Strategies for sizing UCD projects

When discussing strategies for sizing UCD projects with consultants, it quickly becomes evident that there seem to be as many strategies as there are consultants. This document will define and describe commonly used strategies, identify each strategy’s scope (i.e., whether it applies to design, research, or evaluation), suggest situations in which each strategy would be best suited, and identify pros, cons, and caveats to its use.

Detailed Description

The following are commonly used strategies, which are described in more detail under “How To”:

• Actual hours (time and materials)
• Best guess
• Low/high estimates
• Fixed price by project
• Fixed price by activity or deliverable
• Guidelines-based estimate
• Spilt sizing and pricing

Decision Criteria

These strategies aren’t mutually exclusive; they can be combined in a variety of ways.

Some hybrid solutions are discussed, but not exhaustively. Also note that sizing a project and determining what price to quote may also be, to some extent, mutually exclusive. Sizing refers to determining the amount of time (usually measured in person hours) a project will require. The consultant will use this information to then determine what price to quote for the project. The relationship between the two may be direct, or may be influenced by a number of factors that cause the cost to the client to either exceed or be less than the cost to the consultant. For the purposes of this discussion, we will assume a direct relationship between sizing and costing.

How To

Strategies and Best Practices

Actual Hours (Time and Materials). Perhaps the most straightforward sizing strategy is to make a “best guess” estimate of the actual number of hours an activity would require. The estimate would be based on experience or a “feel” for the amount of time required to perform an activity. However, the client would be expected to pay the actual cost of the project. This strategy can be employed for any type of UCD activity.

There are some caveats to the use of the Actual Hours approach. First, it is very difficult to anticipate all the factors that can affect an estimate (as we have discussed above). If the actual number of hours exceeds the estimate, this may cause difficulties with the client. A client may have assumed the consultant’s estimate represented the actual ultimate cost of the project. This expectation may be based on past experience: Often, consultants overestimate the cost of a project, and will ultimately invoice for this price even if the actual cost is less than the estimated cost (another strategy which we’ll discuss shortly). Therefore, the client will come to expect that the estimate is, in fact, an absolute. Deviations in either direction may be looked upon as disruptive to the budget process or otherwise troublesome.

Second, the market for UCD services is becoming more competitive. An Actual Hours estimate does not easily allow for discounts and other adjustments that might be required to bid competitively. It would be best to use such a strategy where there is a flexible budget, as the client would be expected to pay for the actual cost of the project, even if it were to exceed the estimate. It would also be a good strategy to use when you have a trusting or long-term relationship with the client. If the client believes that you are giving both an honest estimate of what is required, and believes that you are making your best effort to stay within your estimate, then this strategy provides a way for you to receive fair compensation for your efforts. Finally, when using this strategy, it is advisable to frequently inform the client of the running total cost to date, and any changes in the project that will significantly affect the estimate.

Best Guess. The final strategy is one in which a consultant essentially takes a guess at what a project will require in terms of time and cost. The guess is not based on formulae or guidelines, but rather a “sense” or “feel” for what is required for the project. This is likely based on experience and knowledge, but applied in a largely unstructured manner. Typically, pricing a project that was sized based on a guess follows a similar strategy. The consultant gets a feel for what the project is worth to the client and fixes a price accordingly.

An advantage to this strategy is that it is quick to perform. Detailed pricing breakdowns are not required. A disadvantage is that a client, upon receiving an estimate with no details about how the price is broken down, may not be confident that the consultant is competent. On the other hand, if the consultant has a reputation for performing high-quality work, the client may not be concerned over the lack of specificity in the proposal. Ultimately, whether this is a good strategy for the client or consultant will depend upon how good the consultant is at guessing. If the consultant severely underestimates the time required for a project, it may result in a schedule overrun. If he overestimates, he will reap more profit than with an accurate estimate.

Low/High Estimates. This strategy is a variation on the Actual Hours/Time and Materials strategy. In this approach, the actual number of hours is estimated and calculated twice: first, based on a best-case scenario and second, on a worst-case scenario. The best-case scenario is one in which it is assumed that there will be no unexpected situations that will affect the estimate and the project will require the estimated number of hours based on other similar projects. The worst-case scenario is one in which all possible factors that might affect the sizing are identified and accounted for. Both estimates are then provided to the client; the client is informed that the eventual cost will probably fall somewhere in-between the two estimates.

