# The Referral Flywheel Is Moving Online. Most AEC Firms Haven't Noticed. **By Dan Cumberland** · Published May 11, 2026 · Categories: AI Strategy > Somewhere in the last quarter, someone recommended your firm. The prospect looked you up— your website, your Google reviews, maybe they asked an AI assistant... ## A Referral Was Made Last Quarter\. You Never Heard About It\. Somewhere in the last quarter, someone recommended your firm\. The prospect looked you up— your website, your Google reviews, maybe they asked an AI assistant about you— and decided, before anyone picked up a phone, that you weren't worth the call\. You never heard about it\. It never showed up as a lost deal, because it was never a deal\. That is the new shape of word of mouth in architecture, engineering, and construction\. A referral is no longer the end of the buyer's process\. It's the start of yours, and the first step happens on a screen you don't control\. The lost referral never enters your CRM as a lost deal, because the prospect simply never called\. No debrief, no go/no\-go meeting, no pattern for the partner group to spot\. The firm goes on believing the relationship engine is running fine\. Before going further, say the obvious thing out loud: referrals still run AEC\. They're just not running the way the partner group thinks\. ## Referrals Aren't Dead— They Changed Shape Referrals are not dead for architecture, engineering, or construction firms\. They still influence the majority of new business— behind up to roughly 65% of new opportunities by industry estimates[7](/blog/blog-moving-architecture#ref-7), and the most common way buyers find a new professional\-services firm is still asking another person, which happens about 71% of the time[3](/blog/blog-moving-architecture#ref-3)\. The relationship engine works\. What changed is what happens after the introduction is made\. AEC leans on face\-to\-face business development more than any other professional\-services segment\. In\-person networking is the single most\-used technique even among the fastest\-growing firms[1](/blog/blog-moving-architecture#ref-1)\. That's why the belief is durable: the thing that built the firm is the thing leaders keep crediting\. What's still true about referrals in AEC: - A person recommending you is still the dominant way new clients first hear your name\. - The seller\-doer— the architect or engineer who also brings in work— still wins real projects\. - Trust transfers through people\. Nobody's arguing that's over\. Here's the connection the partner group hasn't made: "we get our work from relationships" now sits right next to "a stranger is grading our website before they'll talk to us\." Most of this data comes from one research house— Hinge Research Institute, which runs the annual High Growth Study, AEC Edition— so treat these as Hinge's findings, not settled consensus\. The shift is real, and it lands hardest on the median firm\. This is a leadership\-level read on [the realities founder\-led firms are working through right now](/for-founders), not a memo for the marketing department\. So if referrals still work, why are referred leads behaving differently? Because a referral now hands the prospect a homework assignment— and most of the homework happens before you're in the room\. ## What Actually Happens Now When Someone Refers You When someone refers your firm today, the referred buyer does not call you\. They look you up first— your website, a Google search, your reviews, your LinkedIn, increasingly an AI assistant— and they narrow the field before making contact\. In Hinge's research, about 51\.9% of buyers say they have ruled out a referred firm before ever speaking with it[2](/blog/blog-moving-architecture#ref-2)\. The referral got you considered\. Your online presence decides whether you make the shortlist\. Call this the online vetting step— the digital checkout\. It's the moment between "I heard about this firm" and "I'll reach out," where a firm gets included or quietly dropped\. Hinge's referral research draws on 523 professional\-services firms across technology, accounting and finance, management consulting, AEC, and legal[2](/blog/blog-moving-architecture#ref-2), so this is buyer behavior across professional services applied to AEC, not an AEC\-buyer\-specific study\. > **The digital checkout, defined:** the online research a referred buyer does— website, search, reviews, LinkedIn, AI assistants— between hearing your name and deciding whether to contact you\. The research behavior is consistent\. When buyers check out a professional\-services provider, about 81% visit the firm's website, about 63% Google it, about 62% ask friends or colleagues, and about 56% talk to references the seller provides; buyers use roughly 3\.2 research methods on average, and three of the five most common are online[3](/blog/blog-moving-architecture#ref-3)\. ```html-table
How a referred buyer checks you outShare who use itWhat it tests about your firm
Visit your website~81%Can I tell what you specialize in and how you'd help me?
Google your firm~63%What does the wider web say— news, reviews, third-party mentions?
Ask friends or colleagues~62%Does your reputation hold up past the one person who referred you?
Talk to references~56%Do past clients confirm the story?
