General.Zhukov

1) Installation can involve both construction & commercial (non-construction) services. Its not either-or. I suspect this is a key point here. You very likely need both. Legal may be correct in that preparing a building to receive a bunch of radiology machines is, indeed, definitely construction and severable. Altering a building. May need to enlarge doors, add RF/magnetic/radiation shielding, lots of electricity/power work, reinforce load-bearing structures because MRI machines are very, very heavy. etc. This work is very likely necessary, entirely separate from anything involving the actual equipment, and can be separately competed and contracting for. For example, adding RF and magnetic shielding - I'd bet a paycheck there are specialized construction firms that do exactly this, and there's probably a formal inspection & certification process. You are probably also correct. Any work that touches the equipment or could affect it's operation isn't construction. An MRI machine isn't a building or real property. Assembling it, QCing, testing, validating, calibration, maintaining, servicing etc. an MRI machine is commercial services, and obviously not construction. 2) You mention its a best practice for OEMs to oversee install. It's often more than that - OEMs of expensive/complicated/dangerous equipment often mandate how it must be installed. An OEM may state that they will only sell to you if you agree to their terms about installation - which may be very detailed and extensive. So legal can argue whatever they want, but if the OEM won't even sell you their MRI machine unless you agree that they (the OEM) will also install it, then it's a moot point about whether installation is separable. Maybe in theory it is, but if want to buy Brand X MRI, then Brand X is also installing it. Ask the OEMs. 3) Ask people who actually know about this. Run this issue by an actual radiologist, the building manager, an engineer who does this type of stuff, and an OEM rep or two. They will probably tell you immediately what must happen, what must not happen, and what you should do. Then you earn your paycheck by translating that into contracting. Footnote: I am friendly with a radiologist, and he knows his equipment & instruments like a chef knows his knives. He is also the King of radiology, and so would not be amendable to legal or an 1102 telling him about how he "must" go about getting his stuff installed. Your mileage may vary.

I am in the midst of placing an order (for software development) against a single award BPA, FAR 8.405-3 (C) (1). A part of the process is documenting what market research I have conducted. It has been a very long time since I've done this type of thing. I am now wondering - what type of market research do I actually *have to* do, given the procedures I am using? Any? A large portion of market research is about competition, commercial items, small business - all of that was taken care of when the BPA was (competitively) established. My impression is that my market research for order is basically stating my requirement is in scope of the BPA - "I have a need, and so I am using this BPA - which was specifically created to meet this exact need. The End." Is a part of market research explaining why I am using this BPA to get software development services, rather than any of the other innumerable sources available to me? I have my reasons, but I don't think I actually *have to* write down those reasons or justify my decision. (I could have met this requirement via - using some other agency contract/IDEV, a bilateral mod to add scope of my project to an existing order for software development, using one of many different IT GWACs/acquisition vehicles, etc.)

Agreed. 3-5 is about right. COs are well aware of the costs of getting too-much competition, and have techniques to keep the pool to a reasonable number. Some of these techniques are now endorsed by FAI, Periodic Table of Acquisition Innovations. Personal fav is the 'down select' - its like a voluntary and much simplified competitive range.

My agency, within HHS, is a heavy user of GWACs, and I love them. They are great and should be praised. CIO-SP3 & NITAAC are for IT. Purchasing IT services and software using a GWAC is drastically easier and faster than open-market. This is mostly because the GWAC has dealt with compliance and Terms & Conditions (all the stuff that needs to be done unrelated to actually meeting the government's requirement). Examples: Terms of Service. Does the publisher have federally compliant mandatory ToS? If not, you might have to negotiate a bespoke ToS, which is a big undertaking regardless of price. At a minimum you may have to check with legal about it, which brings things to a halt for a week or two. Pricing. A scientific software platform my agency uses a lot has something like 20+ different pricing models. Just figuring out how to accurately code the pricing structure as line items in a federal contract can be a challenge. When the software in question costs $5,000, this extra effort is not a good use of time. For the most part, the pricing in GWACs is federal government friendly. (Note: its been >20 years, we need to have consumption-based pricing in the FAR). FAR 25. IMO, this is the most complicated part of the FAR, by far. If your widget might have FAR 25 apply to it, you really want to get it from a source that's taken care of all that stuff. GWAC vendors are very good at policing each other. You'll hear about it if a vendor is offering you a cheaper but non-TAA compliant printer. Pricing? My guess is on a per-unit basis pricing on GWACs is about the same as open-market. But if you factor into your cost equation time and effort spent by all parties, then GWACs have big cost savings. A major benefit of having redundant (overlapping in scope) GWACs is competition for buyers. Its really important to have several comparable GWACs which we can pick from to procure stuff. Incentives matter.

yes. You haven't encountered policy people who respond "Good lord no, you have to state how pre-award IBR's don't apply when you're buying another $500K of Microsoft Office licenses."?

