What Does It Mean to "Identify an AI Opportunity"?
Identifying an AI opportunity means finding a specific workflow in your firm where applying AI would produce a measurable business outcome, such as hours saved, faster turnaround, or fewer errors, rather than simply adopting a tool because it is available. There is an important difference between using AI somewhere in the firm and identifying a genuine opportunity. Plenty of firms have someone using ChatGPT for a memo or a rendering tool for a quick concept image without ever touching the workflows that actually cost the firm time and money every week. The distinction matters because most firms already have some AI activity happening informally, and that activity can create a false sense of progress. A principal who sees a few employees experimenting with AI tools may reasonably assume the firm is already ahead of the curve, when in reality none of that experimentation has touched the workflows that consume the most hours or carry the most risk. Genuine opportunity identification starts by setting the informal activity aside and asking a more specific question: which of our recurring processes, if improved even modestly, would free up enough time or reduce enough error to matter at the firm level.
Start With Workflows, Not Tools
The question that usually comes before a firm looks for an AI consultant is not "which AI tools should we buy," it is "where could AI actually make a difference in our firm." Answering that well requires starting with the firm's existing workflows rather than a list of trendy AI use cases. A tools-first approach tends to produce scattered, low-impact experiments. A workflow-first approach identifies the handful of processes where AI can move a real number. Consider two firms that both invest in AI at the same time. The first buys licenses for a popular design tool and a writing assistant, distributes login credentials, and hopes adoption follows. Six months later, a handful of people use the tools occasionally, but nothing about how the firm operates has changed. The second firm starts by mapping its workflows, identifies that proposal assembly consumes twelve hours a week across the business development team, and then evaluates which tool, if any, actually addresses that bottleneck. The second firm is more likely to see a measurable result, not because it chose better technology, but because it started from the problem instead of the product.
Where to Look Across the Firm
As context for prioritizing internal opportunities, AIA reported in its 2025 research that 8 percent of architecture firms had implemented AI solutions and another 20 percent were working on implementation at the time. See the AIA research summary; it is a dated sector snapshot, not a measure of your firm's readiness.
AI opportunities tend to concentrate in a predictable set of areas. Mapping these areas against your firm's actual workload is the starting point:
- Design research: gathering precedent, code requirements, and site or material information
- Proposals: assembling, formatting, and customizing proposal content for each pursuit
- Project documentation: organizing, updating, and cross-referencing project files
- Specifications: drafting and maintaining spec sections across project types
- Client communication: summarizing meetings, drafting updates, tracking decisions
- Business development: qualifying leads, tracking pursuits, building marketing content
- Administrative work: scheduling, reporting, and internal documentation
- Knowledge management: capturing and retrieving lessons learned and past project knowledge
- Internal coordination: keeping teams aligned across active projects
A useful pattern emerges once these areas are laid out: the strongest opportunities tend to live in work that is necessary but not differentiating. Proposal formatting, meeting notes, and document organization do not distinguish one firm's design quality from another's, which is exactly why they are safe places to apply AI. Design decisions, client relationships, and code interpretation are where a firm's expertise actually lives, and those are the areas where AI should support the work rather than perform it.
| AI OPPORTUNITY CHECK How many hours does your team spend doing this every week? If nobody can answer that question quickly, it is a sign the workflow has not been evaluated as an AI candidate yet. |
What Makes a Workflow a Strong AI Candidate
Not every task is a good fit. The workflows worth prioritizing tend to share several traits: they are repetitive across projects, time-consuming relative to their value, information-heavy, and prone to inconsistency or error when done manually under deadline pressure. A proposal team assembling the same categories of content for every pursuit is a strong candidate. A designer working through a highly client-specific, judgment-driven concept decision generally is not. A useful comparison illustrates the difference. A proposal team that assembles the same categories of content, project experience, team bios, relevant case studies, for every pursuit is doing structured, repetitive work with a predictable format. That is a strong candidate. A designer working through an early concept for a client with unusual site constraints and a specific aesthetic point of view is doing judgment-driven, context-dependent work that varies by project. That is not a strong candidate, even though both activities happen inside the same firm every week.
What Should Not Be Automated
Some activities should stay firmly in human hands: anything requiring professional judgment, architectural expertise, code interpretation with legal consequences, or direct accountability to a client or jurisdiction. AI can accelerate the research and drafting that surrounds these decisions, but the decisions themselves remain the architect's responsibility. This is not a reason to avoid AI in these areas entirely. It is a reason to be precise about where the boundary sits. An architect can use AI to pull together the code sections relevant to a project type faster than manual research would allow, while still applying professional judgment to how those requirements affect the design. The tool changes how fast the research happens. It does not change who is accountable for the interpretation.
| AI OPPORTUNITY CHECK Is this workflow repeated across multiple projects, or is it a one-off? Repeated, information-heavy workflows are where AI opportunity identification tends to pay off fastest. |
Simple Examples Worth Evaluating
- Identifying and pulling forward repetitive proposal sections instead of rebuilding them each time
- Searching internal project knowledge instead of asking around the office for who worked on something similar
- Summarizing meeting notes and decisions instead of manually writing them up after every call
- Organizing scattered documentation into a consistent, findable structure
- Accelerating early-stage research so more time goes into design thinking, not information gathering
What these examples have in common is that none of them ask AI to make a design decision. Each one removes friction from the work that surrounds the decision, freeing up time that would otherwise go into reformatting a proposal, retracing who worked on a similar project, or writing up notes from a call that already happened. That freed time is the actual return on the investment, and it is worth naming specifically rather than treating "AI adoption" as a goal in itself.
