In part 1, I touched on some of the dangers presented by estimating software along with some advice on how to work and bid safely. In part 2, I will be looking into estimating hazards that are uniquely human. A lot of frustrated and unsuccessful estimators get that way by overlooking human nature. While estimating involves lots of facts and figures, we must keep in mind that we are working with, and working for, people.
Submitted without comment…
In my experience, people define organizational policies according to their outcome. Bureaucracy generates lots of work that is peripheral to the task at hand. In contrast, Leadership aligns people and resources with the task at hand. Within the context of competitive bidding, effective leadership involves communicating expectations that are aligned with the interests of everyone involved. This starts with considering the interests of parties outside of the estimators office.
The four P recipe
- Perspective What do people expect to see? How does that compare with what they actually see?
- Predict How could people do things differently than you might have planned?
- Prepare What can you do to accommodate the inconsistencies, differences, and individual choices of others?
- Perform How can you coordinate the interests of everyone involved to maximize your odds of delivering a successful outcome?
Schools, academies, and trade associations promoting “best practices” in estimating tend to put great emphasis on process uniformity, deference to design professionals and obsequious devotion to every client request. While tidy spreadsheets and good manners are part of being a professional, they hardly define the estimators purpose.
Losing estimators are often telling me how they were “just doing their job” because “their hands were tied“. While some contractors do micro-manage their estimators, this mindset is more common among estimators who prefer to believe their job security is a function of avoiding accountability. If they were making and communicating the right decisions, they’d win more profitable work which is why the job exists in the first place. When a process interferes with your purpose, it won’t be the best practice to follow.
There is no more important safety rule than to wear your reading glasses
Estimates are used to compile and condense a great deal of information into a single number. Even the spreadsheets illustrating what’s going into the single number can be densely packed with information. Since everything must balance utility against clarity, the location of the information in an estimate is almost as important as the quality of the information.
Estimators working by hand are used to categorizing the information according to Construction Specification Institute (CSI) Masterformat guidelines. The Masterformat assigns a unique serial number to commonly encountered building materials arranged so that the materials generally align with similar materials that broadly align with trades. As anyone with experience in actual general contracting could tell you, the CSI divisions aren’t a good indicator of how the work is actually divided among contractors. For example, division 9, finishes may involve a dozen or more trade-level subcontractors, whereas division 14 Conveying systems will ordinarily involve only one.
There are a lot of estimating programs which are configured to organize the Quantity Take Off (QTO) according to CSI divisions. While this is great in terms of adhering to a standard, it doesn’t lend itself to compiling the relevant information to scope subcontractor bids. For example, there are a lot of “flooring” subcontractors (subs) that will install vinyl flooring as well as carpet, but they won’t do any ceramic tile or wood flooring.
The CSI codes place vinyl flooring and carpet in separate areas of the estimate that are often surrounded by completely unrelated trades. This means that on bid-day, the estimator is figuring out how to make one bid apply to two scopes of work which might be separated by hundreds of lines of information. When the deadline is fast approaching and low bids trickle through the door, this creates an arbitrary obstacle that can trip the estimator when they least expect it.
Lizzy was following the instructions perfectly, but then everything went sideways
If you’re using a spreadsheet program to compile your estimate, it’s possible to temporarily move relevant divisions to conform with the sub proposals that are coming in the door. Time invested in building a “working” worksheet that is linked to a “formal” estimate worksheet can make it possible for the estimator to have a streamlined layout for bid-day revisions, without sacrificing the uniformity of a formal layout.
Make sure it works seamlessly because spreadsheet errors on bid-day are serious problems. The “old school” approach to this problem was to print separate pages for every CSI division including a row for bidders and columns to verify, add, or subtract relevant scope items. Each of these sheets were put into binders with tabbed dividers. “Bid tabs” is industry slang for these comparison sheets which “show the math” for how the estimator scoped the bids of every relevant subcontractor on bid day.
Combo bids: Two for one, or double the trouble?
Subcontractors rarely specify which CSI divisions they’re bidding which means the estimator must not only sort the CSI divisions being bid, but must attribute them separately to their estimate. Despite all the heated rhetoric, the subcontractor (sub) is not an employee of the General Contractor (GC). GC’s can “demand” pricing breakouts from subs in direct proportion to the goodwill they’ve cultivated from fair dealing. GC’s cannot afford to ignore competitive bids from subs who are reluctant to provide breakouts that may be used to help a competitor win the job.
This means that GC estimators must be prepared to take the best subcontractor number they can get, even if it combines several “separate” scopes of work. Estimating programs will often generate error messages for any CSI Division that is left empty. If one bid applies to multiple divisions, most programs won’t allow the estimator to group them together. Instead, estimators are forced to use workarounds.
Let’s say a flooring sub’s bid for carpet and vinyl flooring is cheaper than any combination of independent carpet and vinyl flooring bids. They didn’t provide separate prices for vinyl or carpet because they want an “all or nothing” award.
