A Request for Information is a formal written question about contract documents, drawings, specs, or site conditions that need clarification before work can proceed correctly. The core rule for every RFI: keep it to a single subject, cite exact drawing and spec references, and state a requested reply date. Remember that a response, on its own, never authorizes extra cost or schedule time. That last part trips up more contractors than any other piece of the process.
TL;DR:
- RFIs should be kept to one subject and include precise references, attachments, and a recommended interpretation to ensure that they are answered in one round.
- Escalation occurs if the reply date passes without response, with change order discussions triggered if the answer affects cost or schedule, and roles remain fixed throughout the process.
- A well-maintained RFI log tracks status, response times, and potential impacts, with strict definitions for status updates like answered and closed, to prevent delays and missed issues.
- Response times typically range from two to four days for simple RFIs and up to a week or more for complex, multi-discipline requests, depending on project size and staffing.
- Effective RFI management depends on daily screening, weekly review of aging RFIs, and verification of implementation, as software alone cannot prevent closure discipline failures.
Table of Contents
- What Is the Construction RFI Process, Step by Step?
- What Should Every RFI Include?
- How Do You Build an RFI Log That Actually Works?
- What Response Times Should You Expect?
- When Should You Actually File an RFI?
- How Is an RFI Different From a Change Order or Submittal?
- How Do You Keep the RFI Process From Getting Abused?
- Which Tools Actually Reduce RFI Aging?
- Who’s Behind This Guide?
- Ready to Bring Order to Your Next Project’s Paperwork?
- An Editorial Take on Where RFI Management Actually Breaks Down
- Sources
- FAQ
What Is the Construction RFI Process, Step by Step?
An RFI moves through six distinct stages, and the fastest projects treat every stage as a checkpoint, not a formality. Skipping steps is exactly how RFIs turn into schedule delays and, eventually, claims.
- Origination and field screening. A superintendent, subcontractor, or field engineer spots a conflict, an unclear detail, or a condition that doesn’t match the drawings. Before drafting anything, the general contractor’s team screens the question against the drawing set, spec sections, and the existing RFI log to confirm the issue is real and hasn’t already been asked.
- Drafting the RFI. The originator writes a single-subject request, references the exact sheet numbers and spec sections involved, attaches photos or sketches when relevant, and proposes a requested reply date tied to the schedule.
- Submission and routing. The RFI enters the project’s tracking system (or log) and routes through the project engineer to the architect of record, then out to structural, MEP, or civil consultants if the question crosses disciplines.
- Review and response. The architect or engineer answers with a written interpretation. If the answer implies added cost or schedule time, that gets flagged immediately rather than assumed.
- Distribution and drawing-set mapping. The response goes back to the field team, gets logged against the affected drawing or spec section, and any downstream trades affected by the answer get notified directly.
- Closure confirmation. The GC, not the architect, confirms the answer was actually implemented in the field before marking the RFI closed. Field guide practitioner data from Datagrid backs this distinction between “answered” and “closed” as one of the more overlooked controls in RFI management, because an unverified closure can leave a defect uncorrected for weeks.
Escalation triggers matter as much as the steps themselves. If a reply date passes without response, the project engineer escalates to the architect’s principal-in-charge, not just a resend. If the answer suggests a change to cost or schedule, the PM opens a change order discussion in parallel, never waiting for the RFI thread to resolve it. Roles stay fixed throughout: subcontractors originate, the GC screens and tracks, the architect and consultants answer, and the owner’s representative gets copied on anything with cost exposure. Clear subcontractor scope definitions upstream, as outlined in our subcontractor selection guide, prevent a good chunk of RFIs before they’re ever written, since ambiguous scope is one of the most common root causes of an RFI flood.
What Should Every RFI Include?
A well-written RFI gets answered in one round. A poorly written one bounces back with a request for clarification, and that round trip alone can add three to five days to your timeline. The Procore RFI guide and Datagrid’s field guide converge on the same core fields, and for good reason: skip any of them and reviewers have to guess.
Every RFI needs:
- Project identification — project name, number, and the RFI’s sequential number within the log.
- Sender and recipient — who’s asking, who’s expected to answer, and who gets copied.
- Issue date and requested reply date — the reply date should track your actual schedule need, not an arbitrary buffer.
- A single, clearly stated subject — one question per RFI, never a bundled list of unrelated issues.
- Exact references — sheet numbers, detail callouts, and spec section numbers, not “the drawings” or “the specs.”
- Supporting attachments — photos of the actual condition, marked-up sketches, or a redlined detail when the conflict is visual.
- A sender’s recommendation — your proposed interpretation or solution, which speeds up the architect’s review considerably.
- Impact flags — a checkbox or note indicating whether the question, if answered a certain way, could affect cost or schedule.
Pro Tip: Write your recommendation as if you’re the one who has to defend it in a claims dispute two years from now. “We recommend Detail 3/A501 governs over the conflicting dimension on A201” reads as informed judgment. “Please advise” reads as a stalled field and invites a slower response.
