Architecture firms are the most relationship-dependent businesses I've ever studied, and they're the least equipped to talk about it out loud.
The industry benchmark that gets cited by the Society for Marketing Professional Services and repeated in every AE marketing conference is that somewhere between 60% and 80% of revenue at a mature architecture or engineering firm comes from repeat clients. Not from new business development. Not from RFP wins. From the same clients coming back for their second, third, and fifth project.
That's a stunning statistic. It means the vast majority of an AE firm's pipeline is a relationship compounding exercise. And yet — go read the average AE firm's website, look at their marketing collateral, sit in on their weekly BD meeting — you'll find them mostly focused on pursuit tracking, RFP response quality, and awards submissions. All important. None of them the actual engine.
The engine is the repeat client machine. And most AE firms are running it on tribal knowledge and calendar reminders instead of on any kind of system.
Why repeat clients dominate AE pipeline
Take any commercial architecture firm working in office, healthcare, life sciences, higher education, or hospitality. The project life cycle looks something like:
- Master plan / initial engagement. 6-12 months. Client is a large institution — a university, a hospital system, a corporate campus, a REIT, a developer.
- Design development. 12-18 months.
- Construction documentation and administration. 12-24 months.
- Occupancy and post-occupancy work.
Total elapsed time on a meaningful project: 3-5 years, sometimes longer. During that time, the architect and the client's real estate lead, facilities director, or head of construction are in weekly meetings. They travel together. They fight through change orders together. They co-write RFIs. By the end of the project, the client team and the architect team know each other better than most business relationships in any other industry.
Now the project ends. And the client has another project starting. And another. Universities are always building. Hospital systems are always expanding. Corporate real estate departments have rolling capital plans. If the last project went well, the client's default choice for the next project is the firm they just worked with.
That's not marketing. That's not BD outreach. That's the natural physics of the relationship. The firm that shows up on the second project doesn't need to sell — they need to not lose the relationship.
The 60-80% repeat client number is a function of that physics. In fact it might understate the effect, because most surveys don't count relationship transitions (when the client contact moves to a new institution and brings the firm with them) as separate from strict repeat business.
The seller-doer model, and its constraint
Architecture and engineering firms mostly run on a "seller-doer" model. The principal or senior associate who sells the project is also the one who designs it, manages it, and delivers it. That's different from law, where partners are more likely to be pure business developers past a certain seniority, and it's very different from consulting, where the pyramid staffing model separates sellers from doers cleanly.
The seller-doer model has one huge advantage: the client relationship stays continuous from the pitch through the delivery. The person who sold the vision is the person who realizes it. That's why AE clients are so loyal — they're not loyal to the firm, they're loyal to the principal.
The disadvantage is that seller-doers can't do cold outbound. They're chargeable. They're already booked at 80% utilization. They can't be running a sales cadence to net-new accounts on top of that. The seller-doer's BD activity has to be almost entirely relationship activation — asking existing clients about upcoming projects, reconnecting with alumni, attending industry events for high signal-to-noise reasons.
This constraint is why the repeat client machine matters so much. It's not that repeat business is preferred. It's that repeat business is the only kind of BD activity a seller-doer principal can realistically sustain.
Where the leaks are
Now for the hard part. If repeat business is the engine, why do so many AE firms feel like they're constantly scrambling for the next project?
Because the repeat client machine leaks. Everywhere. Here are the leaks I see most often:
Client contact turnover. The person who ran the client side of your last project got promoted, or left, or moved to a new institution. If your firm's institutional memory of the relationship lived only in the head of the departing principal, the relationship is now at risk. The new client-side contact has no memory of your firm's work.
Principal turnover. Your seller-doer principal retires or leaves. Their client relationships walk out the door with them. The firm loses the primary node in the graph and often doesn't know it until the next RFP happens and the firm isn't invited.