A Short Example: Where This Plays Out in Practice
Consider a twelve-person firm that submits roughly three proposals a month. Each proposal takes a business development coordinator six to eight hours to assemble, mostly spent locating the right project photos, updating team bios, and reformatting boilerplate language to match the pursuit. None of that work requires architectural judgment. It requires organization and speed. Applying AI to pull the right content forward and draft a first pass of the narrative sections cuts that time roughly in half, which frees close to ten hours a month, enough to take on an additional pursuit without adding headcount. The design work on each project, the part that actually differentiates the firm, does not change at all. That is what a well-identified AI opportunity looks like in practice: a specific, measurable improvement to a defined workflow, not a firm-wide transformation.
Turning the Framework Into an Internal AI Opportunity Audit
Once you have a list of candidate workflows, the second half of the process is prioritizing them. A simple scoring framework, applied consistently, prevents the firm from chasing whichever opportunity was mentioned most recently in a meeting. The table below outlines the factors worth scoring. None of them require sophisticated tooling to evaluate, just an honest conversation among the people who actually do the work.
Running a One-Week Internal Audit
Firms do not need a formal consulting engagement to complete a first pass at this exercise. A short internal audit, run over a week, is usually enough to produce a workable list: spend a day gathering input from each department on where time actually goes, spend a day mapping that input against the areas listed above, spend a day scoring the resulting list against the factors in the table, and spend the remainder narrowing the list to two or three workflows worth piloting. The output is not a finished plan. It is a short, defensible list that gives the firm somewhere concrete to start, rather than another meeting that ends with general agreement that the firm should probably look into AI.
| Scoring Factor | What to Evaluate |
|---|---|
| Frequency | How often the workflow occurs across projects and teams |
| Time spent | Total hours currently spent on the workflow per month |
| Process consistency | Whether the workflow is done the same way each time, or varies by person |
| Data availability | Whether the information the workflow relies on is accessible and organized |
| Business impact | The effect on revenue, cost, client experience, or risk if improved |
| Implementation complexity | How much change management and integration the fix requires |
| Risk | The consequence of an AI-assisted error in this specific workflow |
| ASK DESIGNHUB Not sure whether a specific workflow scores well on this framework? Bring it to DesignHub. A short conversation is often enough to tell whether it is worth pursuing. |
Why This Is Only the First Step
Identifying opportunities tells you where to look. It does not tell you which opportunities will actually deliver the strongest return, or what should be trained, automated, integrated, or left unchanged. That is the role of a professional AI audit: validating the opportunities the internal exercise surfaces, prioritizing them by realistic ROI, and turning a list of candidates into a sequenced plan the firm can actually execute. Firms that skip this validation step often end up investing in the workflow that generated the most conversation rather than the one that would generate the most value, which is exactly the outcome a scoring framework is meant to prevent.
Frequently Asked Questions
How do I know if my architecture firm has real AI opportunities?
Look for workflows that are repetitive across projects, time-consuming, information-heavy, and prone to inconsistency. If a workflow scores high on frequency, time spent, and business impact, it is worth evaluating further. Most firms find at least three to five such workflows once they look systematically, even if none of them were on anyone's radar as an AI opportunity beforehand.
What architecture firm workflows are best suited for AI?
Proposal assembly, internal knowledge search, meeting summarization, documentation organization, and early-stage research are common starting points because they are repetitive and information-heavy without requiring case-by-case professional judgment. These workflows tend to be strong starting points because the return is measurable within weeks rather than months, which makes it easier to build internal support for further AI investment.
What should not be automated with AI in an architecture firm?
Decisions that require professional judgment, code interpretation with legal consequences, or direct accountability to a client or jurisdiction should stay with licensed staff. AI can support the research and drafting around these decisions without making them. Bringing AI into the surrounding research and drafting work is appropriate. Letting it make the final call on a professional or legal question is not.
How do I prioritize AI opportunities once I have identified them?
Score each candidate workflow against frequency, time spent, process consistency, data availability, business impact, implementation complexity, and risk. The workflows that score well across multiple factors are the ones worth pursuing first. Workflows that score well on several factors at once, rather than excelling on just one, tend to deliver the most reliable results once implemented.
| Bring DesignHub a Workflow If you are unsure whether a specific workflow in your firm is a good AI candidate, that is exactly the conversation to start. DesignHub works with architecture and design firms to identify, validate, and prioritize AI opportunities based on real workflows, not generic use cases. Start a conversation with DesignHub → |