People hire me to design things for them, and I’ve noticed that the most common types of contracts have serious flaws that result in disappointment for the supplier or the customer. So I started looking for a contract structure that makes sense for the type of work I do, and came across so many types, with so many potential variations that I was completely overwhelmed. Which is best? This was a cue for me to do my favorite kind of work – collecting and digesting tons of seemingly disparate materials and compiling them into something manageable.
In this article, I have researched, compiled, and analyzed over 30 types of contract models. I looked at their various benefits and downfalls, pulling out best practices and principles along the way. To wrap up, I’ll arrive at the contract type I feel is best suited to my line of work, that of engineering consulting and product design. Perhaps your industry has different needs and requires a different model.
Disclaimer – I am not a lawyer. Do not take my advice – consult your lawyer. Some of my information is likely wrong and I’m simply a small business owner looking for the best solution to my problem, so if you notice mistakes please point them out.
Problems and Best Practices
Contract problems I have experienced personally
- Ambiguous projects force a definite, fixed scope contract and the supplier loses their shorts.
- Buyers give free reign to trusted suppliers in T&M contracts, then are surprised when the project costs go far above their expectations.
- The supplier gives a 50% design review after months of work, and the cost of manufacturing the design is so high that it’s no longer economically feasible. Lots of money was spent and the whole project is dead in the water.
- Precious time is wasted managing and negotiating change orders, and the customer feels the pain of each one as his costs go up. I want customers feel positive feelings rather than negative ones when working with me. In some industries, ruthless back charges and change orders are the status quo and foster an us-versus-them mentality.
Contract Best Practices and Policies
Because of the problems experienced above, and after extensive research of contract models listed below, I adopt the following policies for contracts.
- Prioritize Success. Everyone’s number one priority should be delivery of a successful project. A successful project contract is, therefore, defined by three characteristics: minimize risk and exposure (liability) for both parties, flexibility to allow for change, and clarity regarding obligations, deliverables, and expectations. A lawyer’s priority when writing a contract is often to protect their client from getting sued – not necessarily to develop a high quality product at a reasonable cost and encourage healthy relationships.
- Incentivize Properly. The success trifecta is cost, quality, and schedule. Customer and Supplier should be incentivized to end projects early and reduce costs while maximizing the quality of the design.
- Share Pain/Gain. Risks and benefits should be shared more equally between Customer and Supplier. This cultivates a relationship of mutual problem solving and teamwork. The party who is willing to carry the most risk should also benefit the most from success. Customer/Supplier 100/0 splits are a recipe for disaster, and 0/100 the same. Anything in between (perhaps 80/20) is a step in the right direction.
- Avoid Fixing Budget, Scope, Timeline (especially all at once). Detailed lists of deliverables should be avoided in contracts. It leads to work on unimportant activities, and conformance to plans rather than cooperation to achieve truly valuable outputs. It also reinforces the false belief that product development can be predictably ordered, rather than a creative discovery. At least one of these factors (budget, scope, time) must be variable otherwise you’re fooling yourself.
- Embrace Flexibility. Reality rapidly intrudes on the best plans. Changing of requirements and specifications should be welcomed. Embrace uncertainty and ambiguity.
- No Rewards/Penalties. Do not accept contracts with rewards or penalties. They foster an us-versus-them mentality between customers and suppliers, rather than cooperation. These contracts are also contradictory to evidence-based management research.
- Build Clear Framework. Scope and acceptance definitions do not need to be set in stone from the beginning, however, a framework for acceptance criteria, scope (re)definition, responsibilities, and expectations should be clear in the contract.
- Mitigate for Complete Failure. While nothing is 100% predictable, a reasonable effort should be made at the beginning of projects to determine economic feasibility. Warren Buffet’s rule No. 1: Don’t Lose Money. Additionally, consider the possibility of a complete relationship breakdown and provide exit measures for both parties. A bad relationship is made worse if you can’t leave it.
