early parts of a sales cycle (for the supplier) or project definition (for the customer) there may not be time for extensive risk analysis, and the contingency planning must take this into consideration—there may be much lower confidence in the base estimates, and many unknown risks (unknown risks refer to events or conditions that have not been anticipated).
Contingency—This is an allowance made for the risk that something will not be undertaken with the planned estimate. Therefore contingency is always directly related to risk. A very important point to note is that for the supplier there must be the opportunity to not use the contingency. If it is known that the contingency will be spent, then it must be put into the base estimate. Suppliers sometimes miss this point, as addressed later, which can give rise to a false sense of expected profit.
Cost—This is the expected cost of undertaking the project. It includes both the cost of the base estimate and the cost of the contingency.
Profit—From the supplier's perspective, financial profit is the difference between what the supplier paid to do the work (cost) and what the customer paid for the work (price). Both the supplier and customer can profit in non-financial ways from a project—for instance a customer may be fulfilling a regulatory rather than business need. This paper just considers financial profit on a single project.
There are now two simple formulae for the supplier,
Cost = Base estimated cost + Cost contingency
Price = Cost + Profit
together with the statement that during the project execution, cost contingency is either spent or converted to additional profit.
It is remarkable how many companies fail to understand the above. Misunderstandings typically come in the form of failing to recognize that contingency is potential cost, failing to understand where and how contingency has been applied, and believing that profit margins can be maintained with a reduced price solely by lowering contingency.
Different Types of Contingency
A risk event, if it occurs, can adversely impact a project in several different ways. If these impact areas are defined, the contingency areas are defined. So what are the major areas in which a risk event can affect a project?
Quality—Lower quality may result, requiring more resource to correct.
Resource—More, or different, resources may be needed to undertake the work.
Schedule—More time may be needed to complete the work. This may, or may not, require more resources.
Cost—The consequence of needing more resources is more cost. (A need for different resources may, or may not, equate to more cost.) The consequences of needing more time may equate to more cost, if, for instance, payment milestones are delayed resulting in the supplier incurring cost of capital charges for the delayed payment. There may also be more cost if there are penalties for late delivery.
Scope—The features of the product may need to be altered; either added to, subtracted from, or changed. Typically more features are added. From the supplier's perspective this may be a result of the supplier failing to reasonably understand the scope of work, or understanding it but failing to plan for some of the scope. Either way, it results in a change to their baseline estimate.
Scope changes generally result in some form of change to one or more of resources, schedule, and cost. This paper considers the impact of additional resources as it affects cost, and does not delve into the subject of contingency planning for resource management across the supplier's organization. The remainder of the paper focuses on cost contingency—however schedule contingency is a significant topic that must be dealt with by both the supplier and the vendor.
Making the Contingency Visible
There are some basic challenges in arriving at a good estimate for contingency, one of which is summed up by the phrase “the less you understand your contingency, the more problems you are likely to encounter.”
Contingency that is hidden to a part of the estimating/management team of the supplier is very dangerous. Problems are caused by contingency hidden in the base estimates, done so possibly because the estimator wanted the best chance of meeting a deadline, or because they assume their managers will not add enough contingency. What needs to be addressed is creating an environment in which it is clear to all how contingency is to be applied.
Adding in contingency twice for the same activity typically occurs when there is a lack of trust, or hidden contingency. Again, an open approach to contingency understood by all involved helps address this.
Sometimes a supplier adds additional resource to a project team, stating this to be contingency. In this situation there is no potential to convert the contingency to profit since it is going to be spent regardless of whether or not it is needed, so the costs for this should be made part of the baseline cost, not the contingency allowance.
Some estimating approaches make allowances for variability in the base estimates, such as range estimating, and the use of simulation techniques such as Monte Carlo. As long as this is known, undertaken in a systematic manner, and accounted for when applying contingency at the overall project level, variability in the base estimates can be accounted for when determining contingency. Approaches based on range estimating at the project activity level may cover contingency for defined activities, but will not cover contingency for missed activities.