``` Firms get ruled out for the same handful of reasons\. The site is a portfolio with no "here's how we help a client like you" story\. Reviews are thin or absent\. There's no sign of relevant specialization, nothing that answers "would this firm understand *my* project?" The chain is short and worth saying plainly: referral → online vetting → shortlist or silent drop → contact\. If that feels like the rules changed overnight, here's the part that should sting: they didn't\. The flywheel has been running through a digital layer for years\. Most firms just funded the offline half\. ## The Flywheel Was Already Half\-Digital This is less of a shock than an overdue bill: most referrals were never about who you golfed with\. In Hinge's AEC research, about 81\.5% of buyers have received a referral from someone who was never a client of the firm, and roughly 94\.5% of referrals are based on a person's knowledge of the firm's reputation, expertise, or visibility— only about 5\.5% come from someone the firm has actually met[1](/blog/blog-moving-architecture#ref-1)\. The flywheel already runs on reputation\. Reputation is increasingly built, checked, and confirmed online\. > **The referral split \(Hinge 2024 AEC High Growth Study\):** \- ~81\.5% of buyers got their referral from a non\-client \- ~94\.5% of referrals are reputation\-, expertise\-, or visibility\-based \- ~5\.5% come from someone who actually worked with the firm Sit with that 94\.5%\. If almost all referrals already ride on reputation, expertise, and visibility, then your website, your content, your search presence, and your reviews were *already* part of the referral machine\. AEC firms have been paying for the offline half— the events, the lunches, the conference booths, the seller\-doer programs— while starving the online half that now decides who survives the vetting step\. The study behind those numbers covered 129 AEC firms representing about $23 billion in revenue, in its ninth annual edition[1](/blog/blog-moving-architecture#ref-1)\. This isn't a thin sample\. That reframes one line item in particular\. Thought\-leadership content isn't a lead\-gen channel bolted onto the side of business development\. It's a referral input— it builds the reputation that makes a non\-client recommend you, and it's the evidence a referred buyer reads when deciding whether you make the cut\. Which raises the obvious question: if the data has said this for years, why hasn't it landed? ## Why Most AEC Firms Haven't Noticed Two things keep this shift invisible\. First, the losses don't show up anywhere— a referred prospect who rules you out online never becomes a "lost deal" in your CRM; there's no line item, no debrief, no pattern\. Second, AEC runs on face\-to\-face business development more than any other professional\-services segment[1](/blog/blog-moving-architecture#ref-1), so leaders pattern\-match "we win on relationships" and stop looking\. You can't read the label from inside the bottle\. There's no measured figure for how many referred opportunities a weak digital presence costs a firm— no study quantifies that\. But it follows directly from the numbers we do have\. If about half of buyers rule out a referred firm before contact, and the ruling\-out happens on the firm's website and search results, then a firm with a flat online presence is losing referred work it will never see counted\. That's reasoning, not a statistic— and it's worth taking seriously precisely because it leaves no trace\. Three things firms believe that the data no longer supports: - "A referral is a sure thing\." — About 52% of buyers rule a referred firm out before speaking with it[2](/blog/blog-moving-architecture#ref-2)\. - "Our website doesn't matter; our work speaks for itself\." — The buyer never reaches the work if the site fails the vetting step\. - "AI search is a publisher's problem, not ours\." — It's now part of how a referred buyer vets a firm\. The headline claim is precise: it's the median firm and the partner mindset, not the marketing department\. Plenty of high\-growth AEC firms have noticed and acted\. And you can't manage what you can't see— the same way you can't [actually measure what AI is returning](/blog/measuring-ai-success) without instrumenting it first\. The vetting step isn't standing still, either\. The place where referred buyers do their homework is changing again— this time toward AI\. ## AI Search Is the Next Turn of the Flywheel The online vetting step is moving toward AI\. Referred buyers increasingly ask ChatGPT, Perplexity, or Google's AI Overviews to surface and summarize firms before they ever click a website— and those tools can put you on the shortlist or leave you off it entirely\. A July 2025 Pew Research Center analysis found that when a Google search shows an AI Overview, click\-throughs to the underlying results drop to about 8%, versus roughly 15% with no AI summary[6](/blog/blog-moving-architecture#ref-6)\. The research is compressing\. Being visible and citable matters more, not less\. > **Answer engines, in plain terms:** tools like Google AI Overviews, ChatGPT, and Perplexity that compress research and hand the buyer a synthesized answer— including which firms to consider— instead of a list of links\. Be honest