No systematic tracking of alumni. Every project has an owner's rep or a facilities director who eventually moves on. Some of them become CEOs of construction management firms. Some of them become chief real estate officers at Fortune 500 companies. Some of them start development firms. If your firm doesn't track where these people go, you miss the compounding effect.
No coordination across principals. Two principals in your firm might both have relationships at the same target client — one from a completed hospital project three years ago, one from an ongoing lab project — and neither of them knows about the other's context. Cross-principal coordination is rare in AE firms because principals are protective of their books.
No signal about live opportunity. Even when the firm knows a former client contact has moved to a new institution, nobody surfaces that intelligence at the right moment. The new institution announces an expansion, plans a new building, receives a capital gift — and the firm's principal who could have called the alumnus with a warm intro reads about it in an industry publication six months later.
Each of these leaks costs the firm real revenue. Together they compound into the difference between a firm that grows at 5% a year and one that grows at 15%.
What SMPS actually measures
The Society for Marketing Professional Services is the industry body for BD professionals at AE firms. Their annual benchmarks are the closest thing the industry has to standardized data. The findings across multiple survey years are consistent:
- Repeat client percentage of revenue: median around 65-75% for firms above $10M in revenue
- Hit rate on shortlisted pursuits: median 30-40%
- Marketing spend as percentage of revenue: median around 3-5%
- BD spend as percentage of revenue: median around 5-8%
Two things jump out. First, the industry spends between 8% and 13% of revenue combined on marketing and BD — a meaningful investment. Second, the vast majority of the revenue outcome is coming from the smallest slice of that investment: the informal relationship activation work that senior principals do outside the formal marketing function.
That mismatch is where the opportunity lives. Every dollar the marketing team spends on awards submissions, project photography, and RFP response templates is defensible. But every dollar spent on making the relationship layer legible to principals — showing them who moved where, who's about to have a project, who from their alumni graph is now at a target client — is going to have a much higher return than the equivalent dollar spent on collateral.
The alumni network as the second engine
If the first engine of AE firm pipeline is direct repeat business, the second engine is alumni. And by alumni I mean two things:
Client-side alumni. The owner's reps, facility directors, capital project managers, and university development officers who have worked with your firm on past projects and have since moved on to new institutions. Every one of these people is a live intro node. When they land at a new employer, the shortest path to that new employer's BD conversation runs through them.
Firm alumni. The former principals, associates, and project architects from your own firm who have left to start their own practices, become owner's reps themselves, or move to competitor firms. In a well-managed firm alumni relationship, these are net-positive nodes: they refer projects back, they team with you on pursuits, and they carry the firm's brand credibly into new contexts.
The economics of the alumni engine are the same as the economics of the customer champion motion in B2B software. Boomerang's aggregate data across customer base shows that when a customer champion moves to a new employer, tracking that move produces meaningful net-new pipeline — averaged across customers, it's the biggest single-motion driver of new pipeline from existing relationships. In AE the equivalent motion is client-side alumni tracking, and it's every bit as valuable.
Most AE firms don't run this motion systematically. They rely on principals reading LinkedIn updates and hoping to notice. That's not a system. That's a hope.
What the operationalized version looks like
Let me sketch what running the repeat-client-and-alumni engine as a system actually looks like at a firm. Not as software but as a workflow:
- A live map of every principal's client relationships, including former client contacts, their current employers, and their current roles. This is a relationship graph, rebuilt from calendar meetings, email traffic, and project records — not from partner data entry.
- A live map of every principal's firm alumni network, including the alumni's current employers and roles, and any known intersections with the firm's current target list.
- Signal firing on live opportunity. When an alumnus's current employer announces an expansion, files a permit, receives a capital gift, or hires a new head of real estate, the principal who owns the relationship gets a notification with context.
- A weekly BD cadence where each principal reviews the top handful of live opportunities in their graph, decides which one to activate, and receives support from the BD team on the ask.
- Institutional memory that survives principal turnover. When a principal retires or leaves, the client and alumni relationships they built are handed to a specific successor with a documented context transfer. The relationships stay with the firm.
