---
title: 'Hard bid vs CMAR vs design-build: how estimating workflows differ'
url: 'https://piper-ai.com/resources/hard-bid-vs-cmar-vs-design-build'
description: 'How GC estimators adjust for hard-bid (design-bid-build), CMAR, and design-build. Scope timing, risk allocation, and bidding process compared side by side.'
contentTypes: [comparison]
publishedAt: '2026-08-06T20:00:00.000Z'
updatedAt: '2026-10-14T08:02:33.683Z'
author: 'Ido Gedanken, CEO'
readingTime: 4
---

# Hard bid vs CMAR vs design-build: how estimating workflows differ

1. Hard-bid prices complete docs once; the lowest responsive bid wins and the GC carries construction cost risk
2. CMAR brings the GC in during design for progressive estimates and a negotiated GMP
3. Design-build prices an integrated design proposal up front; the design-builder holds design and construction risk
4. Align the process to the method: hidden-scope guards in DBB, source-checked assumptions in CMAR, continuous design-cost feedback in DB

**Design-bid-build (hard-bid), CMAR, and design-build change when the estimator enters, what "done" means for design, and who eats design vs cost risk.** Get the method wrong and you apply a sealed-bid checklist to a living GMP, or pad a hard-bid like an early design-build proposal.

This comparison is the side-by-side view. For phase checklists, failure modes, and carry rubrics by method, see the longer [estimating workflows guide](https://piper-ai.com/resources/hard-bid-vs-cmar-vs-design-build-estimating-workflows).

| Feature | Hard-bid (design-bid-build) | CMAR | Design-build |
| --- | --- | --- | --- |
| Contract structure | Separate owner-architect and owner-GC contracts | Separate design and construction contracts; CM advises in design, builds under GMP | Single owner-design-builder contract for design and construction |
| Selection / award | Competitive low bid on complete documents | Qualifications or best-value; GMP negotiated near design completion | Best-value RFQ/RFP; teams propose design and price |
| Estimator timing | After 100% design: one lump-sum bid | Early and ongoing: ROM to DD updates to GMP at ~60-90% | On the proposal team: conceptual price with design, refined as design advances |
| Risk allocation | Owner retains design risk; GC bears construction cost risk | Owner retains much design risk until GMP; GC carries cost up to GMP | Design-builder holds design and construction risk |
| Schedule shape | Sequential; longest typical path | Overlap possible; faster than pure DBB | Heavy overlap; often shortest |

## Hard-bid (design-bid-build)

Owner finishes design, then invites bids. Estimators get **one shot** on the final package:

- Take off quantities and scope from 100% drawings
- Solicit subcontractor quotes on defined packages
- Submit a lump-sum price intended to cover what the documents show

Competition is primarily on price. Missed quantities and silent ambiguities come out of margin.

**Practical notes**

- Run thorough plan checks and RFIs before bid ([scope generation](https://piper-ai.com/resources/construction-scope-generation-guide))
- Confirm every sub priced every [addendum](https://piper-ai.com/resources/construction-addenda-management-for-gc-estimators)
- Track exclusions and unclear items: design risk sits with the owner in theory; change-order reality still needs a paper trail ([exclusions review](https://piper-ai.com/resources/review-subcontractor-exclusions-and-qualifications))
- Use a [coverage matrix](https://piper-ai.com/resources/construction-bid-coverage-guide) so thin trades do not hide behind a low number

DBB is precision on one final number against a frozen set.

## CMAR (construction manager at risk)

The GC is hired early as construction manager (usually on qualifications) then delivers under a GMP.

1. **Early involvement**: cost feedback at schematic and DD; optional early packages for long-leads or sitework
2. **VE and contingencies**: constructability and allowances for unfinished scope while design is still open
3. **GMP proposal**: typically \~60-90% CDs; negotiation may follow
4. **Construction**: CM becomes builder; cost overruns beyond GMP are the CM's problem unless the owner approves changes

**What changes for estimating**

- Multiple estimates that narrow uncertainty, not one sealed bid
- Often open-book justification of line items and assumptions
- Early design errors caught before GMP reduce later fights, underestimating risk still lands on you after lock

CMAR estimation is process control as much as arithmetic: every update should name drawings and pricing sources. That is the same source-backed discipline as [leveling](https://piper-ai.com/resources/construction-bid-leveling-guide) and [final review](https://piper-ai.com/resources/final-bid-review-qa-qc-general-contractors), applied across design stages.