- Iterate Regularly. Development projects should follow the scientific method cycle of iteration: plan, do, check, act. Each iteration should inform the next, and encourage feedback/communication between both seller and buyer on a regular basis. Both parties should strive toward a “fully functional product” at the end of each iteration. In this manner, fatal problems can be realized sooner than later, resulting in less exposure.
- Minimize Wasted Time. Tracking change orders, writing detailed proposals, managing back charges, overly extensive project plans, progress reports, negotiating anything at all – these are not value added activities and should be eliminated or minimized. See Engineering Principle 001 Maximize Value Added Work.
Key Concept: The Supplier Profit Slope
It took a lot of thinking to discover this one concept that I think many contracts get wrong. In most contracts where the price is not capped (T&M, Variable Price, Cost Plus types) the supplier’s profits continue to go up the longer the project drags on. This unnecessarily lengthens projects and increases costs because the supplier’s incentive is to continue the project as long as possible.
Some contracts, such as fixed profit attempt to solve this problem by fixing the profit. This is a step in the right direction but does not solve the problem.
If you want your contracts to be successful, both parties must be incentivized to end the project early. Since the customer already has an incentive to finish early and under budget, the supplier profits must, at the very least, be on a downward slope as the project goes on. Too far and the supplier will start cutting corners and reducing quality to meet the schedule.
Characters and Story
Since contracts are so much fun (sarcasm) and easy to understand (more sarcasm), I’m attempting to make this as simple as possible. We’ll use a couple characters in this story to illustrate the various contracts, and two scenarios for each: best and worst case. A robust contract should work well in both.
Our fictitious customer is a product manufacturer named Hero Products
Our fictitious supplier is a product designer name JM Design. Their burdened hourly rate is $100/hr.
Hero Products hires JM Design to design a product for them. To maintain some consistency across the various contract models, I assume Hero Products has a budget of $100,000, and JM Design hopes to make a minimum 10% profit on the job.
A mind-map of contract types I analyzed.
Fixed Price (Lump Sum) Contracts
These contracts are typically written with a definite scope of work (SOW) and the supplier quotes the work to the buyer at a fixed price. Often the lowest price gets the job, but sometimes they’ll evaluate on several factors such as capabilities and track record. Changes to scope are managed through change orders. The risk is (perceived to be) carried by the supplier since they must deliver the product no matter what happens. But what actually happens is different: the supplier starts severely cutting corners to save costs and ultimately the buyer is the one that suffers because the quality is reduced. Or, equally bad, the supplier bids the job far above the expected costs to ensure they can profit. Again, the universe is out of balance.
Firm Fixed Price (FFP)/Fixed Price Fixed Scope (FPFS)
The most common, FFP contracts have a fixed price and fixed scope. These can work if the job has been done before successfully and the scope is not expected to change.
Best case: JM Designs completes the $100,000 job for $80,000 and pockets $20,000. JM celebrates! Hero finds out through the grape vine they overpaid and aren’t happy. Worst case: JM Designs estimated poorly, goes over cost, and loses $10,000. It would have been better for JM to sit at home and do nothing. They quit their business to join a commune.
Fixed Price plus Incentive Fee (FPIF)
See also Target Cost Contracts below. The same as FFP, but essentially there’s a reward/punishment involved. “Fee” essentially means “profit” for these contracts. The intention is that the buyer and seller share some of the risk and can benefit from outperforming cost and/or schedule goals.
The price is split into two parts: target Fixed Price (FP) and target Incentive Fee (IF). Total = FP + IF. The final price is subject to a price ceiling, negotiated at the beginning. A formula to calculate IF is decided upon at the beginning that determines the profit at the end of the project.
I like elements of this contract type, because the supplier has incentive to reduce costs.
Target Total = Target Fixed Price + Target Incentive Fee
Actual Incentive Fee = (Target Fixed Price – Actual Project Cost) x Seller Share Ratio) + Target Incentive Fee
Actual Total Cost = Actual Project Cost + Actual Incentive Fee
Prior to commencement of work, Hero and JM agree to a target price of $90,000 (FP) and a price ceiling of $100,000. They negotiate an 80/20 Buyer/Seller share ratio, and the fee is $10,000 (IF).
