Pew Research Center reported in 2025 that about a third of American adults had used ChatGPT, and usage among adults under fifty was higher still. That population includes the homeowner planning a backyard renovation, the developer scoping a mixed-use site, the school facilities director with a grant for a green schoolyard, and the city planner writing a request for qualifications. When any of them types “landscape architect” and a city name into an AI assistant, they get a short list of firms with a sentence about each. Your firm is either on that list or it is not, and the list is now being consulted before Google is opened. That is the situation AEO for landscape architects exists to address.
The profession is more exposed than most here for a reason that has nothing to do with the quality of the work. Landscape architecture firms have long marketed through portfolios and referrals, and neither one gives an AI model much to work with. A portfolio is images. A referral is a phone call. The model needs text: answers, mentions, reviews, and structured facts about the firm that it can verify across more than one source. Firms that have never written anything down are invisible to it, no matter how good the built work is.
Why does the AI answer skip good firms?
When an assistant is asked to recommend a landscape architect in a region, it is not reading a ranked list of websites and repeating the top three. It is synthesizing what it can find and verify: firm names that appear in local news, in ASLA award listings, in municipal meeting minutes, in reviews, in directory profiles, and in the firm’s own published material. A firm mentioned in several of those places, with consistent details, is a safe recommendation. A firm with a beautiful website and no footprint anywhere else is a risk the model will not take, because it cannot corroborate anything the site says.
This is why the firm that gets named is often not the firm with the best portfolio. It is the firm with the most legible presence. The good news is that legibility is buildable, and building it is mostly a matter of writing down what the firm already knows.
The Question Ledger
Here is the method. Over the next month, every principal and project manager in the firm keeps a running list of the questions clients and prospects ask, in the client’s words. Not the questions you wish they asked. The ones they ask. “How much does a landscape architect cost for a residential project?” “What is the difference between a landscape architect and a landscape designer?” “Do I need a landscape architect for a commercial site plan in this county?” “What native plants will survive in our soil?” “How does a rain garden work?” “How long does the design process take?” At the end of the month you will have forty to sixty questions. That list is your Question Ledger, and it is the entire content plan for AEO for landscape architects.
The ledger works because AI models are answering these same questions, and they are drawing on whoever has answered them clearly on the open web. Right now, in most regions, the answers are coming from national home-improvement sites and generic directories, because no local firm has written them. The first firm in a region to publish direct, specific answers to its own ledger becomes the source the model cites for those questions, and a firm cited for the question tends to be the firm named for the recommendation.
Answer each question the way the client asked it
Take the top twenty questions from the ledger and publish a page for each. The format matters. The question is the title, in the client’s words. The first paragraph is the direct answer, two to four sentences, with a number or a specific if one exists. Then the explanation, the caveats, and what the firm does about it. A page titled “How much does a landscape architect cost in Central Texas?” that opens with a real fee range, explains what drives the range, and describes how the firm structures fees will be read by clients and cited by models. A page titled “Our Design Philosophy” will be read by nobody.
Be specific to your region. A national site can say “native plants vary by climate.” Your firm can say which twelve species it specifies most often in your county and why. That specificity is what makes the firm the better source, and it is what a national site can never match.
Make the firm’s facts consistent everywhere
The second pillar is duller and just as important. AI models cross-check firm details across sources, and inconsistency reads as unreliability. The firm name, address, phone, founding year, principals’ names, services, and service area should be identical on the website, the Google Business Profile, the ASLA firm directory, LinkedIn, Houzz if you use it, and every other listing. Add structured data to the site (LocalBusiness or ProfessionalService schema with the same details) so the facts are machine-readable. Fix the old listing with the previous office address. This takes an afternoon and removes the most common reason a firm gets omitted.
Get mentioned where the model looks
Third-party mentions are what turn a firm from “claims to exist” into “verified.” For a landscape architecture firm, the highest-value mentions are local news coverage of completed projects, ASLA national and chapter award listings, municipal records that name the firm on a public project, university or extension service pages that cite the firm’s work, and reviews on Google that describe specific projects. Each one is a place the model can look and find the firm’s name attached to real work.
Pursue them on purpose. Pitch every public project opening to the local outlet that covers the problem the project solves. Enter the chapter awards every year. Ask residential clients for a Google review that names the project type and the neighborhood. Offer the principal as a source to reporters covering heat, flooding, and public space. Six months of this produces a citation trail that no competitor with only a portfolio can match.
Which projects should the site actually describe?
Portfolio pages are where AEO for landscape architects most often fails, because firms present projects as image galleries with a one-line caption. Rewrite the ten most relevant project pages as short case narratives: the client’s problem, the site’s constraints, what the firm designed, what it cost or what it captured or what it changed, and where it is. A model can extract “designed a two-acre stormwater park in this city that captures runoff from this many acres” from a narrative. It can extract nothing from a gallery. Include the project type and the city in the page title so that a question like “who designs green schoolyards in Ohio” has a page to land on.
Where firms get this wrong
Three mistakes account for most failed attempts at AEO for landscape architects, and each is easy to avoid once named. The first is writing for peers. A page that explains a stormwater design in the language of the profession, with references to hydrology and planting typologies, impresses other landscape architects and answers no client’s question. The client asked how a rain garden works and whether one would fix the wet corner of their yard. Write for the client, in the client’s words, and let the expertise show through the specificity of the answer rather than the vocabulary.
The second is treating the website as a portfolio with words attached. Firms add a blog, publish three pieces about design philosophy, and stop when nothing happens. The Question Ledger method works because every piece answers a question someone is asking a machine right now. A philosophy piece answers nothing. If a page does not begin with a question a client has asked, it does not belong in the program.
The third is inconsistency across the firm’s own channels. The site says the firm was founded in 2009, LinkedIn says 2011, the Google profile lists a service the firm no longer offers, and the ASLA directory has the old office address. Each discrepancy is small. Together they tell a model that the firm’s facts cannot be trusted, and a model that cannot trust the facts will not name the firm. An hour spent reconciling every listing is worth more than a month of content.
What a fee page looks like when it is done right
Because cost is the most common question in every ledger, it is worth being precise about the page that answers it. The title is the question, with the region: “How much does a landscape architect cost in the Portland metro area?” The first paragraph gives the answer in two or three sentences: a typical fee range for residential design, a range for a full design-and-construction-administration engagement, and a note on how the firm structures fees (hourly, percentage of construction cost, or fixed fee by phase). The second section explains what moves a project from one end of the range to the other: site complexity, permitting, the level of construction documentation, whether the firm is managing bids and construction. The third section says what a client gets at each phase, so the fee is attached to deliverables rather than floating. The last section invites the client to send the site and a budget for a specific estimate.
A page built this way answers the client’s question, gives the model something checkable to cite, and pre-qualifies inquiries so that the firm’s first conversation with a prospect starts from a shared understanding of the budget. Firms that publish it report the same thing: fewer inquiries from people who could never afford the work, and more from people who read the page and decided the fee was reasonable.
What changes when it works
The first sign is a new kind of inquiry: a prospect who says they asked ChatGPT or Perplexity for a landscape architect and the firm came up, often with a summary of what the firm is known for. The second is that the summary is accurate, because the firm wrote the material it was built from. The third is that the inquiries are better qualified, because the prospect arrived already understanding the process, the fee range, and the kind of work the firm does. Answer the questions clients actually ask, keep the firm’s facts consistent, get mentioned where the model looks, and describe projects as stories with numbers. That is the whole discipline, and in most regions no landscape architecture firm has done it yet.