Single Award IDIQ normalization

[FLContracts]

I'm working on an IT help desk procurement and have a question about CPFF (LOE)  single-award IDIQ. The solicitation provides both estimated labor hours and required labor categories/mix that all offerors must use in their proposals. An industry partner has informally suggested this might constitute improper normalization under GAO case law, citing Information Ventures, Inc., B-297276.2 (Mar. 1, 2006).

Here's a hypothetical example: Say Company A proposes an innovative solution using automated ticketing and AI-powered initial response that requires fewer staff members, while Company B proposes a traditional manual approach requiring more staff. If the RFP mandates a required LOE for all to propose, it makes cost realism much easier. They state they are at a disadvantage since they need to propose required hours and mix, even though their technical solution could accomplish the same work with less staff.

My PCO has never seen this before. We are just setting a required max QTY and all work will be defined at the Task Order level. Thoughts? 

After reviewing Information Ventures Case and our single-award IDIQ structure, I believe our approach is distinguishable:

The hours in our RFP represent ceiling capacity over the contract period, not a fixed requirement. Each task order will determine actual hours needed based on requirements. Most importantly, technical approach is the most important evaluation factor, with cost/price less important. The ceiling hours don't restrict innovative approaches.

[Don Mansfield]

CPFF/LOE for help desk services? Why aren't you using FAR part 12?

[formerfed]

In your example, Company A has a valid point.  Actually I’ve been involved with many solicitations for help desk support.  The majority just include workload data with the solicitation, which should be readily available for the government to include.  Let offerors propose staffing resources based on their technical approach.  Some agencies include the governments LOE as information as well as a basis for their evaluation.  

[C Culham]

16 hours ago, Don Mansfield said:

CPFF/LOE for help desk services? Why aren't you using FAR part 12?

I think this is very important.  I think it has been forgotten that the Federal Streamlining Act of 1994 established a preference that agencies procure commercial services when they are available to meet the needs of an agency.   This preference is acknowledged in the FAR at FAR 12.101.  I think market research would suggest that help desk services can be acquired by firm fixed priced, firm fixed price with economic adjustment or time and materials contracts that are structured as an IDIQ.

[uva383]

Done a lot of CPFF term type awards, used the approach that you are describing and we included verbiage that stated if an offerer deviated from the government labor mix they were required to detail the rationale for the deviation or they were advised that their proposal would likely be found unacceptable. I can find the verbiage we used next week when I’m back in the office if you like. 

[FLContracts]

To clarify - this isn't a typical IT help desk. The requirement involves specialized Level 2/3 support for classified systems and secure facilities. 

This is why it's being procured under FAR Part 15 rather than Part 12. The specialized nature of the support, security requirements, and integration with government systems/processes makes this fundamentally different from commercial help desk services.

[joel hoffman]

12 hours ago, FLContracts said:

To clarify - this isn't a typical IT help desk. The requirement involves specialized Level 2/3 support for classified systems and secure facilities. 

This is why it's being procured under FAR Part 15 rather than Part 12. The specialized nature of the support, security requirements, and integration with government systems/processes makes this fundamentally different from commercial help desk services.

What do you intend to evaluate for a part 15 source selection and what is the basis of award? 

[C Culham]

Deleted and rephrased.

[C Culham]

12 hours ago, FLContracts said:

To clarify

Thank you....

I guess I might still wonder.....

PGI 216.104 Factors in selecting contract type.

See the policy tab for Principal Director, Defense Pricing, Contracting, and Acquisition Policy memorandum dated April 1, 2016, entitled “Guidance on Using Incentive and Other Contract Types,” when selecting and negotiating the most appropriate contract type for a given procurement.

[FLContracts]

The source selection includes two primary technical factors:

1. Technical Understanding of SOW - where offerors must demonstrate their knowledge of the specialized Level 2/3 support requirements, cyber security framework, and classified systems integration

2. Organizational Experience - focused on demonstrating relevant experience with similar classified system support and security requirements

Within the non-cost factors, Technical Understanding is most important, followed by Organizational Experience. Past Performance and Small Business Participation equal in importance but less important than the technical factors.

Most importantly - the non-cost factors combined are significantly more important than cost.