Yes, HHSAR does not explicitly say it has to cover "everything" in FAR 7.105. The actual text leading to the generation of thousands of APs full of mostly empty or otherwise useless content is found in the AP template instructions and the obscure "Directive for the Acquisition Plan." For all acquisitions not requiring SPE review and approval, and are equal to or greater than the simplified acquisition threshold, including task orders under Indefinite Delivery, Indefinite Quantity (IDIQ) contracts (FAR 7.105),OPDIVs may use this AP Template or their own template which must meet, at a minimum, the AP content requirements outlined in the FAR 7.105 and HHSAR 307.105. Which many people interpret to mean, everything in 7.105 must be covered by every AP. The HHS AP template, which we use in a slightly modified format, has 64 text fields, I counted. Gotta fill in all 64 text fields, every time, even if this task may be repeatedly pasting in "Does not apply". In some cases, the AP content does not ever apply- like, never ever, not even once. Its Pre-Award IBR, we don't do that and the AP has four questions about it. Yet, we paste in "does not apply" a few thousand times a year. Its dramatic overkill and makes the process ridiculous. Many thousands of hours wasted by poor PM's and CORs and such wondering "What is an IBR" and ultimately concluding, yet again, it doesn't apply.

Situation: My Department, HHS, mandates all acquisitions over $250K have a written AP which must cover everything in FAR 7.105. My OPDIV (unnamed, but hi Ted if you are reading this) essentially passes down this mandate unchanged. My OPDIV does perhaps 95% commercial fixed price contracts, mostly orders using various existing government-wide vehicles (GSA, NITAAC, Best-In-Class, etc.). We very rarely use FAR 15 procedures or award 'contracts'. Results: HHS's policy regarding acquisition planning is terrible for my OPDIV, given our actual workload as described above. Its drastic over-kill given the relatively simple stuff we do. Time-waste 1: Large swaths of 7.105 are never applicable or extremely rarely applicable to us, yet we have to include them in an AP, every time. Time-waste 2: We get thousands of acquisition plans a year. Want to buy $500K Dell Computers from NASA SEWP? You need a 20+ page Acquisition Plan. Somebody has to waste their time writing, and reading, all those APs. If everything is important, nothing is: Everyone knows much of the AP is completely unimportant, and leads to all involved discounting the importance and usefulness of the AP in its entirety. Ignoring important stuff: Actual acquisition projects and programs - which cover multiple contract actions - do not have acquisition plans at all. All the trivial procurements do, but not the project. For example, the acquisition project is to to acquire a very large Widget. Our big widget which consists of components A, B, & C, assembled together by contractor D, and operated by contractor E. That's five acquisition plans (A,B,C,D & E), but does the acquisition project to get the Widget have an AP? No. Questions: What the heck was HHS (or anyone involved) thinking? Is ignoring a thoughtless and counter-productive policy that bad a thing to do? How can the harms be fixed or mitigated, given my higher-up's narrow & literalist interpretation of the policy & the FAR? If you were involved in Department or OPDIV policy regarding acquisition plans, what do you think?

Yes. Circumstance matter. Prime A. I would like one widget please, delivered ASAP to Kwajalein Atoll, Marshall Islands. Sub: That will be $50,000/ea. Prime B. I would like one thousand widgets, delivery in six months, to the big depot across the street from your factory. Sub: That will be $10,000/ea.

The FAR itself doesn't prescribe exactly when a COR must be appointed for FFP. Note the very large caveat in 1.602-2 "as appropriate." So, per the FAR, the CO could reasonable determine a COR for your order wouldn't be appropriate. So the short answer is "No." However.. not end of story. Other regs, policies or whatever could mean you do need a COR in your situation. For example your department states you need a COR over a certain $ threshold, regardless of contract type. And your order is > $. Or something about that specific contract or project. Say, the Program Office wants all orders issued against a particular IDIQ to be assigned a particular COR, for centralized contract admin or whatever. So the longer answer is "No, but also maybe."

Well, this isn't my area of expertise, so I definitely could be wrong about this. I think the GVT must determine the probable cost. GVT cannot take the proposal at face-value, right? GVT may find that the proposal is, in fact, the most probable cost. But that is a result of mandatory cost realism analysis. In the case of an obviously low-ball proposal, GVT would certainly not agree with whatever the offeror states, and would find that the probable cost is higher (or has offeror has elevated risk, or both). Right?