That last point is where most firms get burned. Principals leave. Client relationships walk out with them. The firm has no formal handoff mechanism. Two years later the firm is losing pursuits it used to win because the relationship layer eroded silently.
What good BD software should do for AE firms
If you're evaluating BD tools for an architecture or engineering firm, the right question is not "does this help us respond to more RFPs." The right question is "does this help the seller-doer principal spend less time on data entry and more time on activation."
Specifically the tool should:
- Read existing signals (email, calendar, past projects) rather than requiring principals to enter data.
- Map the relationship graph across principals so cross-principal coordination is legible.
- Surface job change signals at real-time cadence so alumni moves become BD moments.
- Coordinate the warm intro request workflow so principals ask each other for intros in a consistent format.
- Stay out of the client-facing writing. Principals write their own emails to clients — that's part of the trust — and AI-drafted messages read wrong in this industry.
The tool should not:
- Add another layer of data entry.
- Generate outreach on the principal's behalf.
- Reduce the principal's relationship to a lead score.
- Optimize for outbound volume, which is not the game in AE.
The compounding case
I'll close where I started. If 60-80% of AE firm revenue comes from repeat clients and alumni, and if most firms are running that engine on tribal knowledge, the arbitrage available to a firm that operationalizes the engine is enormous.
A firm running the repeat-client machine as a system doesn't just win the immediate next project. It compounds. Every closed project adds new alumni nodes. Every alumni move opens a new door. Every warm intro closes at 25-30% higher rates than a cold pursuit. Over five years, the difference between a firm that runs this engine and a firm that doesn't is not five points of growth. It's often two or three times the total revenue.
Firms that get this right build durable BD advantages that RFP response quality can't touch. Firms that don't spend a rising percentage of their revenue chasing new business they should have won as repeat business — and burn out their seller-doers doing it.
Frequently asked questions
What percentage of AE firm revenue comes from repeat clients? Industry benchmarks from SMPS and similar sources consistently show 60-80% of revenue at mature architecture and engineering firms comes from repeat client engagements. This includes strict repeat business (same client, next project) and relationship-transition business (former client contact moves to a new employer and brings the firm with them).
What is the seller-doer model? The seller-doer model is the AE industry's default structure where the principal or senior associate who sells a project is also the one who designs and delivers it. The advantage is continuous client relationship from pitch through delivery, which drives loyalty. The constraint is that seller-doers are chargeable — they can't run cold outbound cadences on top of billable work, so BD has to be almost entirely relationship activation.
How should AE firms track client contact turnover? Systematically, not through principal memory. When a client-side contact — an owner's rep, facilities director, capital project manager — moves to a new employer, the firm should know within days, not months. That's where a job-change signal layer creates the biggest value in AE. The alumnus's new employer often becomes the firm's next account.
Why don't AE firms run alumni networks as a BD channel? Most treat alumni as a marketing responsibility — an event, a newsletter, a LinkedIn group. That's not a BD channel; it's brand maintenance. Running alumni as a BD channel means mapping every alumnus to their current employer, intersecting that map with the firm's target account list, and surfacing live opportunities at principals' desks in real time.
What does SMPS actually measure? The Society for Marketing Professional Services publishes annual benchmarks covering marketing/BD spend as percentage of revenue, hit rates on shortlisted pursuits, repeat client percentages, and staffing ratios. The recurring findings: firms above $10M in revenue see 65-75% of revenue from repeat clients, spend 8-13% combined on marketing and BD, and win 30-40% of shortlisted pursuits.
How is this different from law firm BD? The seller-doer model is the biggest structural difference. Law firm partners past a certain seniority can shift toward pure BD roles. AE principals cannot — they're needed on delivery. That constraint makes the AE relationship activation motion tighter and shorter. Every BD hour has to convert. The upside is that the client relationships in AE tend to be even stickier than in law, because the project delivery cycle is longer and more collaborative.