Best case: Actual project costs come in at $70,000. IF = (90,000-70,000) x 20% + 10,000 = 14,000. Hero Products pays JM Design a total of 70,000 + 14,000 = $84,000. Hero saved $16,000 and JM pocketed a “free” $4,000. Everyone drinks beers. Worst Case: The project goes way over what everyone expected and JM Design stops working once project costs hit $90,000. The project isn’t done. IF = (90,000-90,000) x 20% + 10,000 = $100,000. Depending on how the contract is written, Hero Products must contractually pay JM Design $100,000 even if the project is not done, or JM must finish all scope of work items regardless of cost to them. Either way, someone goes home and cries alone in a dark room because their contract has failed them.
Fixed Price with Economic Price Adjustment (FPEPA)
Similar to FFP. Usually used with projects that span multiple years and/or projects that are based heavily on commodity prices. In theory, the contractor wouldn’t touch the job unless they had some protections against commodity inflation. The price is tied to inflation, the price of steel, or any other reliable financial indexes.
JM Design quotes Hero Products for $100,000 today and gets the job. The price is tied to the commodity price for steel. Steel costs go up 10% in a year when they’re finished, and the final bill is $110,000.
Fixed Price Variable Scope (FPVS)
This type of contract is used on projects where the supplier has a strict budget, but do not have a clear scope of work, or where the scope will likely change as the project progresses. It is best to prioritize the scope items from most to least important so the contractor will do the most critical items first. Less-critical scope items may never get done. It offers more scope flexibility than FFP contracts, but with the final price fixed there is no opportunity for the buyer to end the project early and save money. This contract model is advantageous for suppliers. Optional Scope contracts are a variation of this model.
Hero Products hires JM Design for an agreed upon $100,000 to work 5 requirements for two months. Best Case: JM works on the list of 5 requirements given by Hero, but halfway through the project Hero Products wants to change one of the requirements completely. No problem. JM Design shifts, completes them all, and keeps their costs under $100,000. JM Design pockets the remaining cash. Worst case: JM Design is only able to finish 3 of the 5 requirements (hopefully the most important 3), and the project still costs $100,000. The remaining 2 requirements are never completed.
Agile Contracts Primer pg 35 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Fixed Price Optional Scope (FPOS)
Time, cost, and quality are fixed, but the scope is allowed to absorb the uncertainty. This type of contract can be kept short. This is very similar FPVS models.
Our two parties agree on a $100,000 contract due in 2 months, and write up some quality standards for verification of quality work. Hero Products doesn’t have clear scope requirements, but that’s fine as long as there is excellent communication. Best case: Similar to FPVS, JM Design may only finish some of the expected scope, and there might be items outstanding that Hero wants to finish. Worst Case: JM Design does such a poor job that Hero Products never hires them again, and scraps all their work.
Fixed Price Variable Scope Progressive (FPVSP)
The same as FPVS, with multiple phases or iterations in which the scope is re-defined. These iterations could be as often as every two weeks, and allow the buyer to change the scope regularly to accommodate the changing needs of the project and/or organization. The customer is protected from too much exposure because they can they can terminate the contract at any iteration.
Our two parties set up an agreement the same as FPVS, with the addition of regular intervals every two weeks. They write the framework for managing each iteration: how requirements and scope are added/removed/reprioritized, how billing is performed each iteration, and mechanisms for acceptance criteria. Best Case: Hero Products begins by assigning JM Design its 5 scope requirements, but halfway through the project realizes that 2 of them are irrelevant, and replace those with 2 other requirements that are relevant. JM Design finishes the 5 (changed) requirements earlier than expected and Hero Products terminates the contracts when they get what they need. Worst Case: Hero Products makes the 2 requirements change as before, but JM doesn’t make the progress that was expected. Hero has the option to cut the project when their 3 requirements are met, or on the other hand, they can simply continue additional iterations until the work is complete using the established framework.
