Faster RFPs Won’t Save the Future of PBM Procurement

Procurement technology has become very good at making Request for Proposals (RFPs) faster. Digital platforms can distribute questions, collect responses, manage deadlines, coordinate evaluators, and organize documents. Artificial intelligence can draft questions, summarize proposals, extract information, identify differences, and accelerate analysis which leads to Faster RFPs. Vendors can use similar tools to prepare responses more efficiently.

These are meaningful improvements, but they raise a more important question for the future of Pharmacy Benefit Manager (PBM) procurement: What if some of the work we are making faster no longer needs to happen at all?

For decades, procurement technology has largely focused on improving the execution of an established process. Paper became digital. Physical delivery became email. Email gave way to procurement platforms. Now AI is being layered onto those platforms to automate increasingly sophisticated tasks. Each generation has made procurement more efficient, yet much of the underlying information flow remains remarkably familiar.

A requestor creates questions. Vendors assemble responses. Information is submitted. Procurement teams and consultants organize it. Different formats and terminology are normalized. Responses are compared. Decision-makers receive the resulting analysis.

Technology keeps getting better at moving information through that process. The future of PBM procurement may depend on asking whether the information needs to make that journey in the first place.

Six Years Ago, We Were Still Shipping Binders

Not long ago, the inefficiency was much easier to see because it was physical. Members of the Rapid Request team remember PBM procurements before the COVID-19 pandemic in which completed RFP responses were expected to be printed, organized into binders, reproduced in multiple copies, packed into boxes, delivered to the UPS Store, and physically shipped to the requestor by a specified deadline. Failure to meet the submission requirements could mean disqualification.

This was not a procurement practice from the distant past. It was happening only about six years ago. Moving those submissions to email was unquestionably an improvement that made faster RFPs. Hundreds of pages no longer needed to be printed, assembled, copied, packed, carried to the UPS Store, and shipped simply to put information in front of a requestor.

Rapid Request’s CEO half-jokingly points out that, for PBM procurement, email may have been the industry’s biggest process innovation in 60 years. That is an obvious exaggeration, but it captures something important.

The delivery mechanism changed dramatically, while the underlying information process did not. The requestor still created the questionnaire. PBMs still recreated information in response. The requestor still received separate documents that needed to be organized, interpreted, normalized, and compared.

The paper disappeared. The process remained.

Today, procurement is undergoing another technological transition, and this one may be considerably more powerful than replacing shipping boxes with email. The question is whether we will use that technology to rethink procurement or simply to make faster RFPs.

Faster Does Not Necessarily Mean Modern

There is nothing wrong with faster RFPs. If technology can reduce the time required to create a questionnaire, distribute it, collect responses, coordinate evaluators, or perform analysis, organizations should benefit from those efficiencies.

Artificial intelligence expands those possibilities significantly. An AI system can potentially summarize hundreds of pages in seconds, locate information buried in attachments, compare narrative responses, identify missing information, draft follow-up questions, and help procurement professionals navigate information sets that would otherwise require hours of manual review. Those capabilities can create real value.

But speed and modernization are different objectives.

Imagine a PBM already maintains detailed information about its clinical programs, networks, service model, implementation capabilities, reporting, organizational structure, and other foundational characteristics. A requestor issues an RFP asking for that information, and the PBM uses AI to help adapt and rewrite its existing information into the requestor’s questionnaire.

The requestor receives the response and then uses AI to extract that information from the PBM’s documents, summarize it, reorganize it, and compare it with information submitted by other PBMs. The technology on both sides may perform its job extremely well.

But step back from the technology for a moment. Information existed, was transformed into an RFP response, was transmitted, and was then extracted and transformed again.

Making every transformation faster is useful. Eliminating unnecessary transformations is potentially much more valuable.

EXECUTIVE PERSPECTIVE
Automating unnecessary work makes the work faster. Modernizing procurement asks whether the work needs to exist.

We Are Using AI to Solve Problems Technology Can Eliminate

AI is particularly good at making inefficient information structures more manageable. That is both its strength and a potential distraction.

Unstructured documents create enormous amounts of work in procurement. Vendors describe similar capabilities differently. Responses appear in different locations. Definitions vary. Financial information may require normalization. Supporting evidence may be spread across spreadsheets, attachments, exhibits, and narrative responses. AI can help make sense of that complexity, but there is another way to reduce it: structure the information before analysis begins.

If comparable vendor information already exists within defined fields and categories, technology does not need to spend as much effort discovering where the information is located or determining which pieces correspond to one another. That distinction matters enormously for PBM procurement.

A modern PBM evaluation can involve financial guarantees, formulary strategy, pharmacy networks, specialty pharmacy, clinical programs, implementation, reporting, member experience, contractual terms, organizational capabilities, and many other considerations. Some of that information genuinely needs to be customized for a particular requestor, but much of it does not.

A PBM’s fundamental service model does not become a new service model every time another employer asks about it. Its organizational capabilities do not cease to exist between RFPs, and its clinical programs are not created when a questionnaire arrives. Yet the traditional RFP architecture frequently treats information as though procurement must create it before procurement can evaluate it.

That assumption made more sense when information was difficult to maintain, distribute, secure, and search. Technology has changed those constraints.

The Better Question Is Why the Information Keeps Starting Over

Most procurement processes begin from the requestor’s side. The organization identifies a need, develops requirements, constructs an RFP, identifies potential vendors, distributes questions, and waits for responses. Only then does the requestor begin receiving detailed information about available solutions.