The GC Estimator needs to enter the “combo” bid into the estimate but this raises several issues. Everything they enter in as a quoted value will be documented which means the Project Manager (PM) running the job will expect to find a subcontractor bid for the exact same amount in the bid file. If the estimator arbitrarily divides the quoted amount into plausible-looking amounts for carpet and vinyl respectively, there’s no bid in the file that actually matches either number. Now the estimator could put the entire bid amount into just one of the CSI divisions. That solves the problem of quoted numbers matching bids in the file. However, this causes two new problems. First, the default of most estimating programs is to “select” the lowest available bid in each entered quote. If the combo bid was entered into the carpet division, it would likely be higher than the carpet-only bidders because it’s also including vinyl flooring. This means the default setting for that division must be overridden in order for the estimate to select the combo bid. The second problem is that the vinyl flooring division needs to have a quote entered and most programs will not accept zero as a valid bid. Some estimators enter $0.01 for the quote as a workaround because no PM would go looking for a one penny quote for the vinyl flooring.
CSI Masterformat is tremendously helpful for design and management professionals who want a uniform system for coding information. Many Project Management programs include estimating functionality which not only imposes the CSI structure, but also includes the accounting structure for the job that follows. The estimating program’s lack of flexibility means that on bid-day an estimator might enter a one penny bid for a subcontract amount which later causes administrative issues in accounting and project management.
Breakouts are the leading cause of breakdowns
Alternates can multiply the estimators labor to an incredible degree. In their simplest form, Alternates are a request to add or subtract something to the project. In their most complex form, they’re a multi-dimensional problem that generates its own risk for the bidder.
For simple additions or subtractions, the alternate needs its own mini-estimate to address what’s going on. When the changes become more convoluted, the Alternate essentially replaces the original bid. Estimating programs may feature user-defined breakout tags which allow the estimator to sort, group, and compile the different breakouts into different schemes that reflect the alternate. Unfortunately, many estimating programs with breakout functionality are unable to compile multiple breakdowns into a cohesive estimate. This is very common for trade-specific estimating programs.
For example, let’s say there is an alternate which substantially changes the vinyl flooring scope. Some areas grew, other areas got smaller. As there are several alternates pertaining to the vinyl flooring, the estimator would have breakouts defined by the rooms involved.
Rather than a single line item for all the vinyl tile in that alternate, the program would output each room’s vinyl flooring separately. As silly as it sounds, some estimating programs will not compile the breakdown information into a cohesive estimate the way it does for an ordinary bid.
“With our new mirror technology you can double your horsepower!”
When GC estimators call the subcontractor wanting to make changes to the Alternates, the Subcontractor ends up going into intense “manual override” to answer relatively simple questions. The sub is usually under incredible pressure to answer quickly because the deadline is rapidly approaching. It’s much worse when the GC calls the sub whenever they are away from their desk, and unable to wrangle a simple answer from an obstinate program. I know of at least one competitor who guessed at a breakout price on bid-day that dramatically under-bid one portion of a project. That mistake was the first of many cascading events that ended in bankruptcy. Learn from their mistake, professional estimators do not guess! It’s much better to replace a lost opportunity than it is to “win” a project that imperils your company’s survival.
Bigger blocks, fewer breakdowns
One successful strategy to counteract an estimating programs clunky breakdown system is to use the definable breakdowns for complete alternates. Picking up on the earlier tile example, the estimator would conduct separate breakdown-level QTO’s for each alternate separately.
Let’s say there were four rooms pertaining to the base bid and two alternates.
In the base bid, rooms one, two, and four get vinyl flooring.
In Alternate #1, rooms two, three, and four get vinyl flooring
In Alternate #2 rooms one, three, and four get vinyl flooring.
This means that one definable breakdown would be named “base bid” and the estimator would conduct their QTO for the rooms as normal. Then the estimator would name a definable breakdown “Alternate #1” and would do a QTO for rooms one, three, and four. Note that this repeats the QTO of room four. Finally, the estimator would name a definable breakdown “Alternate #2” and would do a QTO for rooms one, three, and four. At this point, every room has been measured twice, and the vinyl flooring has arguably been estimated three times.
However, the estimator can now output their reports by the individual breakdown with all the pertinent information correlated normally. This means that the Alternates will display the total vinyl flooring as a single line item, tremendously simplifying the information you’re reading at Mach 6 when the GC calls. Only in estimating do we have situations where taking the long way around gets us to our destination faster than a direct path.
Quoth the vendor: “It costs more”
Quoted goods pertain to items with requirements that influence the price such as custom-built equipment. Some quoted goods are unique materials represented by an agent or a firm that promotes the material to design professionals to secure exclusive sales rights. Wherever competition and transparency are discouraged, artificial pricing hikes are sure to follow. As a result, the quoted goods can constitute an out sized proportion of the total estimate.