There are several advantages of the low/high approach for both the consultant and the client. Giving both low and high estimates protects the consultant somewhat by building in a “padding” that may or may not be required. If additional costs are required, the consultant will not be required to “eat” the costs him or herself; the costs can be passed on to the client. The client is forewarned about those possible higher costs and can budget accordingly. Also, if an additional activity is required, it has already been accounted for. There is no need to rework the estimate or create a new statement of work. Finally, giving a low/high estimate can help develop a trusting relationship between the consultant and client. A client who expects that the high estimate is in fact the actual quote will be pleasantly surprised if a lower cost is assessed. The client may come to trust a consultant that doesn’t automatically take everything (s)he can get.

There are also some disadvantages to the low/high estimate strategy. Experience seems to tell us that there is virtually never a project that isn’t subjected to unexpected situations. Therefore, creating a low estimate may be unrealistically optimistic. Further, meeting a cynical client’s expectations that the high estimate is in fact the “real” estimate may lead to the client’s eventual mistrust. The client may think the low estimate was simply a ploy to win the bid.

Another potential disadvantage of offering both a low and high estimate is that it might lead to a client’s attempts to negotiate each of the factors identified as those potentially causing the disparity between the low and high estimates. This may be especially true if the consultant chooses to share her sizing worksheet with the client. Detailed negotiations can be unduly cumbersome and may negate the benefits of the low/high strategy; therefore, sharing the worksheet with the client is not recommended.

The low/high approach can be used for all types of projects except for those that require an exact quote (for example, some government and other non-flexible RFPs).

Fixed Price by Project. The Fixed Price by Project strategy is one in which a consultant calculates and quotes a total price for the project. Regardless of how much time the consultant requires to complete the project, the eventual price tag will equal the estimate.

There are a variety of means by which a consultant can determine the price of a project. A consultant can create a detailed worksheet containing time estimates for each of the constituent activities required to complete the project. The consultant then multiplies the hours by a desired hourly rate to calculate the total price. Or, a consultant may set a price based on a typical price for a particular activity. For example, a consultant may have performed a given activity, such as a usability test, many times prior (see “Guidelines-Based Estimate,” below). (S)he can then use the price that has proven to yield the profit desired, and has been acceptable to clients. In any of these cases, a consultant may subject the estimate to a “sniff test;” that is, a subjective judgment to determine whether the price seems too high to be acceptable to the client or be too low to be profitable to the consultant. The price is then adjusted to maximize both profit and desirability to the client. Additionally, a consultant may perform market research to determine what a client would be willing to pay for a given activity, or what his/her competitors charge, and set a fixed price based on that information.

One advantage of a Fixed Price by Project strategy is that less time is spent tracking and reporting hours and costs against an estimate. There will be no need to “break the bad news” to the client and incur negative repercussions because an estimate was too low. However, it is useful to track hours for the consultant’s own benefit, as doing so can help inform future sizing efforts. Another advantage is that a Fixed Price by Project strategy can work to a consultant’s advantage if the project takes less time than expected.

On the other hand, a consultant can incur a considerable loss if the time required far exceeds the time (s)he originally estimated. Another disadvantage is that some clients try to get the most for their money, and thus attempt to extract additional tasks, deliverables, or iterations from the consultant for no additional cost. A related potential problem is that the client may interfere with the consultant’s ability to work efficiently. Excessive review, management, or iteration may instead bog down a project that could have been completed efficiently. The onus is therefore on the consultant to protect him/herself by writing a detailed statement of work with clearly specified activities and deliverables, including a description of a process for renegotiating the contract to account for any new client requirements or deliverables. Also, for equivalent projects, a fixed price estimate is often higher than a time and materials estimate because the contractor assumes more of the project risk.

A Fixed Price by Project strategy can be used for virtually all types of projects. It may be best suited for projects that have few resources for project management, projects with fixed budgets, and for projects for which the consultant has a good knowledge of and control over the factors that can cause a project to exceed its expected time and resources requirements. It is also more amenable to adjustment (e.g., extending a discount to the client), and thus is well-suited to highly competitive bidding environments.