This section focused on approaches to applying cost contingency to projects. While discussing how to apply contingency, whether or not this results in a viable project is an issue not addressed by this paper. There are several different types of contracts between suppliers and customers, and they impact cost contingency in different ways. The two most prevalent types of contracts between commercial companies are considered, and how they affect the view of cost contingency.
Time and Materials (t&m)
How It Should Work
This should be as simple as the supplier provides people and the customer pays for all the time the people spend working on the project for the customer. The supplier's only risk is that of being unable to provide qualified staff—the customer owns all other risks. Hence the customer must have his or her own view about how much cost contingency to set aside. The customer must also set aside contingency for additional scope—this is a separate type of contingency.
How to Help Make It Work
Many times the above clean view of t&m work is clouded by expectations that the supplier is really on the hook to produce deliverables. The supplier must be clear in their own mind what they are really contracted for, and ensure that everyone on their team is consistent about this, and that this is communicated and discussed with the customer—this helps both the supplier and the customer understand where the contingency should lie (with the customer). Any disputes about this approach can be resolved before the work starts, rather than during the project. It is possible to create a hybrid t&m contract in which the supplier takes on more risk (and might need to set aside contingency for this), but this is challenging to do well and should be avoided. In most situations they are primarily the result of poor customer management; the aim should be to keep the contact simple and understandable.
There are some particular areas to be cautious about, where suppliers need to exert skill and manage the customer's expectations. Failure to manage these situations can result in additional cost for the supplier. However, with the exception of the warranty situation discussed next, the supplier should not have set aside project contingency for these events—they are primarily the result of poor customer expectation management. Rather, the supplier needs to become more adept at their business. These areas are:
1. The supplier recognizes that some additional resource is needed for the expected workload—this could be a piece of equip-ment, something as simple as a technical manual, or additional staff effort. It may be difficult to have the customer pay for this, since no expectation was set that this resource was needed.
Exhibit 1. Contingency Ownership Table
2. If there is a change in scope, and this comes as a surprise to the customer, they may be unwilling to fund the change in its entirety, having the perception that it is valid to hold the supplier partially liable for failing to foresee this.
3. The customer fails to meet an obligation and expects the supplier to work around this at no extra cost.
4. Some warranty is required at no additional charge. Sometimes customers demand, and suppliers acquiesce, that there be a warranty on work performed on a t&m basis. Although this goes against the concept of a pure t&m approach, this situation does arise. Any effort expended on warranty from the supplier's side costs money, and this money needs to be budgeted for. The correct approach is for the supplier to take this cost into consideration when estimating their profit from the project.
5. The customer is significantly dissatisfied with the work, and just refuses to pay some portion or all of the invoices. There are many reasons why this situation might arise, and when it does it leads to serious negotiations to attempt to rectify the situation and get the project back on an even keel. The result may still be that some monies are uncollected. Contingency planning for this kind of dramatic situation is generally not done on a project-by-project basis, but rather is done with general management reserves held at the corporate, rather than project, level.
Fixed Price (fp)
How It Should Work
From the supplier's perspective, this is as simple as “we provide products and/or services and the customer pays us a fixed fee based on an agreed payment schedule.” It is generally clear to all suppliers that in this situation there is a need to have cost contingency, however this does not mean that supplier's adequately account for it!
How to Help Make It Work
As with t&m contracts, there are some specific situations that require careful customer management by the supplier. Poor management results in higher cost to the supplier, and setting aside contingency for them is generally inappropriate—rather the supplier needs to learn how to improve their customer management skills. A couple of these challenging situations are:
1. The customer fails to meet an obligation, the consequence of which is additional cost to the supplier. For example, late review comments on documents, delayed delivery of test data, etc. The supplier will be more likely to be able to get and agreed change order if they have consistently reminded the customer of the need for timely delivery, and the consequences of failing to deliver.
2. Disputes over whether a change request is really additional scope requiring additional funding, or just clarification of agreed scope, which the customer is not willing to pay more for. The extent to which the supplier sets aside contingency for this type of risk is largely based on the extent to which the supplier believes the scope has been well defined, and their previous relationship with the customer.