Going back to the original normalization question for a single-award CPFF LOE IDIQ: After reviewing the Information Ventures case and our structure, I believe providing labor hours/mix in this context isn't improper normalization because these are ceiling hours that represent maximum capacity over the contract period, not fixed requirements that must be used. The actual work will be defined through future task orders, where the contractor can propose efficient solutions using fewer hours. The technical approach is the most important evaluation factor - we're evaluating innovation and efficiency upfront. A contractor proposing AI/automation can receive higher technical ratings for innovation, execute task orders more efficiently, use fewer hours than ceiling, and be selected based on technical merit. Unlike Information Ventures where all proposals were forced to match a baseline, we're setting maximum qty of hours. The ceiling hours don't prevent selecting the best technical approach - they just establish maximum contract capacity to properly scope the IDIQ.

Thoughts?

[formerfed]

An offeror has an approach which they assert will provide a lower cost than your estimate.  But if you require them to propose something greater without any justification other it makes cost realism evaluation easier, you are on shaky grounds.  You are not truly evaluating potential cost of performance.  Instead ask offerors in their proposals to justify deviations from your estimates.  Do you have technical people that can evaluate that? 

[FLContracts]

@formerfed - You raise a good point about evaluating the true potential cost of performance. The challenge is that as an IDIQ, we can't precisely define all the work upfront - the specific requirements will be established through individual task orders. That's why we've set ceiling hours as a maximum capacity for the full IDIQ. This is a standard IDIQ structure that most of NAVWAR uses. 

[RJ_Walther]

11 hours ago, FLContracts said:

The actual work will be defined through future task orders, where the contractor can propose efficient solutions using fewer hours.

I think the industry partner has a valid concern that they will be unable to make it to the future task order stage where they would be able to propose an efficient solution using fewer hours because of the IDIQ evaluation criteria.

I assume an innovative, AI-driven approach is more expensive per labor hour (say, a technician supported by a multi-channel AI chatbot) than a more traditional approach (say, a technician sitting at a desk with a checklist), but will require significantly fewer labor hours to support the same number of customer service requests. If the AI approach is $200/labor hour and requires one FTE and the traditional approach is $100/labor hour but requires four FTEs but all offerors are required to price their proposal on the basis of four labor hours the traditional approach will have a much better evaluated price. You do say your ceiling hours don't prevent selecting the best technical approach, but are you going to be able to defend a 100% price tradeoff to go with the AI-driven solution (which is what would happen in my example, which don't think it is necessarily unrealistic for an AI chat-bot supported help desk versus a traditional help desk) against a technical approach that is obviously unrealistically-priced (i.e. the offeror with the AI-driven approach was forced to propose an unrealistically high number of hours which don't align with their technical solution) to GAO in the face of a protest?

You say that the "ceiling hours [...] represent maximum capacity over the contract period" and "[you've] set ceiling hours as a maximum capacity for the full IDIQ." I assume this is the "reasonable maximum quantity based on market research, trends on recent contracts for similar supplies or services, survey of potential users, or any other rational basis" from FAR 16.504(a)(1) but since your market research seems to indicate that labor hours don't align well with capacity because of the different technical approaches you may need to look for a better way to measure capacity and use that to create a more reasonable maximum quantity. For example, assuming you have historical information on the number of service tickets supported maybe extrapolate the historical growth in service ticket numbers over the past three years forward over the life of the IDIQ and add 10% as a safety factor. I'm not sure whether you could subsequently convert your solicited maximum quantity in service tickets to a maximum quantity in hours based on the awardee's particular technical approach at award, if that makes it easier to administer, but that could be worth researching too.

[formerfed]

2 hours ago, FLContracts said:

@formerfed - You raise a good point about evaluating the true potential cost of performance. The challenge is that as an IDIQ, we can't precisely define all the work upfront - the specific requirements will be established through individual task orders. That's why we've set ceiling hours as a maximum capacity for the full IDIQ. This is a standard IDIQ structure that most of NAVWAR uses. 

The way many offices handle this situation is using sample (hypothetical or actual) tasks which all offerors must propose against.  While not inclusive of all work, they are representative of the overall work covered by the solicitation.  

[joel hoffman]

12 hours ago, formerfed said:

The way many offices handle this situation is using sample (hypothetical or actual) tasks which all offerors must propose against.  While not inclusive of all work, they are representative of the overall work covered by the solicitation.  

Or could be an actual initial task. Of course if the actual work is classified or sensitive that wouldn’t be feasible.