Agile Contracts Primer pg 33 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Fixed Price Per Iteration (FPPI)
This model is somewhat common among agile contracts, where the iterations (sprints) are a regular occurrence, often every two weeks or every month. Requirements are defined and agreed upon before each iteration, or there are no predefined requirements at all – the team simply works on the next highest priority item. You might say this model is a progressive FFP contract.
Fixed Price Per Unit of Work (FPPUOW)
Units of work can be defined in several ways: per story point, per feature point, per function point, or any metric of value to the customer. Points here refer to points as defined by Scrum or Agile methodologies. When a unit of work is accomplished, the supplier gets paid. The buyer can end the contract at any point when they feel they have what they need. In theory the point values should relate to business value impact, but this can be difficult to ascertain, especially without a history of previous projects to draw from.
I don’t see how this model can work unless the scope is well defined, and a full backlog of items is clear before the project starts. Why? If points need added during the project, the contractor will be incentivized to fudge the point assignments in their favor, bumping up the point count so they get paid more for less work. There is incongruity between point values and hours worked that cannot be easily resolved without a third party point auditor (read: more wasted work).
Agile Contracts Primer pg 27 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Cost Plus Contracts
Cost plus contracts reimburse the contractor for the expenses they incur, usually plus some agreed upon amount. This model is often used by general contractors and construction outfits. Similar in nature to Time and Materials contracts, but the profit is located outside the hourly rate.
Cost Plus Fixed Fee (CPFF) aka Fixed Profit
Expenses are reimbursed at cost, plus a fixed fee which is profit for the contractor. These types of contracts are typically used for exploratory or R&D projects. The contractor is completely insulated from escalated costs, but since their accounting books are visible to the customer they can’t inflate costs because of risk like they might in a fixed price contract. The customer needs to be comfortable with final project costs far in excess of their expectations, and have tighter mechanisms for ensuring the project is completed in an efficient manner. This model was used in the US Space Program.
Hero Products hires JM Design to make a prototype of a new product. The product is so new that they have no idea how much the technology will cost to develop, and agree to pay a fixed fee (aka profit) to JM Design of $10,000. Best Case: JM Design makes the product in record time and project costs add up to $70,000. Hero Products pays a total of 70,000 + 10,000 = $80,000 to JM Design. Worst Case: The technology needed to build the prototype is more costly than everyone hoped and project costs mount to $110,000. JM Design is paid 110,000 + 10,000 = $120,000.
Cost Plus Incentive Fee (CPIF)
Similar to CPFF, but the contractor is rewarded when they achieve certain agreed-upon targets by changes in the fixed fee. Also similar to FPIF, except charges are at-cost rather than fixed. The contractor is incentivized to reduce costs since they are rewarded/punished by cost overages.
Cost Plus Award Fee (CPAF)
Similar to CPFF, but the award is based on subjective or other technical acceptance criteria. This type of award structure often leads the contractor to cut corners or sacrifice quality to meet the award criteria.
Hero Products hires JM Design to make a test go-cart. If the go-cart meets a certain power to weight ratio goal then they get an award fee of $5,000. Best Case: JM Design makes the go-cart for a project cost of $70,000 and meets the power to weight ratio goal. Hero Products pays JM Design a total of $75,000. Worst Case: JM Design doesn’t seem to care about the project costs and project costs go up to $100,000. Of course, JM Design also finds a way to use thinner metal for the frame and meets the power to weight ratio to get the extra $5,000. Hero Products takes the go-cart for a spin (after paying the $105,000) and the frame cracks in three places.
Cost Plus Fixed Rate (CPFR)
Similar to CPFF, labor rates are determined before the job starts and are based on the contractor’s history and labor costs. The profit needs to be built into the labor rate. I was not able to find very much information regarding this type.
Cost Plus Percentage of Cost (CPPC)
Similar to CPFF, but rather than a fixed profit fee, they are awarded a percentage on top of project costs. This contract setup is easy for a seller to abuse, since they have no incentive to control costs, and are in fact awarded by increased costs. To me it seems a buyer would have to be stupid, or really trust their supplier to use this type.
