Founder in Eleven

One unique issue faced by professional services firms and IT consulting firms surrounds estimates.  Clients need services firms to estimate the timeframe for initiatives in order to justify its costs, many times relying on multiple vendors to engage in competitive bidding.   Professional services firms need to have a staff comfortable with a consistent methodology to create estimates and manage both the projects and client relationships.  This post will present best practices for estimating and working against those estimates to embrace an “Estimates are always wrong” mantra, which improves client relationships.

Estimating: A Background

Estimating is a key part of professional services as clients and firms want to know…

• How much time and effort will this project take for both the client and services teams?
• When will the project be done?
• How much will this project cost?

People make rough estimates every day, whether it be figuring out one’s commute or time for household chores. .  Understanding how to estimate requires understanding different variables that impact the estimate.  Good estimating engages with as many variables as possible and communicates those variables to all affected parties.  Here are three methods for coming up with estimates:

• Sampling: Based on previous experience, if task A took 20 hours and task C took 30 hours, and the team is trying to come up with an estimate for task B that is harder than A but easier than C, the estimate should be more or less 25 hours.
• Bottom-up estimating: What tasks are needed for this project?  Can the team estimate out all tasks from the samples, add the numbers, then come up with an overall estimate?
• Top-down estimating: Similar to sampling but on a larger scale, what projects are similar to this one?  How long did those projects take, and can we build an estimate on this project based on those similar projects?

While useful, these methods have their own issues. Top-down estimates tend to be too broad an estimate that is difficult use to manage a project due to lack of tasks and estimates per task.   Bottom-up estimates can either miss a task (underestimating), or result in assigning large numbers to every task (overestimating).  The best estimates combine the status and task-level predictions of bottom-up with the big-picture predictions of top-down estimates. 

Here are some more best practices for making estimates:

• Share your top-down/bottom-up assumptions with your client
• Include a surplus of input/reviewers (for potential missed tasks or more samples)
• Cross-foot all spreadsheets, since bottom line numbers can be wrong due to missing inclusions

Why Overestimating Is Worse than Underestimating

One analogy worth sharing with you team, particularly the geeky IT folks, is Scotty from the original Star Trek series.  When the captain would ask him how long something would take, he would always quote a time (say, 2 days) and the Captain Kirk would push back for something less (Scotty, you have 2 hours).  The thinking from Scotty’s perspective is “whatever number I give to the captain, he will  ask for something less, so I better start out with a high number.”  This approach can be represented by the question, “Should I give a larger number to make sure I have enough hours?”  Too often, the bigger number is justified in regards to being cautious or avoid the issues with underestimating; this includes:

• Difficult conversations with the client or management regarding the effort going over budget
• The professional services team writing-off of the surplus amount
• A client forcing to ask for an additional budget from management due to a low estimate
• An additional budget not being available as it has been allocated somewhere else

Given the above, it is easy to see the incentive to give large numbers to avoid conflict with the client and business, but worse issues that come about from overestimating. These issues include:

• “Roll-up overestimating,” where overestimated low-level tasks are overestimated in bottom-up estimating,are totaled together and then contingency is added to the overestimates, skewing the total even further..  Often times, roll-up estimating feels negligible (instead of 4 hours for this task let’s say 6 to be safe).  While the hours seem small, if this method is applied to all tasks, the total estimate could be as much as 50% higher than normal.
• Project cancellations due to high estimated cost.  Larger estimates also tend not to get funded because of the size of the project and the risk of large numbers.
• Forming overestimating habits in employees that are developed and enforced throughout the company, resulting in more overestimates.

Overestimating is worse than underestimating as it builds and reinforces bad habits from both the estimator and the person asking for the estimate.  The estimator, avoiding a difficult conversation down the line, puts extra time/cost into every estimate.  Others involved in the estimate might add their own overage before passing the estimate up the management chain. The person asking for the estimate might either trust it inappropriately or, like Captain to Scotty, always assume that the estimate is overestimated and push for reduced numbers.

The goal of an estimate is to be as accurate as possible (not high nor low).  I would always tell my team that overestimating could result in TSG not getting a project or losing competitive bids.  As the business owner, I would rather have an aggressive but accurate estimate, risk having a difficult conversation with the client or write-off an overage rather than have an overestimate that resulted in us not getting the work or the client not getting the internal funding.

“Estimates Are Always Wrong”

One best practice to help deal with conducting projects or initiatives based on an estimate is to get all resources, including clients, to understand that the majority of estimates are proven wrong over the course of time.  Some factors that drive differences in estimates and actuals are:

• Project Scope and Scope Creep
• Assumptions that are valid at the time of the estimate, but prove out to be incorrect over time
• Capability and availability of resources, both the firm’s and the client’s
• Difficulty in estimating unknowns
• Difficulty in estimating complex tasks
• Difficulty in estimating capability of staff and learning curve of staff
• Difficulty in estimating new technology and functionality
• Underestimating difficulties in data consistency

Creating a methodology and culture that results in a realistic top-down and bottom-up estimate (given the known assumptions) needs to be combined with a client communication plan (see my “Weekly Status” post) highlighting to all parties that the estimate was created with the best information, intentions, and experience available at the time, though both the consultants and client need to work together to achieve the initiative objectives, timeframe, and cost.

Embracing “Estimates are always wrong” relieves the stress of the estimator to come up with every variable and unknown when most times many data points are not known.  Another benefit of the “Estimates are always wrong” mantra is that it informs the client that the estimate is fluid and will demand their continued involvement to keep the project focused.

Plan the Work, Work the Plan

Given an “Estimates are always wrong” mantra, professionals and clients will understand the fluidity of day-to-day decisions that go into managing a project against its estimates.   Estimators will also be relieved of the uncomfortable discussion with management or the client after the estimate is shown to be wrong (which is always the case). With a quality bottom-up estimate, professional services should be reporting project effort and costs against the tasks in the estimate along with estimates to complete.  When individual tasks go under or over their estimates, or scope changes/assumptions are proven wrong, client and services can work together to address the issue and “work the plan.”  As presented in the Weekly Status Post, weekly and daily discussions with the client give the team the ability to adjust the effort based on any new findings.