That's not how cost realism analysis works. What the offeror gave you is cost information of some type. That is step 1 of a 3 step process. 1) Gather cost information 2) Do cost realism analysis, which has two big parts a) Identify understated costs/prices b) estimate probable cost. Defined as " the Government’s best estimate of the cost of any contract that is most likely to result from the offeror’s proposal." 3) Use results of cost realism analysis for evaluation and source selection. Taking cost info from the offeror, and stating "it is the probable cost and satisfies cost realism" is either, in my opinion: 1) False. Since the Government did not in fact generate a probable cost estimate, and did not perform the analysis. 2) True. Which is also false. Because if that statement is true, it means Government did generate a probable cost estimate and found the offeror's proposed cost to be exactly correct, and completely realistic. Which, of course, isn't what happened. Take a gander at https://www.dau.edu/pdfviewer?Guidebooks/CPRG/CPRG-Volume-4.pdf Chapter 8, which goes over Cost Realism is excruciating detail. Better yet, send it to your lawyers and have them read, no doubt for the first time.

Yes, this is correct. Even easier than a new (but functionally identical) order is adding a line item to an existing order via option exercise.

Assume you have no constraints or nuance of any type, can do whatever you like here, and are simply choosing between a) New order for CY2, or b) Mod existing order to add funds for CY2. Then I think a) is the better practice. Its simpler. The documentation is cleaner and easier to follow. Doing a mod, perhaps to avoid the paperwork involved in issuing a new order, is adding complexity. And that's something to avoid. Of course, there is no such thing as "no constraints." The most obvious one is what you referenced - Severable services contracts obligating appropriated funds cannot have a PoP exceeding one year.

Very skeptical about that IG report. A single manufacturer of specialized parts purchased mostly, perhaps exclusively, by the US Government. The buyer procures these items in small quantities as needed (so it seems from IG report). There is a large risk / opportunity cost for the manufacturer here. Accordingly, a large profit when it pays off. Don't like those high prices? Don't buy in the least cost-effective way possible. I doubt cost analysis accounts for the opportunity cost of the manufacturer. Must be some (likely substantial) cost for the manufacturer to keep making and storing these widgets, just so the (single?) buyer can purchase a few whenever they need them. Monopsony - Monopoly pricing is going to be messy and probably sub-optimal for one or both parties. Econ 101 here. Analogy: A family member has a really nice vintage Porsche. He needed some rando part for the engine. From what I gather, this part is exclusively used by a few models of Porsche made in the early 2000s. This part (forget its name or what it does, not a car guy) is literally a small piece of metal - no moving parts. Weights a pound or two. The material costs couldn't have been more than a few dollars. The Porsche mechanic in our area had one in stock, and charged (approximately, this was a few years ago) $200. $200 for a $5 part. $5 for the part. $195 for having a small retail warehouse of obscure German car parts located within a short driving distance. This would be excess profit, no?

This is a ridiculous case. The underlying but obvious and insurmountable constraint is the cost and effort involved in switching from Microsoft to a different platform. USDA can't afford to consider other alternatives. Likely the CO found this so obvious and trivial that they didn't consider stating it in the D&F.

As an addendum to this discussion, True story about the last time I had to deal with travel on a contract. Travel was a topic of discussions with the contractor and program office. It was all very confusing to them, and if I am being honest, to me too. We were going around in circles, not entirely unlike this thread. Ultimately, the contractor proposed that travel should just be FFP. We crunched numbers and came to something like $1,500 per "travel event." If the contractor flew on a private jet, we don't care. If they eat ramen and stayed in their sister's spare bedroom, we also don't care. Made program happy, because this makes travel like an optional FFP line item - something they know and understand (unlike ODC). This was a commercial open-market services & software contract valued at about a $1M. Did not use FAR 15 procedures. So far as I could tell, what we did was legal and compliant, and resulted in a fair and reasonable price. Although I wouldn't bet my paycheck on that.