Variable Price Contracts
I have noticed this terminology is used more in the software realm of agile and scrum, but the concept is very similar to T&M or cost plus models, if not exactly the same.
Variable Price Variable Scope (VPVS)
Essentially all aspects of the project are flexible: cost, timeline, and scope. The agreement can detail the framework on which the project is completed including the people deployed, the labor rates, and other factors. The customer can change the scope requirements at will, and assumes most of the risk in this model. There is a distinct possibility that the scope keeps extending further and further without an end in sight, which could be good or bad depending on the customer’s needs.
This model is essentially the same as T&M, but usually the supplier isn’t supplying physical materials.
Hero Products hires JM Design to build them a prototype car that runs on biofuel, and has no idea how long it will take or how much it will cost. Hero agrees to pay $100/hr for JM Design’s services. Best Case: JM Design finishes the design for the prototype car within a reasonable budget, and Hero Products wants to take the project the next step into production. It’s easy to extend the project and add extra features. No additional contracts are needed. Worst Case: Since there is no incentive for JM Design to finish early, they drag on for an extra month and charge an additional $16,000.
Agile Contracts Primer pg 33 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Indefinite Delivery Indefinite Quantity (IDIQ)
This model is used mostly for on-call services such as Architect-Engineering (A-E) and terms may last for years. The buyer usually does not know the quantity or timing of their needs. The terminology is mostly used in government contracts, and something around $20B/year is spent in IDIQ contracts by the US government. This model is very similar to VPVS. Within IDIQ, there are other models which I will not go into, such as multiple task single award, multiple task multiple award, and single task.
The city government of Townsville has ongoing yearly needs for architect and engineering (A-E) services for their government buildings. They don’t know the exact needs they will have, but do know that they will have A-E needs. They post a 3 year IDIQ request for proposal on their local online bidding system and JM Design bids on it and wins. JM Design then completes work on an as-needed basis at the terms agreed to in the contract.
Variable Price Variable Scope Progressive/Phased (VPVSP)
The “progressive” component of this contract means the project is completed in iterations. Typically these are shorter duration – two weeks, for example – and the scope can be adjusted by the customer each iteration. The progressive element helps keep the buyer and seller on the same page regularly since they are meeting bi-weekly at a minimum. If project costs begin to get out of hand, the buyer can adjust the scope to remove items, or make the call to cancel the project. Hopefully in this manner, if a project is doomed from the start, at least you can find that out sooner than later.
Often the intent in each iteration is to deliver a “fully functional product”. The customer should be able to end the project and walk away with a working product. This is a bit easier in software projects than physical projects, but the concept is to deliver something you can sell/use, even if it’s a very stripped down version. Perhaps the first iteration is the basic model, and subsequent iterations add various features that are nice to have.
Billing can run the spectrum of fixed price per iteration, T&M per iteration, a dollar value per point value, or other models.
Similar to VPVS, except the two parties agree to two week sprints. This requires more oversight time from the Hero Products, but since the contract model is iterative they are able to stop the project early and minimize their losses.
Agile Contracts Primer pg 33 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Capped Price Variable Scope (CPVS)
Similar to VPVS, this model is used when the buyer doesn’t have a clear scope of work, but does have a strict budget that cannot be exceeded.
Agile Contracts Primer pg 34 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Capped Price Variable Scope Progressive (CPVSP)
Adding the “progressive” aspect to a CPVS contract creates an iterative model where the overall project has a price cap, but the pricing per iteration can be T&M, fixed, or whatever you want.
Agile Contracts Primer pg 35 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Capped Price Variable Scope Progressive with Non-Binding Release Backlog (CPVSPNBRB)
What a ridiculous acronym, I love it. This contract type has both parties create a backlog of release goals before the contract is even written. The backlog is then added to the appendix and used to estimate the overall project costs, along with a common vision. The creation of the backlog is commonly created in a separate, preliminary contract. The backlog is non-binding, meaning the supplier is not required to deliver everything in the contract – which is really a benefit to the buyer since they can change the backlog at any time to meet the needs of the project. This model has a lot of cross over with target cost contracts.