The AIA G716 instructions specifically note that the requested reply date should align with actual contract timelines, not wishful thinking. Padding every RFI with an aggressive 24-hour deadline trains reviewers to ignore your dates entirely.
How Do You Build an RFI Log That Actually Works?
The AIA G716–2004 form provides the industry-standard fields, but most practitioners add a few columns that the base form doesn’t cover. Priority level, an assigned reviewer name, and a linked change order ID (when applicable) turn a static form into a working management tool.
Your log needs to track more than the RFI content itself. It needs to show where every request stands and who’s accountable for moving it forward.
| Log field | Purpose |
|---|---|
| RFI number and subject | Unique identifier, tied to a single issue |
| Origination date | Starts the clock for turnaround tracking |
| Status | Open, pending, answered, closed, or void/cancelled |
| Response required by | The requested reply date from the RFI itself |
| Response date | Actual date the answer was received |
| Cost/schedule flag | Yes/no marker for potential impact |
| Linked documents | Drawing sheet, spec section, related change order |
Status definitions need to be strict, not loose. “Answered” means the architect replied. “Closed” means the GC confirmed the answer got built or implemented correctly in the field, a distinction Datagrid’s practitioner guidance treats as a genuine control point rather than paperwork. Only the GC’s project engineer should have authority to move an RFI to “closed.” Anyone can request “void” for a duplicate, but that status also needs GC sign-off to prevent someone burying a legitimate question.
Maintenance can’t wait for a weekly meeting. Daily screening of new entries against the existing log catches duplicate questions before they consume a reviewer’s time twice. A mandatory field requiring submitters to note what they searched in the log and drawing index before filing a new RFI builds deduplication into the audit trail itself, rather than relying on someone remembering to check.
What Response Times Should You Expect?
Response benchmarks vary by project complexity, but the practical threshold most experienced project managers work toward is one week for a standard, single-discipline RFI. Anything routinely running longer than that starts eroding schedule float fast.
Benchmark snapshot: Research using Aconex project data shows that larger, more complex projects generate more RFIs per day and take longer to turn them around than smaller jobs, meaning your benchmark should scale with project size and discipline count rather than apply a single fixed number across every job.
Project size and staffing directly shape turnaround. A single-architect project with a compact consultant team can often answer straightforward RFIs in two to four days. A multi-discipline commercial job with structural, MEP, and civil consultants layered in routinely pushes past a week once an answer requires input from more than one party, a pattern Procore’s RFI guide reflects in its own guidance on setting realistic response windows in the contract.
Four metrics tell you whether your RFI process is healthy or drifting toward trouble:
- Frequency per day — a sudden spike often signals a documentation gap somewhere in the drawing set.
- Aging distribution — how many RFIs sit in each age bracket (0 to 3 days, 4 to 7, 8 to 14, 15-plus).
- Percent overdue — RFIs past their requested reply date as a share of the open total.
- Mean days to close — not just to answer, but to full field-verified closure.
A log where 20% or more of open RFIs sit past their requested reply date usually points to a design-review capacity problem, not a field problem. That’s a staffing conversation with the architect, not another reminder email.
When Should You Actually File an RFI?
Not every field question deserves a formal RFI, and filing one for the wrong reason clogs the log and slows down the questions that genuinely matter.
- File an RFI when: drawings conflict with each other, a spec section is missing information needed to proceed, a concealed condition doesn’t match what’s shown, or a code requirement appears to contradict the design.
- Don’t file an RFI when: the question is about means and methods (how you build it, not what to build), it’s a supplier-only clarification that doesn’t touch the contract documents, or it’s a trivial field question that has no bearing on design intent.
- Before submitting, check: search the log for a duplicate, confirm the drawing set doesn’t already answer it in a section you missed, and verify the question actually needs an architect or engineer, not just a phone call to your own foreman.
That last checklist step alone prevents a meaningful share of the trivial RFIs that pad logs on busy projects.
How Is an RFI Different From a Change Order or Submittal?
An RFI asks a question. A change order authorizes a change in cost or time. Confusing the two is where projects get into trouble, and it happens more often than most PMs expect.
- RFI vs. change order: the AIA G716 instructions are explicit that neither the RFI nor its response authorizes changes to project cost or time. A change order or construction change directive is required separately, even if the RFI answer clearly implies added work.
- RFI vs. submittal: a submittal seeks approval of a specific product, shop drawing, or material for use. An RFI seeks clarification of what the documents actually require.
- RFI vs. RFP/RFQ: a request for proposal or quote solicits pricing or a bid response from a vendor or subcontractor. An RFI has nothing to do with pricing; it’s a documentation and interpretation tool.
When an RFI response implies extra scope, document it immediately: note the RFI number on a pending change order log entry, notify the owner’s representative in writing, and don’t proceed with the implied extra work until that change order is signed. Treating an RFI answer as informal permission to proceed is the single most common way contractors lose a legitimate cost claim later.