OK, CO's aren't avaricious. They are risk-averse. Pro Tip 1: If we are talking about a commercial contract, don't call it "G&A." As others have noted, very explicitly cite 212-4 Alt 1. Pro Tip 2: When conducting price analysis for a competitive commercial contract, which includes travel as part of T&M line item, nobody cares about your "G&A" rate on travel. Honestly, just make it zero, and mark up some other T or M element (legally and following the rules, of course), so net change is zero. For the record, CO here who dislikes travel as ODC in my commercial contracts. Because: 1) The line item including travel is now T&M, which triggers some rather onerous procedures per 12.207 (b). In particular, getting HCA approval if the PoP > 3 years. makes the contract no longer fixed price. Executive dashboard KPI needles move, in the wrong direction. 2) If the conditions of 12.207 (b) (1) (i) - basically, it must be competitively awarded - cannot be met, then T&M, and therefore travel, is verboten, or its not commercial. 3) I do not like dealing with, and know very little about, indirect costs. For example, G&A vs. Material Handling. What's the difference? Do I even need to care for competitive commercial contracts? I'd rather not think about it at all. 4) My customer, the requiring activity hates, just hates, both obligating money upfront for theoretical travel which may never occur AND not obligating money upfront for actual travel which needs to happen but for which there is no longer sufficient remaining money.

This is what I was going to write, except its already been written. Amen. I have an action right now that I am about to award, this is its actual timeline: Identification of need -> 'acquisition package' to contracting office: Maybe 2 months Receipt of Acquisition Package -> Solicitation: 3 months Solicitation - Award: 2 weeks What is the PALT here?

I suspect DoD just wants better contracting writing software, and felt obligated to add a buzz word. I don't think this document would be much different if AI were swapped out with another trendy tech "Machine Learning CWS" "Blockchain CWS" "Robotic Process Automation CWS" Mark my words, some of us will still be 1102s when we see RFIs for "Open Source CWS" - My opinion is that this should have happened 20 years ago. "Augmented Reality CWS" "Quantum Computing CWS"

Asking for a PPQ from a reference is like asking someone to go out on a date with you. If they don't want to go on a date with you, you shouldn't keep asking them to go on the date with you. Ask someone else.

You're right about the regs. Its allowed (I think). But... The problem is not what any particular reg does or does not state, its what the COR thinks they state. Talking to the COR will be more productive than trying to prove a negative (that the FAR, et all. does not prohibit them doing a PPQ).

Irrespective of what you are buying, three methods to finding an appropriate PSC. 1) Ask the seller. If they are experienced with the federal government, they'll know. 2) Reference previous contracts for the same thing, and see what PSC was used then. In house contracts or FPDS. 3) Punt. Ask Program Office. If particularly important and difficult, ask for help from Legal.

This overlaps with the much ballyhooed 'Smart Contracts.' Which run on the blockchain, the blockchain I tell you, blockchain! /sarcasm Smart contracts execute pre-determined actions when predetermined conditions have been met and verified. Basically, (if -> then) contracts that are entirely outcome driven. Much research these days looking into how smart contracts can be used by Government agencies, and presumably some of that research is about funding and legal authority. Also, https://govlab.hks.harvard.edu/pay-success

See Veterans Healthcare Supply Solutions, Inc. B-409888. Quote, from GSA: "If the procurement at issue is for a single SIN, then it would stand that both/all members of a CTA would need to have the particular SIN on their respective FSS contract." My very unofficial understanding of this protest and how CTA's work is: 1) Despite the quote, my position is that a CTA member who does not hold the referenced SIN isn't necessarily ineligible. The CTA's proposal would need to demonstrate how the CTA member helps provide the solution. Like, for a given HACS need, I could see how a solution may involve closely-related SINs (like CDM or PKI) and have a CTA member who has those SINs, but not the cybersecurity. In contrast, I doubt a CTA member who has a unrelated IT SIN (say, A/V equipment) could legit join a CTA for cybersecurity. 2) This protest is from 2014, and GSA has changed how the schedules work since then, so this reasoning may not be fully valid in 2021. 3) The solicitation matters, since it can customize what types of CTAs can respond. For example, an RFQ may state that all CTA members must hold a specific SIN. 4) There are lots of protests about CTAs.

In terms of the basic legal part, you definitely CAN enact a contract mod to increase the value from $385M to $1B, under the right circumstances. Such as getting the funds to cover the increase, and getting God-level approval for a huge, non-competitive, out-of-scope modification (possibly an equitable adjustment?) that reflects poorly on everyone involved, having OIG parachute in, and other unpleasant things. Basically, you want stuff that you thought would cost $10. And now it seems like it will cost $30. As Vern notes, you have limited options: 1) Get another $20 2) Get only about 1/3 of the stuff you had planned on getting. 3) Pick up your marbles and go home (terminate, find a different means of acquiring the stuff). Quarrelling about whose at fault for the variance between $10 & $30, stalling for time, or being unhappy with situation won't change these options.