Agile Contracts Primer pg 34, 39 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Target Cost Contracts
Target cost contracts were developed to build stable, long-term relationships with suppliers. Used by Toyota, these contracts are built with an initial planning step where the scope is identified as best as possible. A “target cost” is determined as a basis for writing the contract.
Customer and supplier work together to analyze all the potential project requirements and costs. The cost of changes in scope increases should be included to get as realistic of a cost as possible. The supplier and customer share intimate details of their cost estimates. From there a target profit is calculated, say it’s 15%. Finally, an agreement is made how cost under/overage will be shared, perhaps 60% buyer 40% seller.
Adjustment = (Actual Cost – Target Cost) x Customer Share of Cost Difference
Customer Payment = Target Cost + Target Profit + Adjustment
The buyer in this case will need to make sure they objectively agree with the proposed targets, since an unscrupulous supplier might fudge the numbers upward in their favor.
Once the project is started, there is an open-book policy and all cost information is shared in near-real time. At the conclusion, the profit is adjusted in accordance with the agreed-upon metrics for sharing costs. If costs are higher than anticipated, both parties share the pain, and if lower, both share the benefits.
It seems to me this model can only work when the costs can reasonably be estimated. Research and Development jobs with unpredictable costs are not a good candidate.
Target Cost (TC)
Hero Products wants JM Design to work on a medical device. In a preliminary FFP contract, Hero hires JM for $5000 to do a feasibility study and give their best estimate of the complete product development. Hero and JM work together on a budget and come up with a target cost of $86,364, and a target profit of 10% for JM at $8,636, a total target project cost of $95,000. At this point the contract is written and JM Design commences work.
JM Design’s actual costs end up being lower than expected at $70,000. Adjustment = (70,000 – 86364) x 60% = -$9,818. Customer Payment = 86364 + 8636 – 9818 = $85,182. JM Design’s profit ends up being 22%. Both parties pay less than expected and both benefit. The entirety of costs for Hero Products are $85,182 + $5,000 = $90,182
Side note: If Hero Products has determined through the feasibility study that the development costs must be under $50,000 for the product to be profitable on the market, they would know very early on that this project should get the axe.
JM Design can’t manage to keep the costs low and they come in at $100,000. (Although, it is important to note that JM is highly incentivized to keep costs low in this scenario.) Adjustment = (100000 – 86364) x 60% = $8182. Customer Payment = 86364 + 8636 + 8182 = $103,182. JM Design’s profit is a meager 0.3%.
Agile Contracts Primer pg 36 by Arbogast, Larman, and Vodde; www.agilecontracts.org
Customer/Supplier Capped Target Cost (CCTC/SCTC)
In this version, the customer or supplier’s costs have a maximum allowable amount.
Adjustable Target Cost and Target Profit (ATC)
An additional element can be added to target cost contracts which allows adjustments to the target costs or target profits by negotiation. It is recommended these changes are made on a moderate schedule to avoid changing too often, and a framework for making the change is in place before the contract is signed.
Adding on to the Target Cost example above, we add the ability to periodically adjust the targets. Best Case: Everybody realizes halfway through the project that they forgot a big scope item and the project is a failure without it. They estimate the cost of the added scope and adjust the target cost accordingly. Worst Case: same as earlier.
Time and Materials Contracts
One of the most advantageous model for suppliers, T&M jobs are simply charged to the customer at cost, plus a profit built into the unit rate. Material costs are billed to the job, typically with a markup (often 15-35%) to cover the overhead of purchasing, handling, and storing the material. This method is often used by lawyers, architects, contractors, and other consultants.
It doubly benefits suppliers because it relieves them of having to put together accurate estimates in proposals, which comes with a lot of stress, is time consuming, and is usually charged to overhead. See my engineering principle maximizing value added work for more on this thought.
These contracts can be good, however, where there is trust between parties and the client needs short work done at a moment’s notice. There is a beauty in the simplicity of these contracts, and they have their place.
Time and Materials (T&M)
As described generally above.