Exhibit 1 demonstrates who owns contingency for t&m and fp contracts. The major difference between t&m and fp is in the ownership of overruns on the base estimates and warranty, which is expected given the nature of the two contracts. However, as noted above, if both parties are unclear about their responsibilities, additional costs can arise for the supplier, that could, but should not, be allowed for via contingency.
Approaches to Cost Contingency on Fixed Price Contracts
Let's take a look at some typical approaches to applying cost contingency to a supplier's financial plan for a project. This involves consideration of the two categories we have previously identified for risks:
1. How confident the supplier is in their base estimate.
2. The supplier's view of project risks.
The amount of contingency the supplier adds is significantly influenced by factors such as how familiar they are with the type of work being proposed, how much they understand the specifics of this particular project situation, and how familiar they are with the specific client. Some of this uncertainty is addressed in the base estimates. Let's assume that the base estimates are the best estimate at the present time for the supplier (by definition this is always the case, as time goes by the level of confidence in the estimates increases). A supplier can almost always give an estimate—the question is how confident are they in this estimate, how should they address contingency, and how valuable the estimate is if there is an extremely high degree of uncertainty.
There is a strong relationship between the base estimate and the overall contingency as mentioned earlier, but for now the two are separated.
Exhibit 2. Contingency Allowance Table
Given that a base estimate does exist, how might the supplier add in cost contingency? The following approaches are considered:
1. Single consistent simple formula.
2. Fixed percentage based on overall risk assessment.
3. Fixed extension of project team.
4. Variable formula based on a detailed risk assessment.
5. Variable formula based on history.
Single Consistent Simple Formula
The simplest approach is to add a set percentage to the base estimate, which can either be added to the effort (which is then translated into higher cost) or directly to the cost. Adding contingency to the effort requires the additional step of deciding how to cost the additional effort. Typically one of two approaches are used, either costing at the average cost rate of the staff on the project, or profiled so that the contingency effort is applied in equal percentages to each staff category on the project. This percentage is typically in the order of 15% to 25%. About the only advantage to this approach is that it is very simple, particularly if applied to the base cost. The disadvantage is that it does not account for differences in projects; some are inherently riskier than others, and the same project will have an estimate with a lower confidence factor earlier in the sales cycle than later.
Fixed Percentage Based on Overall Risk Assessment
This approach attempts to categorize the project into a set of predefined overall risk categories and apply a consistent percentage based on the risk, such as X% for low risk, Y% for medium risk, and Z% for high risk. As with the first approach, the percentage can be applied to the base effort or the cost of the base effort. Clearly the challenge here is to figure out how to categorize the overall project risk. Expert judgment is typically used. In the author's experience of proposing numerous fixed price projects, a general rule of thumb has been not to set the contingency below 10% nor higher than 25%. There is a certain risk premium for undertaking fixed price work, and even when the situation is very well understood, a 10% minimum is recommended. The 25% maximum is not to indicate that a higher percentage is inappropriate, but that if a higher percentage is appropriate, then serious questions must be asked about whether the project is well enough contained such that a fixed price approach is appropriate at all. The advantage of this approach is that with a company that undertakes many projects, this does generally provide a reasonable contingency factor, provided that a reasonable amount of risk analysis has been undertaken to correctly categorize the project. A disadvantage is that with only limited tying of risk analysis to contingency, an incorrect categorization may result.
Fixed Extension of Project Team
This approach is specifically geared toward short-term projects. As an example, for a project that has two people on it for six weeks, a possible delay of up to two weeks for both people may be considered a reasonable contingency. However, this amounts to 33% of the effort, and would therefore break the 25% upper ceiling mentioned in the previous approach. This demonstrates that the smaller the project, the more important it is to work with effort contingency and then translate this into cost contingency, rather than just focus on cost contingency.