## Design-build

One design-builder holds design and construction under a single contract.

1. Owner shortlists via RFQ, then RFP for technical + price proposal
2. Team prices from performance criteria and evolving design, estimate and design move together
3. After award, construction can start on early packages while later design continues
4. Design-builder owns design errors, omissions, and cost overruns (unless the owner adds scope)

**Implications**

- Cost model and design are integrated; items may be less granular, oriented to performance
- Design changes feed the estimate immediately, undocumented decisions become buried cost
- Conceptual accuracy matters: too low and you panic later; too high and you lose or leave money on the table

DB compresses schedule and coordination when the team stays aligned. It also demands version control and a living baseline, freeze points and allowance cutoffs from the [workflows guide](https://piper-ai.com/resources/hard-bid-vs-cmar-vs-design-build-estimating-workflows).

## How the paths differ

| Step | Hard-bid | CMAR | Design-build |
| --- | --- | --- | --- |
| 1 | Owner finalizes 100% design | Owner hires CM for precon | Owner issues RFQ/RFP to DB teams |
| 2 | GC prepares lump-sum estimate | GC provides cost feedback and VE during design | Teams propose integrated design + price |
| 3 | Award to lowest responsive bidder | GC submits GMP (~60-90% design) | Design-builder awarded single contract |
| 4 | Build to the docs | Build under GMP; owner pays approved changes only | Construction proceeds as design finalizes |

Across all three, the estimator's core job does not change: **complete scope coverage and clear assumptions.** In hard-bid, that means catching every drawing and spec section before the sealed number ([solicitation](https://piper-ai.com/resources/construction-bid-solicitation-guide), [scope gaps](https://piper-ai.com/resources/spot-scope-gaps-before-you-carry)). In CMAR and DB, the same discipline tracks how evolving design and trade inputs move cost and risk, and how the [carry](https://piper-ai.com/resources/carry-number-methodology-construction-bids) is justified.

**Growing share** Owners shifting to CMAR and design-build for speed and single-point accountability. Estimators still need method-specific checklists

## Align the process to the method

| Method | Estimator focus | Common miss |
| --- | --- | --- |
| Hard-bid | One final bid; catch every scope item and addendum | Bidding blind on ambiguities; thin trade coverage |
| CMAR | Iterative estimates + GMP; manage allowances for unfinished work | Locking GMP without a reconciliation log of assumptions |
| Design-build | Concurrent design-cost; living baseline | Vague performance brief; undocumented design drift after proposal |

## Where Piper fits

The method decides when design freezes and who carries which risk. It does not change what has to be true before you stand behind a number: the scope you are asking each trade to carry is defined, the bids you received are compared against that scope, and the assumptions you priced are still visible when the documents move.

Piper is the AI operating system for preconstruction. It reads the drawings, specifications, addenda, quotes and estimates for the project in front of you, together with your company's own standards, and keeps one understanding of that project across scope, leveling and final review, with every finding linked back to the document it came from. In hard-bid that shows up as hidden-scope guards before a sealed number goes out, in CMAR as assumptions you can trace when the GMP is negotiated, and in design-build as a read that survives the design still moving underneath it.

Piper does not choose your delivery method, write the contract, or replace the estimator's judgment on contingency. It keeps the project understanding current so that judgment is made against what the documents actually say.

### When should we use this comparison vs the longer workflows guide?

Use this page for a side-by-side risk and timing snapshot. Use the [estimating workflows guide](/resources/hard-bid-vs-cmar-vs-design-build-estimating-workflows) for checklists, failure modes, KPIs, and carry rubrics by method.

### Does hard-bid mean zero contingency?

No. Complete docs and strong coverage support a leaner carry; incomplete docs or thin trades still need an honest risk allowance.

### Is CMAR always open-book?

Often, but not always. Even when it is not, progressive estimates should still name sources and assumptions. That is what makes the GMP defensible.

### Who carries design errors in design-build?

The design-builder, unless the owner adds scope or the contract carves out exceptions. That is why freeze points and a living estimate matter more than in sealed-bid DBB.

### Do bid spreads mean the same thing on CMAR and DB?

Classic high-low IFB spreads apply mainly to hard-bid. For CMAR/DB, watch dual proposals, conceptual-to-GMP deltas, and trade buyout spreads inside the price, see the [bid spread report](/resources/construction-bid-spread-report).