JM Design is hired by Hero Products to design a product at $100/hr. With no need for fancy scheduling and planning operations, JM starts immediately on the work. Best Case: The two parties trust each other and expect the project to cost $100,000, JM Design is an honest operation and finishes the project earlier than expected at $80,000. JM is not rewarded for finishing early and Hero Products gets all the benefit. Worst Case: The two parties still trust each other, but the project takes much longer than expected and JM Design sends a bill for $120,000. Hero Products is surprised the costs were so high and has to do a lot of work to manage to unexpected costs.
Time and Materials – Not To Exceed/Capped/Cost Ceiling (T&M-NTE)
Buyers who have experienced the abuses of T&M will be quick to add a “not-to-exceed” clause into their contracts. This is basically a capped price where the contractor cannot charge more than the cap even if supplier expenses go over it. Not surprisingly, these capped price jobs almost always hit the cap. Why would the contractor not slow down, work more hours, and extend the schedule if they think the customer has more money in their budget? Equally bad on the other side, if the contractor hits the cap and the job isn’t done they have no reason to continue working. For this reason, it seems to me that NTE jobs are inherently flawed – a band aid for a broken model.
Expanding the T&M example above, Hero Products puts a cap of $100,000 on the project. Best Case: Same as above. Worst Case: JM gets to the end of the $100,000 and the project is not finished. They estimate that the project will take another $20,000 to finish and Hero has no choice but to agree to the additional cost since they must finish. In this case, at least Hero Products has the option to spend the extra money to finish or not.
Time and Materials Fixed Scope (T&M-FS)
This model is essentially the same as regular T&M, but the buyer knows exactly what work they want done and puts together a scope of work before the job starts.
Time and Materials Variable Scope (T&M-VS)
This model is essentially the same as regular T&M, but the scope is written down and variable during their project.
Time and Materials Variable Scope with Cost Ceiling (T&M-VSCC)
This model is the same as T&M-VS, but a cost ceiling is added.
The possibilities of joint venture contracts are many, but outside the scope of this post. I’ve included it here since it is technically an option to get projects off the ground.
As if there weren’t enough contracts models out there, many of the sources I found embrace a multi-phase concept. The permutations are nearly endless. If you can’t find one that suits your needs – just combine two in separate phases to get what you want. One common example is to perform a preliminary feasibility study (product specification, market analysis, prelim budget, …) in the first phase using a FPFS, then move to a progressive T&M model once work commences.
Analysis of Contract Types
Using the best practices I laid out earlier as metrics, I put together a ranked list of all the contract types above. 1 is good, 0 is bad.
I realized there were two of my metrics that were mutually exclusive: you can’t have an equalized sharing of the project risk if the costs can’t be reasonably predicted. It’s a big problem, how do you share the risk on projects where the scope is highly variable?
There must be a compromise between these two mutually exclusive factors, and I think there is a solution: to reduce the risk to the supplier enough that they feel comfortable with the deal. I will discuss this below.
Conclusion: The Ideal Contract Model
The two models that offered the best combination of features were Adjustable Target Cost and CPVSPNBRB (despite the length of the acronym). I have decided to use these as a basis for what I call a Progressive Target Cost Contract. Here are the key mechanics of this contract type.
Before the contract is signed:
- An hourly rate is agreed upon for service types.
- A risk split is agreed upon. This is likely somewhere between 50/50 and 95/5 Customer/Supplier. 50/50 can be used in more predictable circumstances where the supplier has high confidence of success, while 95/5 more closely resembles T&M and is better for uncertain projects like R&D where the supplier doesn’t want to take on a lot of risk. 100/0 and 0/100 splits are not acceptable.
- A target profit percentage is agreed upon.
- A framework is established for quality control, acceptance criteria, protocol for addition/removal/change of scope items, roles and responsibilities, and expectations.
- A blank space is left on the contract for target cost, estimated in the next phase.
At this point both parties sign the contract.
Here’s an example of what happens for each type of split. Notice in the 50/50 split, the supplier profit slope is steep. What does that mean? The supplier really benefits from finishing early, but also really suffers if the project goes over.