about the limits\. Most AI\-search data today is about consumer and publisher traffic, not AEC buyers choosing firms\. This is the leading edge of where buyer research is going, supported by the fact that a large and growing share of B2B buyers now use AI tools somewhere in their research\. There's a real counterpoint too: AI search doesn't remove the need to contact a provider— a buyer still picks up the phone— so local\-service work is somewhat insulated\. What AI search changes is *who makes the shortlist* before that call happens\. A firm with thin content can be filtered out without a person ever evaluating it\. The practical move holds regardless of how fast this lands\. The same things that make you checkout\-ready for a human— clear positioning, real expertise on the page, named specializations— are what let an AI assistant surface and describe you accurately\. This is also [where AI can augment marketing without replacing the people doing it](/blog/ai-marketing-automation): the technology makes a firm's real experts more findable; it doesn't manufacture expertise the firm doesn't have\. None of this means the answer is to become a marketing company\. It means adding one layer to a machine that already works\. ## What "Checkout\-Ready" Looks Like— Without Becoming a Marketing Company The fix is additive, not a trade\. Keep doing what AEC firms do best— in\-person networking is still the single most\-used business\-development technique even among the fastest\-growing firms[1](/blog/blog-moving-architecture#ref-1)— and add the layer the referred buyer now checks\. Being "checkout\-ready" means a referred buyer can tell, in two minutes on your site, what you specialize in, who you've done it for, and why you'd understand their project— before they ever email you\. The checklist, kept short on purpose: 1. **Clear positioning** — a visitor can tell what you're known for, not just what services you list\. 2. **A website that explains how you help this kind of client** — not just a gallery of finished projects\. 3. **Visible proof** — reviews and ratings, named expertise, results relevant to the buyer's project type\. 4. **Steady thought\-leadership content**, published by the firm's actual experts\. 5. **Findability** — you show up in search and can be accurately surfaced and summarized by AI assistants\. This isn't aspirational\. High\-growth AEC firms already do it, and the gap shows\. ```html-table
High-growth AEC firmsAverage-growth AEC firms
Leads generated online~63%~12%
SEO maturity68%+ rate it mid or highLargely undeveloped, ad hoc
Marketing investment~10% of revenue~5% of revenue
``` High\-growth AEC firms generate roughly 63% of their leads online, versus about 12% for average\-growth firms[1](/blog/blog-moving-architecture#ref-1)\. Over 68% of high\-growth firms rate their SEO maturity at a mid or high level, while no\-growth firms' SEO is "largely undeveloped, unstructured, and ad hoc"[4](/blog/blog-moving-architecture#ref-4)\. And high\-growth firms invest meaningfully more in marketing— roughly twice the share of revenue, around 10% versus 5%— with stated priorities that include social media marketing, developing thought leaders, and marketing technology and AI/automation[5](/blog/blog-moving-architecture#ref-5)\. The digital layer isn't optional anymore\. It's the dividing line\. One honest caveat\. At the very top of the market— large public and institutional work won on QBS or RFP and past performance— a partner's relationships and the firm's project list still dominate, and Googling the firm matters less\. Even there, the "have I heard of this firm, do they look credible" check happens online\. And for the $20–100M firm chasing mid\-market private work, the digital vetting step is decisive\. Producing thought\-leadership content at the cadence the new flywheel requires is where AI\-assisted content systems earn their keep\. Used well, they let a firm's real experts publish— turning the principal's actual thinking into articles and answers faster than a busy seller\-doer can manage by hand— without hiring an agency or becoming "a marketing company\." Used badly, AI just pumps generic slop into the internet pipes, and a referred buyer can smell it\. The point is the experts' real thinking, faster— not synthetic filler\. If you're weighing that investment, [a clear framework for when an AI investment is worth it](/blog/ai-decision-framework-founders) beats a vendor demo as a starting point\. \([Dan Cumberland Labs](https://dancumberlandlabs.com) works with founder\-led firms on exactly this kind of build\.