Most of a public finance RFP response is firm-level prose that changes slowly: the approach, the credentials, the team, the fee structure. What changes with each solicitation is the issuer, the transaction and the evidence that ties your experience to both. The work is finding the closest winning response and editing the delta.
What in an RFP response is actually the same every time?
Four blocks, and each one describes the firm rather than the solicitation.
The approach is the first. How the group structures a negotiated sale, how it thinks about pricing, how it works with the issuer's financial advisor: those positions do not change because a different authority is asking. The credentials page is the second. Comparable financings, prior issuers, awards and rankings are firm assets, refreshed when something new closes rather than rewritten for each response. The team biographies are the third, and they change when people join or leave. The fee schedule is the fourth, and even that repeats in structure: the same not-to-exceed line, the same reimbursable categories, the same day-one and day-two wording.
That repetition is not laziness. It is what a firm is. The issuer is buying a record and a set of people, and most of what describes them was written before this solicitation arrived. The best practice that governs the process assumes as much: GFOA's guidance on selecting and managing underwriters for negotiated bond sales tells issuers to put the objective evaluation criteria, the fee format and at least three references into the RFP itself. The response then answers on those headings, and the headings are the same on the next one.
What changes from one solicitation to the next?
The issuer, the transaction and the evidence.
The issuer changes the objectives. A city refunding for savings is not the same client as a utility financing new capacity, even when both use the same selection process and the same evaluation headings. Its financing plan, its credit story, the officials who will read the response and the advisor running the search all differ, and the best response speaks to that.
The transaction changes the substance. Size, security, structure, call features, timing, the rating path and the working group: each one changes which experience is relevant and which figures have to be current. The evidence changes with it. The comparable financings you cite, the references you name, the numbers you carry from the model, the conflicts and pay-to-play attestations: all of those are specific to this issuer and this financing, and a prior response's version of them is at best a template.
There is a fifth change that is easy to miss: what the issuer chose to ask that is not standard. A question about a specific credit concern, a request for a particular sensitivity, a local policy the authority wants addressed. The standard blocks cannot answer it, and the response that wins usually does.
Why is the repeated part dangerous to reuse unread?
Because the standard block is standard to prepare, never safe to reuse unread.
A prior response carries another issuer's name, another transaction's figures and another team's roles. The sentences can read perfectly and still be wrong, in the way a familiar label on the wrong document is wrong. A certificate naming the parent entity is not the certificate the issuer asked for. A financing table from a different structure is not this structure's table. The failure is quiet because the prose is polished and the document is real.
The discipline is the same one a checklist needs: templates for the ask, never for the answer. Reuse the headings, the structure and the language that describes the firm. Rebuild the facts. Every number should trace to the run it came from, every reference should be one the issuer can call, and every claim about experience should name the financing that supports it.
That is also why the first draft should be routed, not sent. The banker who wins the mandate is the one who can see which parts of the precedent still apply and which parts changed with the issuer. The draft gives that banker a starting point; it does not remove the judgment.
How should a desk start from its last winning response?
Start from the closest win, not the most recent one.
Closeness has three dimensions. The transaction: the same kind of financing, the same security, a similar size. The issuer: the same kind of authority, a comparable credit and a similar set of objectives. The evaluation: the same criteria and weights, because those decide which sections matter. A response that scored well on a different rubric is a weaker precedent than one that scored well on this one.
Then edit in order. Mark the sections that are firm prose, the sections that are transaction-specific and the sections that are issuer-specific. The first group can move with light editing. The second needs the current figures, checked against the run. The third needs the issuer's own record: its financings, its officials, its plan. The order matters because it puts the cheapest work first and leaves the reading attention for the parts that are actually new.
The workflow on the software side runs that sequence from the other end. RFP automation starts each section from the closest winning response, takes the figures from the DBC run in the house format, and routes the draft to a named reviewer. The retrieval is the part a system does well. The delta is the part a banker does, and on a public finance desk it is the part that decides the mandate.
What should a public finance group measure about reuse?
Measure the edit, not the reuse.
A reuse percentage on its own says very little. A team can start from a strong precedent and still ship a weak response if nobody checked the figures. The useful measures are the ones that describe the delta: how many sections changed, how many numbers were verified against the model, how many review comments a section drew, and how long the response took from the solicitation arriving to the draft going out for review.
Those measures also tell a group when its precedent is getting old. A credential page that has not changed in two years is either a stable firm or a stale one, and the review comments will say which. The point of reusing the standard block is to spend the saved time on the quarter of the response that is new. A group that measures the saving but not the spending has measured the wrong half.
What else is worth asking?
What parts of an RFP response carry over from a prior win?
The firm-level blocks: the approach to a negotiated sale, the credentials page, the team biographies and the shape of the fee schedule. They describe the firm rather than the issuer, which is why they survive from one solicitation to the next with light editing.
Can a firm reuse a losing response?
Yes, and it should read one first. A losing response shows which criteria the issuer weighted, what the firm said that did not land, and where the draft leaned on generic language. Reusing its structure is fine. Reusing its claims unread is not.
How do you know which prior response is closest?
Match on the transaction and the issuer before the calendar. A response to the same kind of financing for the same kind of issuer, with the same evaluation criteria, is a better start than last month's response to a different structure.