On the 95/5 side, the supplier profit slope is not so steep. There is not as much punishment for the supplier if the cost goes higher than expected, but they will also not benefit as much if they finish early. This seems to be a better model for highly unpredictable scope projects like R&D. Notice the tail end of this split is longer – the supplier profit is positive longer than the 50/50 split.
- Time is billed hourly.
- A preliminary backlog of scope items is created in order to get an estimate of workload. The backlog is prioritized highest to lowest by value added. This backlog should be high level; it should be just detailed enough for a relatively accurate estimate. Too much detail would be a waste since it’s likely to change anyway.
- A target budget is established based on the backlog. The customer and supplier work together on this and both must agree by negotiation on the target.
- An iteration cycle is agreed upon – typically two weeks to a month.
Work phase (iterative):
- Scope is flexible and non-binding. Tasks can be added/removed/re-prioritized at will. Large additions of scope (or many small additions) should be accompanied by a re-negotiation of the target costs.
- The supplier works to deliver a “fully functional product” at the end of each iteration. A functional product, understandably, may not be not possible in early iterations but later iterations should work toward this goal.
- The customer can choose to end the project at the end of any iteration.
- The supplier can also choose to withdraw from the project at the end of any iteration, but they forfeit their share of the profit split if they do.
- The customer is billed T&M per iteration.
- Contract metrics (target cost, target profit, rate split, etc.) can be changed each iteration upon agreement of both parties.
- The project costs are calculated and the difference in budget is split at the agreed upon amount. The supplier and customer share the pain/gain based on their share of the split. If the project is under the target, the supplier gets their split of the remaining budget as a reward.
What Happens If?
What happens if the customer wants to end the project early?
The project can be concluded at the end of each iteration. As a protection to the supplier, they get their share of the split if the contract is pulled unexpectedly.
What happens if the supplier wants the customer to end the project early?
The assumption is that the customer already wants to reduce costs, so the supplier should work hard to understand the needs of their customer and give them what they want. If the customer doesn’t have what they need, then the project is not yet successful and more work needs to be done. In a worst case scenario, the supplier can pull from the contract and forfeit their share of the profit, or negotiate some other way out.
What if the customer doesn’t have what they need when the budget is reached?
The project can continue without major difficulties if the customer wants to increase the budget. The supplier continues to work T&M so their costs are guaranteed to be covered, but their profit margins will decrease the longer the project drags on.
What happens if the project is under the target cost?
The remainder of the budget is split at the agreed upon amount. This is the supplier’s reward for completing the job early.
What happens if there is a large change in expected scope and the target cost is no longer accurate?
The supplier should keep their wits about them and negotiate a change in the target cost. If the customer will not agree, the supplier can end the contract without reward, or continue working in hopes of gaining the
What happens if the buyer’s available budget is below the estimated target cost?
The scope should be reduced to match the available budget.
What happens if the project scope is so variable that it can’t be estimated at this time?
The customer and supplier should come up with a target cost anyway, even if they know it’s wrong, and the split weighted more heavily toward the customer (95/5). The target cost should be adjusted as the project progresses.
What happens if the project goes so long the supplier profit becomes negative?
The first thing to note is that the split was likely weighted too heavily toward the supplier. If the customer carried more weight then the tail of the project is longer (see the diagram above). At this point the supplier has the option to leave the project, which ethically makes sense. No one should be expected to pay to operate a business. Here’s one option: the supplier agrees to continue work for zero profit (not negative profit), but at the previous T&M rate or a reduced T&M rate.
Resources and Further Reading
Pfeffer, J., Sutton, R., 2006. Hard Facts, Dangerous Half-Truths And Total Nonsense, Harvard Business School Press
Austin, R., 1996. Measuring and Managing Performance in Organizations, Dorset House
Kohn, A., 1993. Punished by Rewards, Houghton Mifflin
Herzberg, F., 1987. “One More Time: How Do You Motivate Employees?” Harvard Business Review, Sept/Oct 1987
Agile Contracts Primer by Arbogast, Larman, and Vodde; www.agilecontracts.org