[FLContracts]

@joel hoffman - Unfortunately we can't share a representative task order due to the classified/sensitive nature of the requirements.

[C Culham]

On 12/13/2024 at 10:19 AM, FLContracts said:

Thoughts? 

So here is what I think.  My thought is based on the continued discussion in the thread and a review of several GAO decisions that reference Ventures.

Your approach with regard to the RFP wording appears to move the issue of normalization to proposal preparation by each offeror to make your cost realism analysis easier rather than evaluating each proposal on its own approach and the probable costs (cost realism analysis) associated with each approach offered. 

If the government understands an offer why couldn't the government in doing a cost realism analysis create its own sample task and apply each offer to that task based on each offerors technical approach?

[FLContracts]

I appreciate your perspective.

I think there's a critical distinction between a completion contract, where the government is buying a defined deliverable by a unique technical approach, and a LOE contract, where the government is purchasing a specified level of labor hours. In Information Ventures, the agency normalized costs by adjusting offerors' proposed hours to match a government estimate without considering their unique technical approaches for meeting a defined performance spec.

At NAVWAR, we mostly do non-commercial software, and  for the software development, we are using a LOE IDIQ contract where offerors must propose a specified level of hours over the ordering period. We're not buying a defined deliverable, but rather labor hours to perform tasks within a general scope of work.

Given that, it's appropriate for us to specify the labor hours and mix in the RFP. We're not normalizing different technical approaches, but rather defining the very nature of the LOE IDIQ contract. Offerors still have flexibility to propose unique technical solutions within that LOE framework, which we will evaluate qualitatively under non-price factors.

It's impossible to predict every scenario over a 5+ year contract. 

At NAVWAR, we routinely use this approach for non-commercial software development because our requirements and the cyber threat evolve rapidly. Specifying the LOE ensures we have a predictable level of resources to meet emergent needs. If we believe this process is normalization, than LOE IDIQ non-commercial software contracts at NAVWAR, will need to significantly change their ways. 

[formerfed]

@FL_Contracts I think I got the wrong impression of what you’re doing.  I assumed from your initial post and the offerors question that your solicitation covered a rather specific and overall defined need.  You had an overall LOE that everybody had to propose against but an offeror questioned their technical approach reduced the overall effort required.  The overall scope is covered by the resulting contract and work is broken down by acseries of known task orders.  

But now I believe you are just establishing an IDIQ contract with specific labor categories and the LOC stated will serve as the max.  Then you anticipate a variety of tasks orders which will be negotiated on an individual basis with the contract labor rates.  I’m also assuming you don’t know the precise nature of future task orders, some of which may not even be known at the present,  so the purpose of the contract is to provide speed and efficiency in contracting for requirements as they arise.

Is that it?  If so, I don’t see any problems.

[FLContracts]

You're correct, I am establishing an IDIQ contract with specific labor categories and using the LOE as the contract maximum. 

The offeror's question about normalization gave me pause, as I want to ensure we are structuring the procurement appropriately and in accordance with applicable case law.

That's why I wanted to tap into the collective wisdom of WIFCON to see if there were angles I hadn't considered or potential pitfalls to avoid. Information Ventures and the concept of improper normalization are complex topics, so I appreciate the chance to clarify my approach and get the group's input.

[formerfed]

What you’re doing is not normalization in accordance with Information Ventures, Inc., B-297276.2 unless I’m missing something.  You’re not asking for offerors to propose a level of effort.  Furthermore you won’t be analyzing or adjusting offerors LOE or mix of categories.  

[C Culham]

On 12/18/2024 at 2:32 PM, formerfed said:

What you’re doing is not normalization in accordance with Information Ventures, Inc., B-297276.2 unless I’m missing something.  You’re not asking for offerors to propose a level of effort.  Furthermore you won’t be analyzing or adjusting offerors LOE or mix of categories.  

Ditto.  I just have to add that normalization in Ventures, as well as where I read that Ventures is cited in other cases has to do with proposal evaluation.   This said the pitfall could occur if you normalize during proposal evaluation.  You are probably aware of this but just sounding off anyways as it seems most protests are lodged against agencies for not following evaluation properly.    It sounds like you, with your additional detail, that you are not normalizing in the solicitation wording ergo the ditto.   

I hope the discussion helps you apply sound reasoning to your approach in case the entity providing feedback to you protests your approach during the solicitation stage.