Competitive and Portfolio Research with AI — AI for Architecture and Design

What you'll learn

Use AI research tools to build a structured competitive landscape and precedent study ahead of a pitch or design review
Apply AI to synthesize a firm's own project data into a clear portfolio narrative for a website, award submission, or credentials package
Identify the specific failure mode of AI misattributing a project to the wrong firm or architect, and describe the verification step that prevents a misattribution from reaching client-facing or marketing material
Build a verified competitive research summary for a real pitch using AI as a research and synthesis accelerant

Before a pitch, a design review, or an award submission, you usually need to know what else is out there: what similar projects have been built, how competing firms position themselves, and what precedent language reviewers or clients respond to. AI research tools like Perplexity can pull this together far faster than manual searching, and ChatGPT or Claude are useful for synthesizing your own project data into a clear portfolio narrative. The specific risk in this domain is attribution: AI tools can confidently state that a specific building was designed by the wrong firm, or blend details from two similar projects into a single inaccurate description, and repeating a misattribution in a pitch or marketing document is a credibility problem that is entirely avoidable with one verification step.

Where AI Speeds Up Competitive and Precedent Research

Building a competitive landscape. Given a project type and market, AI can help identify firms working in a similar space, general positioning themes, and precedent projects worth reviewing further — a genuinely useful starting map for research you would otherwise build from scratch.

Precedent study synthesis. Once you have identified real precedent projects, AI can help you organize a structured comparison: program, scale, material approach, and design strategy, making it easier to articulate how your proposal differs from or improves on precedent.

Portfolio narrative drafting. Given your firm's actual project data — scope, square footage, completion date, and outcome — AI can help draft a clear, consistent portfolio narrative for a website, credentials package, or award submission, saving significant time over drafting each project description from scratch.

Tip

When asking AI to identify who designed a specific building or project, treat the answer as a lead to verify, not a fact to cite. Confirm attribution through the firm's own published portfolio, a reputable architecture publication, or the project's own public documentation before including it in any pitch, comparison, or client-facing material.

Precedent Research for an Award Submission — Landscape Architecture Practice

Design Director, landscape architecture practice

Context

A design director was preparing an awards submission for a completed public plaza project and wanted to position the submission narrative against a set of comparable award-winning precedent projects from the past several years, to help the jury understand where the project sat in the current landscape of public space design. Manually researching and organizing that comparative context had historically taken several days of searching award archives and publication sites.

Action

The design director used Perplexity to build an initial list of comparable award-winning public plaza projects from the past five years, including the AI tool's stated attribution of which firm designed each one. Before including any project in the submission narrative, she independently verified the firm attribution for each one against the awarding body's own published records and the firm's own portfolio site, and discovered that the AI tool had misattributed one plaza project to a different firm than the one actually credited in the award archive.

Outcome

The corrected list of five verified precedent projects gave the submission narrative accurate comparative context, and the design director's practice avoided what would have been a visible, embarrassing error had the misattributed project appeared in a formal awards submission read by jurors familiar with the actual body of work. The research phase took about a day and a half including verification, still meaningfully faster than the practice's prior manual process, and the practice adopted a standing rule that every project attribution used in a public-facing submission is checked against a primary source before inclusion.

Knowledge check

A designer asks an AI tool which firm designed a well-known mixed-use development and receives a confident, specific answer naming a firm. The designer includes this attribution in a competitive positioning slide for a client pitch without checking it further. What risk has the designer accepted?

Select one answer.

Competitive research verification

Before

AI states Firm X designed Project Y. This is included directly in a pitch deck comparison slide without further checking.

An unverified AI attribution is treated as fact and used in a client-facing document, creating credibility risk if the attribution is wrong.

After

AI states Firm X designed Project Y. Before inclusion, the attribution is checked against Firm X's own published portfolio and a reputable architecture publication. The attribution is confirmed (or corrected) before appearing in the pitch deck.

The same AI-generated lead is used as a starting point, but only reaches the client-facing document after independent verification against a primary source.

Exercise

~15 min

Your Task

Choose a project type relevant to a current or upcoming pitch. Ask an AI research tool to identify three to five comparable precedent projects and their design firms. For each one, independently verify the firm attribution against the firm's own published portfolio or a reputable publication before including it in any research summary. Note whether any attribution required correction.

Success looks like

Every firm attribution in your final research summary is confirmed against a primary source, not just the AI response
You can explain why a confident, specific AI answer about a well-known project is not sufficient verification on its own

Watch out for

Skipping verification for projects that feel familiar or well-known, since familiarity does not guarantee the AI's attribution is correct
Verifying only the firm name and not double-checking other specific details (completion date, scale, scope) that may also be blended or inaccurate

Your reflection

Did you complete this exercise? What did you find? (Saved locally in your browser)

Quick check

Why does this lesson recommend treating an AI-stated project attribution as 'a lead to verify, not a fact to cite'?

Select one answer.

Key takeaways

AI research tools like Perplexity meaningfully speed up building a competitive landscape and precedent study, and general-purpose tools like ChatGPT or Claude help synthesize your own project data into a clear portfolio narrative.
Treat any AI-stated project attribution as a lead to verify, not a fact to cite — misattributing a project to the wrong firm is a documented AI research failure mode, particularly for similar or adjacent project types.
Verify every attribution used in client-facing or public material against the firm's own published portfolio, the awarding body's records, or a reputable publication before inclusion.
A visible, credibility-damaging correction from a client or reviewer who knows the actual facts is avoidable with one verification step — the cost of skipping it is disproportionate to the time it takes.
Use AI to accelerate the research and synthesis work, and reserve your own professional judgment and verification for anything that will appear in front of a client, jury, or public audience.