Variable Formula Based on Detailed Risk Analysis
This approach is much more direct about tying contingency to risk. Project risks are identified, and then expected consequences and probability of the risk occurring are estimated. For each risk the probability of the risk is multiplied by the cost of the consequence. These are then added up to provide the overall project contingency. Exhibit 2 illustrates this approach. There are a couple of caveats to note. All the known risks are accounted for, an attempt made to quantify them, but what about the unknown risks? Expert judgment is used to make an estimated allowance for this. This is put into the risk list as a risk with 100% probability of occurring. The second caveat is to note that this is not an exact science, and good judgment must be applied to the calculated contingency amount. For example, if it is generally believed that the project is high risk, with a base estimate cost of $1,000,000, and a detailed risk analysis indicates a contingency of $50,000, something is probably wrong. The contingency would be expected to be higher. In this instance, the approach must be revisited.
The advantage of this approach is that it ties contingency to risk. In fact, a supplier may be able to use this risk table, or a variant of it, to discuss contingency with the customer, illustrating those risks that have the greatest potential impact on the supplier's costs. Working together on risk response planning, the customer may be able to undertake risk containment actions that help minimize the probability and/or impact of the risk, thus increasing the likelihood of project success. The supplier may in turn be able to lower its cost and/or increase its profit.
Another advantage of this approach occurs during project execution. There are often pressures on project managers to lower remaining contingency in order to show more profit. The approach of tying contingency directly to remaining project risks allows the project manager to justify the remaining level of contingency—justify first to themselves, then to their management.
Variable Formula Based on History
For those situations in which the work is very familiar to the supplier, and history has demonstrated that this work is very consistent in terms of following the base estimate, the contingency may be well understood as a percentage of the overall effort. However, each new customer is different (different people behave differently in similar situations), and this must be taken into consideration.
Contingency and Cash Flow
A supplier needs to understand their potential cash flow, in order to understand the cost of capital on money the project is borrowing. On large projects this can be a significant cost, especially in terms of a percentage of the planned profit. To forecast the cash flow, the supplier needs to forecast the potential contingency spend profile. Planning can range from simply allocating the contingency across the project duration in proportion to the effort profile, to defining time windows for each risk and using this to profile the potential contingency spend.
An example approach to a supplier's financial plan, demonstrating the link between risk and contingency, allowing for contingency cost profiling, and showing the impact on profitability, can be found on the Westford Consulting website, www.westfordconsulting.com.
Examples of a Shared Approach to Contingency
A few examples of real-life situations in which joint discussion of contingency occurred are as follows:
1. A supplier bid fixed price in a relatively open book manner whereby the staffing profile and major project risks were shared with the customer upfront. This allowed the customer to see and discuss the major components of project risk. On several occasions the customer undertook actions to reduce project risk, the supplier reduced the contingency, and the cost was lowered to the customer.
2. An internal IT department of a major global organization had been providing a single cost estimate to the business without contingency, and not tracking and forecasting their spend particularly well. They moved toward the “variable formula based on detailed risk analysis” defined above. They found they were able to have sensible discussions with the business about allowing contingency for potential cost overruns that were not related to scope increases (this was a first). However, they felt that they could not add contingency for the “unknown risks,” and had to work around this issue.
Real life is not as simple as we might like, but progress can be made toward a shared understanding of project risk that yields gains in project planning and allows for less conflict during project execution.
It has been the author's experience that contingency is a misunderstood subject for many people working for both IT supplier companies and for customers. This causes many problems during the project execution stage. With a drive to more fixed price contracts during recent years, these problems are exacerbated. Understanding is the key to a more accurate depiction of contingency, and to improved project performance. First the supplier needs to understand how they are applying contingency, then the supplier and customer can develop a shared understanding of project risks to minimize them, reducing the need for contingency, leading to a greater likelihood of project success.
Winters, Frank 1999. The Estimation Process—A Discussion. Project Management Institute Consulting SIG Newsletter Volume 1, Issue 1.
Proceedings of the Project Management Institute Annual Seminars & Symposium
October 3–10, 2002 • San Antonio, Texas, USA