There is an obvious historical logic behind that sequence. Without a practical mechanism for vendors to maintain detailed, current information that qualified buyers could securely discover and compare, the request itself had to initiate the information exchange.

Modern technology makes another model possible. Vendors can maintain information continuously while access remains controlled. Information can be structured consistently and updated without recreating an entire proposal. Qualified requestors can search available solutions, comparable characteristics can be viewed together, and analytical technology can work with information that already has defined relationships.

The starting point changes from collect first, understand later to understand what already exists, then ask what is still necessary.

That does not eliminate procurement-specific questions. It makes those questions more valuable. Instead of spending significant portions of an RFP establishing foundational information about every vendor, procurement teams can concentrate more attention on the issues that actually depend on the organization conducting the procurement.

PfRs Change the Information Architecture

Proposal for Requestors (PfRs) are designed around that different starting point. Rather than waiting for every requestor to initiate a new information-collection exercise, qualified vendors create structured, standardized proposals that can be maintained and updated within a secure marketplace accessible to qualified requestors.

The proposals are reusable, and that characteristic is more important than it may initially appear.

Under the traditional model, information frequently moves like this:

Requestor asks → Vendor responds → Requestor organizes → Requestor compares

PfRs introduce a different sequence:

Vendor maintains → Requestor discovers → Requestor compares → Requestor asks what remains necessary

The distinction is not simply technological. It changes where work occurs.

A PBM can maintain foundational information about its offering rather than repeatedly recreating it. A consultant can investigate available vendors before constructing a bid list. A requestor can compare relevant capabilities before determining which questions require deeper investigation.

Formal procurement can then focus on what actually needs to be specific to the opportunity: pricing, organization-specific requirements, contractual provisions, implementation considerations, finalist questions, negotiations, and due diligence. RFPs can remain part of that process. In fact, better information before an RFP can produce a more focused RFP.

The objective is not to remove rigor from procurement. It is to separate rigor from repetition.

EXECUTIVE PERSPECTIVE
Don’t just improve how procurement information travels. Improve how the information exists.

AI Becomes More Valuable After the Process Is Redesigned

None of this is an argument against artificial intelligence. It is an argument for applying AI at the right layer.

AI working against structured procurement information has a different starting point than AI asked to interpret several unrelated documents. Instead of spending much of its effort extracting and normalizing information, technology can concentrate on analysis.w

A requestor might use AI to identify meaningful differences among qualified vendors, investigate which offerings align with particular priorities, summarize tradeoffs, identify gaps requiring follow-up, or navigate supporting information more efficiently. A consultant could use those capabilities while applying market knowledge and client-specific judgment, while a procurement professional could spend less time locating information and more time evaluating it.

That is a fundamentally more interesting application of AI than simply producing an RFP faster. AI should enhance the decision-making layer rather than become an increasingly sophisticated patch over an unnecessarily repetitive information layer.

The sequence matters: structure first, compare second, analyze third, and apply judgment throughout.

General-Purpose AI Is Not a Procurement Strategy

The rapid availability of general-purpose AI tools introduces another consideration. Organizations already have access to powerful systems capable of analyzing documents. Microsoft 365 Copilot, for example, can operate within an enterprise environment and access organizational information according to a user’s existing permissions. Microsoft states that prompts, responses, and data accessed through Microsoft Graph are not used to train its foundation models. Its enterprise protections can also incorporate existing permissions, sensitivity labels, retention policies, auditing, and administrative controls.

Those are meaningful protections, but they do not eliminate the need for procurement governance. Microsoft documents controls governing access, data protection, retention, auditing, and the use of agents and connected services.

The lesson is not that organizations should avoid Copilot or other general-purpose AI systems. It is that having access to AI is not the same thing as having an AI procurement strategy.

PBM procurement can involve confidential pricing, contracts, proposal information, financial analysis, and other sensitive organizational or vendor data. Organizations should make deliberate decisions about which tools are approved, who has access, what information can be processed, how interactions are retained, and how AI fits within existing governance policies.

A purpose-built procurement environment has another advantage as well: it can organize AI around the procurement process rather than asking procurement professionals to organize the process around a general-purpose AI tool. That distinction will become increasingly important as AI becomes commonplace.

The Future Is More Time for Judgment, Not More Time for Process

The strongest procurement processes of the future are unlikely to be those that collect the most information. They will be the ones that transform information into insight with the least amount of unnecessary administrative effort.

That means examining the entire procurement workflow rather than simply automating individual tasks within it. Consultants should be able to spend more time advising clients. Procurement professionals should be able to spend more time evaluating vendors. Subject matter experts should be able to spend more time examining the questions where their expertise actually matters. Executive leadership should receive clearer information about meaningful differences and tradeoffs.

Vendors should similarly be able to spend more time explaining what genuinely differentiates their offerings instead of repeatedly reconstructing information they have already provided elsewhere. Technology can help accomplish all of those things, but only if procurement is willing to use technology for more than acceleration.

Six years ago, eliminating boxes of printed RFP responses represented meaningful progress. Today, AI gives organizations the ability to process those same kinds of responses at extraordinary speed. The next evolution should go further.

The future of PBM procurement will not be defined by how quickly organizations can complete an RFP. It will be defined by how much of the work surrounding that RFP still needs to exist.

---

Rapid Request is developing Proposal for Requestors (PfRs) to give qualified requestors and consultants access to structured, reusable vendor proposals before repetitive information collection begins. The goal is not simply a faster RFP. It is a better starting point for procurement.