How Do You Keep the RFI Process From Getting Abused?
Contract language does more heavy lifting here than any software ever will. A well-drafted RFI clause should require single-subject requests, give the GC the right to return an RFI without review if it’s duplicative or incomplete, and define a specific response window tied to the project schedule, a recommendation echoed throughout Navigant’s Construction Forum research on RFI-driven delay claims. Building those expectations into the contract documents, similar to how commercial construction contracts define scope and payment terms upfront, removes a lot of ambiguity before ground ever breaks.
Operational discipline matters just as much as the contract language itself.
- Daily screening catches field questions before they become formal RFIs that didn’t need to exist.
- Weekly aging reports surface bottlenecks in design-review capacity while there’s still time to fix them.
- A fixed escalation ladder (say, day 3 reminder, day 7 escalation to principal, day 10 owner notification) keeps pressure consistent instead of reactive.
- Upstream technical validation on the owner or architect side blocks invalid or duplicate RFIs before they consume review time.
Pro Tip: When a response comes in late, log the actual delay in days against the schedule immediately, not months later when you’re building a claim. Contemporaneous documentation is worth far more in a dispute than a reconstructed timeline.
The Navigant report’s central argument is worth internalizing: RFIs are routine and necessary, but left uncontrolled, they become one of the more common bases for delay claims on a project. Contract language and daily habits are what keep that from happening.

Which Tools Actually Reduce RFI Aging?
Most mid-size and large projects now run RFIs through a project management information system with a dedicated module, rather than email threads and spreadsheets. Those modules route requests automatically, flag overdue items, and maintain the searchable log that duplicate-checking depends on.
- Validation agents cross-check an incoming RFI against existing project files and prior entries, and Datagrid’s field data shows this step alone eliminates a meaningful share of duplicate and trivial submissions before they ever reach a reviewer.
- Smaller projects without a full PMIS can still run a disciplined process with a shared spreadsheet, as long as fields are enforced (no blank reply dates, mandatory reference numbers) and naming conventions stay consistent across every attachment.
- Electronic data management, according to research on RFI processing time, can meaningfully cut the hours spent per RFI compared to manual tracking, simply by removing the routing and re-filing overhead.
Who’s Behind This Guide?
This guide draws on field experience managing RFI logs across active renovation and commercial projects, written by Arienne for a general contracting team at Axenia Construction.
Ready to Bring Order to Your Next Project’s Paperwork?
An RFI process only works when someone owns it every single day, not just when a deadline slips. If you’re planning a renovation, tenant buildout, or commercial project in the DC, Maryland, or Virginia area and want a general contractor who treats documentation discipline as part of the job rather than an afterthought, Axenia Construction’s general contracting services are built around exactly that kind of accountability, from the first RFI to final closeout.

An Editorial Take on Where RFI Management Actually Breaks Down
Most advice on RFIs focuses on the form itself, single subject, exact references, requested reply date. That advice is correct, but it’s not where projects actually lose time. The real failure point is closure discipline. Teams answer RFIs promptly, then forget to verify the answer got built correctly in the field, and that gap sits invisible in a log marked “answered” for weeks.
The conventional wisdom oversells software as the fix. A PMIS module routes faster, but it doesn’t stop a superintendent from filing a means-and-methods question dressed up as a design clarification, and it doesn’t force an architect’s office to staff up when RFI volume spikes on a complex job. Those are contract and staffing problems wearing a technology costume.
If you take one thing from this guide, make it the daily and weekly rhythm: screen new questions every day, review aging every week, and never let “answered” substitute for “verified.” That habit costs nothing and prevents more schedule damage than any log template ever will.
— Arienne
Sources
- Impact & Control of RFIs on Construction Projects — Navigant Construction Forum / CMAA
- Request for information frequency and their turnaround time in construction projects (research using Aconex data)
- Construction RFI Workflow: Field guide for GCs | Datagrid
FAQ
What Are the Steps in the RFI Process?
The core steps are origination and field screening, drafting a single-subject request with exact references, submission and routing to the architect or consultants, review and response, distribution back to the field, and GC-confirmed closure once the answer is verified as implemented.
What Format Does a Construction RFI Follow?
Most projects use a format based on AIA Document G716–2004, which includes project identification, sender and recipient, issue and requested reply dates, a single stated issue with drawing/spec references, and space for the response.
How Long Does the RFI Process Take?
Straightforward, single-discipline RFIs often turn around in two to four days, while the practical threshold most PMs plan around is about one week; multi-discipline or complex-project RFIs can run longer based on Aconex project data.
What Is the Difference Between an RFI and an RFP?
An RFI asks a question to clarify contract documents or site conditions, while a request for proposal (RFP) solicits pricing or a bid response from a vendor or subcontractor. They serve entirely different functions and shouldn’t be confused during procurement or construction.