Quoted goods can be material exclusively, or they can be materials plus some labor or service. “Parts and Smarts” is industry parlance for a proprietary system of components that the contractor must install themselves, according to the design and programming requirements of the quoting firm. This is most common in fire-alarm and HVAC controls systems. “Turnkey” proposals are generally understood to be standalone quotes to deliver a completely built system. At the trade-level estimators desk, it’s critical to correctly attribute labor hours to the quotes you expect to receive.
Trade-level estimating program defaults can be very complex. For example, a fire alarm vendor supplying a “parts and smarts” quote will provide the fire alarm devices which the electrical contractor must install on a dedicated system. The electrical contractor is expected to furnish the junction boxes, conduit, and wire, for a fire alarm system that has not been designed yet. Estimating programs might have a “helpful” default for fire alarm takeoffs however they will only quantify the quoted goods. This means the estimator must carefully supplement the “fire alarm” takeoffs with all the parts and pieces that the fire alarm vendor omitted. Unless the whole system is attributed to a dedicated breakout, the quoted aspects of the fire alarm will be separated from the costs to furnish and install all the stuff needed to make the vendor’s quote work.
It’s good practice to conduct separate breakout estimates for any quoted goods that involve bidder groups with inconsistent levels of scope delivery. For example, the breakout combined with parts and smarts quotes can be directly compared to turnkey proposals.
Getting more information out of less data
Reading along, it would be easy to conclude that the best approach is “more breakouts”. Being better informed certainly helps when making decisions. To serve its purpose, the estimate must be a condensed explanation of what a project entails. Specifically, the estimate should reveal what is driving the cost, duration, and risk, of the project. I’ve encountered plenty of estimates that were so detailed that they buried the meaningful project attributes. This can be described as the “noise to signal ratio”. If you’ve ever been listening to a radio station when an adjacent station intruded, you can appreciate how difficult it is to understand what’s being said.
The Request For Proposal (RFP) may list alternates the owner requested alongside breakdowns the Architect wants to see. The intention and implication of each may serve different purposes which occasionally makes them difficult to understand.
I’ve seen projects with twenty or more breakout requests on the RFP get whittled down to three alternates in the course of a single exchange with the client at the job walk. Clients and Architects aren’t always considering the quality or the context of the information they’re requesting in the RFP. It’s often easier to generate a long list of things they might want, than it is to consider which things they would actually be willing to combine. There’s also a tendency to be additive rather than reductive when tasked with writing a wish list.
For example, lets imagine a project which is comprised of three connected buildings named A, B, and C. The client asks that all buildings be included in the base bid. They then ask for an alternate to move building B to the other side of building A, and to omit building C altogether.
Their second alternate request is to build only building’s A and B as originally aligned, omitting C altogether.
At this point, we’re up to three prices due on bid day. To bid them separately, all the estimating for buildings A and B would be repeated for all three prices.
In contrast, we could arrive at the same answers by answering two questions. What does building C cost? and “What cost difference is there in moving building B’s alignment with Building A?
That’s one breakout, and one alternate which is never repeated elsewhere in the estimate. More importantly, the estimate for building C generates 100% of the accuracy with 50% of the data compared to estimating A and B together. It’s probably a whole lot easier to review an estimate for building C against the drawings than it is to check a “combo bid” against multiple buildings. If your process is the same for all the buildings, the check on building C will be instructive towards determining if there are issues with your estimate for buildings A and B.
It’s also very significant to note that the building alignment question is pulled out as a line item cost. This allows careful consideration of what the result implies without the “noise” of building A and B’s total factoring in. I really can’t stress this enough because alternates are often sparsely documented by the design team. It’s fairly common for a complex alternate to be completely and exclusively defined in a few sentences on the RFP. What may sound like a simple “add this” or “take away that” alternate request can generate a long list of subtle consequences to the project. The knock-on effect for the client is sticker shock. Estimators who’ve carefully constructed their approach to reveal the subtleties are better equipped to present a solid explanation.
Savvy estimators will have already noticed that this advice could lead to a situation where you win the job and the client selects one of the alternates. Now when you go to hand off the estimate to a Project Manager (PM), you don’t have a single estimate which perfectly reflects the contract scope of work.
Your options will depend heavily on your software. In some cases, an estimator can copy the Building C breakout into the base bid and “multiply” the new version by -1 thereby generating a subtraction amount in all takeoffs. When grouped with the original total, and the relocation alternate, the output would be reconciled to the actual quantities needed.
Without question, this will require additional work, however it’s important to note that most estimators don’t win every bid. Spending a bit of extra time on those you win is an easy trade to make when you’re sinking less time into the losing bids. Negotiated agreement or “sure shot” bids should be done so that the estimates can be handed to a PM without confusion, rework, corrections, or delay.
Estimating is about controlling risk to secure profitable work. We can worry about risk created by the limitations of people and machines, or we can build our operations to accommodate them. I’ve provided a few examples to show how applying the four P’s can lead to opportunities that competitors only saw as obstacles.
For more articles like this click here
© Anton Takken 2017 all rights reserved