**Fixed Price by Activity or Deliverable. **Another variation of the Fixed Price strategy is to fix a price not by the project, but by the deliverable. Each distinct deliverable is identified and priced separately. For example, a usability test is a deliverable. A presentation of results is a deliverable.

Advantages and disadvantages include all of those identified for the Fixed Price by Project approach, plus a few more. An additional advantage to this approach is that since activities and deliverables are smaller units than an entire project, it is easier for both the consultant and the client to “get their arms around” them than an entire project. The primary disadvantage is that a client may get the impression that they can simply pick and choose the activities or deliverables they want or can afford. A solution typically requires numerous deliverables that have interdependencies. Choosing among them is not simply a matter of deciding what is affordable. Consultants can help the client understand the activities and deliverables by explaining and documenting why each is necessary, and explaining the risks of eliminating them.

Guidelines-Based Estimate. A Guidelines-Based estimate is one in which a standard amount of time is specified for each activity. A total cost estimate is derived by multiplying the number of “units” by the amount of time and the consultant’s rate per hour. For example, a typical usability test may be characterized as requiring ten hours per user (including planning, observation, data analysis, report writing, and presentation of results). Using this guideline, a consultant can quickly estimate that a 6-user test will require 60 hours; he can then multiply 60 by his hourly rate to calculate the total cost estimate.

Another example is the “per screen” estimate for interaction design projects. If a screen is estimated as requiring 20 hours to design (including requirements gathering, lo-fi and hi-fi prototyping, and evaluation), then the estimated number of screens can be multiplied by 20 and the consultant’s hourly rate to derive a total cost estimate. For example, a consultant may add 20% to the estimate for a usability test because the code that is being tested is unstable.

Advantages of the Guidelines-Based method is that it is quick and easy to perform, and easy to communicate to the client. It projects an air of fairness: It is similar to, for example, going to a mechanic who always charges the same price for an oil change. A disadvantage is that the method is only as good as the guidelines used. That said, a consultant’s guidelines can be modified through experience and professional exchange. For example, a consultant may learn that the amount of time required to perform an activity changes as more units are performed and will therefore adjust her estimate to take this into account.

A guidelines-based estimate is best used for projects consisting of activities or deliverables that are readily containable (e.g., a usability test, a screen). Conversely, this strategy may not be ideal for large projects that include many activities that are difficult to delimit (e.g., giving advice, generating and evaluating alternative visual designs). It is best used for small projects in which activities are easy to identify and quantify. Small redesign projects are well-suited to this strategy because the deliverable is, to some extent, a known entity, as opposed to projects in which something new is being designed. It is more difficult to estimate the number of screens in a web site that does not exist than for one that is being redesigned.

**Spilt Sizing and Pricing. **Because user-centered design is a process that is, by definition, driven by ongoing data gathering, it is virtually impossible to know at the inception of the project what precisely will be required to achieve the project goals. Some of the most significant factors such as the user groups and their requirements are often unknowns early in the project. Once they are identified, however, the rest of the project can be more easily sized. Therefore, a sizing strategy that accommodates this split may be beneficial. A Split Sizing and Pricing model postpones the need to size the latter project activities until enough information is gathered upfront. Once the early user research and other early conceptual design activities are performed, the consultant can then size the rest of the project. Ideally, the first part of the project should be charged on a Time and Materials basis. The rest of the project can then be sized using any of the methods discussed here.

There are several advantages to the Split Sizing and Pricing model for the consultant. First, it allows the consultant to easily charge the client for project planning, something that is often performed without compensation. Second, it protects the consultant from having to commit to an estimate without first having the information required for an accurate sizing. Ideally, as much user research and early conceptual work as possible should be done during the first phase. Unfortunately, many clients may be reluctant to agree to this type of pricing. A client may require a total estimate upfront before agreeing to the contract, or prior to receiving internal approval to fund the work. Or, a client may feel that paying for the first part of a project without any idea of the cost of the second part effectively obligates them to keep the consultant on through the second part, regardless of cost. They may be wary that if the second part is too expensive, and they choose to work with another consultant, they would lose money or productivity by having to start the project again.

Split Sizing and Pricing is most applicable to projects that are end-to-end in nature, that is, that follow a project from its early stages through later data gathering or design activities. It is also better suited for very large projects because they typically have more unknowns.