\) If you want the short version— the questions a principal actually asks when this lands— here they are\. ## FAQ— Referrals, Online Vetting, and AEC Firms The questions principals ask most when the referral pipeline starts behaving differently: ### Are referrals still important for architecture and engineering firms? Yes\. Referrals still influence the majority of new business in AEC— behind up to roughly 65% of new opportunities by industry estimates[7](/blog/blog-moving-architecture#ref-7)\. What changed is that referred buyers now research firms online before making contact, so a referral alone no longer guarantees a conversation\. ### Why do referred leads sometimes never call? Because buyers vet referred firms online first\. About 51\.9% of buyers report ruling out a referred firm before ever speaking with it[2](/blog/blog-moving-architecture#ref-2)— often because the firm's website or content didn't make clear how it could help them\. The opportunity never registers as a lost deal, because it was never a deal\. ### Do clients Google an architecture or engineering firm before hiring it? Yes\. When buyers check out a professional\-services provider, visiting the firm's website and Googling the firm are among the most common steps— and three of the five most common research methods are online[3](/blog/blog-moving-architecture#ref-3)\. A weak online presence can disqualify a firm before any conversation happens\. ### What share of referrals come from people who actually worked with the firm? A small minority\. About 81\.5% of buyers have received a referral from someone who wasn't a client, and only around 5\.5% of referrals come from someone the referring person personally worked with— roughly 94\.5% are based on the firm's reputation, expertise, or visibility[1](/blog/blog-moving-architecture#ref-1)\. ### Should AEC firms stop in\-person networking and switch to digital marketing? No\. In\-person networking remains the single most\-used business\-development technique even among the fastest\-growing AEC firms[1](/blog/blog-moving-architecture#ref-1)\. The shift is additive: keep relationship\-based BD and add a strong digital presence so referred leads actually convert\. ### How is AI search changing how clients pick AEC firms? Buyers increasingly use AI tools and AI\-generated search summaries to research and shortlist firms, which can surface— or omit— a firm before a human ever evaluates it\. A July 2025 Pew analysis found AI Overviews sharply reduce click\-throughs to underlying sites[6](/blog/blog-moving-architecture#ref-6), meaning the answer is delivered before anyone visits a page\. Visibility and credible, citable content matter more as a result\. One last thing, for the partner who's still not sure this is worth a line in the budget\. ## The Bill Was Always Coming Due The referral engine isn't dying\. It's relocating— half of it now lives on a screen the firm doesn't control, and the firms that maintain that half are the ones surviving the vetting step\. Keep the relationships\. Build the digital layer the referred buyer now checks\. That's the whole move\. If mapping that layer— positioning, content cadence, where AI fits without becoming a marketing shop— is more than the partner group wants to figure out alone, an AI strategy partner can [map the right approach to a firm's actual workflows](/services/ai-strategy) and hand you a plan you own\. ## References 1. Hinge Research Institute, "5 Takeaways from the 2024 AEC High Growth Study" \(2024\) — [https://hingemarketing\.com/blog/story/5\-takeaways\-from\-the\-2024\-aec\-high\-growth\-study](https://hingemarketing.com/blog/story/5-takeaways-from-the-2024-aec-high-growth-study) 2. Hinge Research Institute, "Rethinking Referral Marketing: A New Research\-Based Approach to Referrals" \(n\.d\., based on a 523\-firm dataset\) — [https://hingemarketing\.com/blog/story/rethinking\-referral\-marketing\-a\-new\-research\-based\-approach\-to\-referrals](https://hingemarketing.com/blog/story/rethinking-referral-marketing-a-new-research-based-approach-to-referrals) 3. Hinge Research Institute, "New Study Highlights: How Buyers Buy Professional Services" \(n\.d\.\) — [https://hingemarketing\.com/library/article/new\_study\_highlights\_how\_buyers\_buy\_professional\_services](https://hingemarketing.com/library/article/new_study_highlights_how_buyers_buy_professional_services) 4. Hinge Research Institute, "The Digital Divide in AEC Lead Generation Strategies" \(2023\) — [https://hingemarketing\.com/blog/story/the\-digital\-divide\-in\-aec\-lead\-generation\-strategies](https://hingemarketing.com/blog/story/the-digital-divide-in-aec-lead-generation-strategies) 5. Hinge Research Institute, "5 Key Findings from the High Growth Study, AEC Edition" \(2025\) — [https://hingemarketing\.com/blog/story/5\-key\-findings\-from\-the\-2025\-high\-growth\-study\-aec\-edition](https://hingemarketing.com/blog/story/5-key-findings-from-the-2025-high-growth-study-aec-edition) 6. Pew Research Center, "Google users are less likely to click on links when an AI summary appears in the results" \(July 22, 2025\) — [https://www\.pewresearch\.org/short\-reads/2025/07/22/google\-users\-are\-less\-likely\-to\-click\-on\-links\-when\-an\-ai\-summary\-appears\-in\-the\-results/](https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/) 7. SMPS Foundation & Stambaugh Ness, "AEC\.BD: Building Business Development Success in a Post\-Pandemic World" \(2024\) — [https://www\.smps\.org/wp\-content/uploads/2024/08/AEC\.BD\-Final\-Report\-1\.pdf](https://www.smps.org/wp-content/uploads/2024/08/AEC.BD-Final-Report-1.pdf) ; figure as summarized at Stambaugh Ness, "AEC Firms Win More Work With A Seller\-Doer Referral Strategy" — [https://www\.stambaughness\.com/blog/aec\-seller\-doer\-referrals\-strategy/](https://www.stambaughness.com/blog/aec-seller-doer-referrals-strategy/) --- Source: https://dancumberlandlabs.com/blog/moving-architecture/