Trie Is Built Around Business Physics

Business physics is the idea that growth and pursuit work obey structural constraints—not wishful planning. Edward Hess and Jeanne Liedtka argued in The Physics of Business Growth (2012) that organic growth is governed by natural laws: uncertainty is the only certainty, and firms need mindsets, systems, and processes built for exploration—not stability. Adrian Bejan’s constructal law adds that flow systems evolve to reduce resistance over time. Trie extends that tradition at the operational layer: context cannot be invented at send time, tool handoffs dissipate energy, institutional knowledge compounds or decays, and outbound work obeys format gravity. The product is architecture for those constraints—not generic chat.

Key takeaways

1. Hess and Liedtka’s Physics of Business Growth (2012) frames organic growth as governed by uncertainty, experimentation, and internal systems—not linear planning.

2. Bejan’s constructal law: flow systems persist by evolving toward easier movement—a useful lens for why tool friction kills pursuit velocity.

3. Most AI tooling ignores these constraints and optimizes for draft speed—a metric that rarely correlates with pipeline output.

4. Context is conserved: it must be captured before reuse. Cold prompts violate this law and produce generic output.

5. Institutional knowledge follows compound dynamics: small capture investments today reduce marginal cost on every future pursuit.

6. Trie’s architecture—company brain, meeting agent, Connect, local-first desktop—is operational physics, not a feature checklist.

Every pursuit team has a physics problem dressed up as a tooling problem. BD leads ask for better AI. GTM operators ask for fewer tabs. Founders ask why token bills spike when pipeline output still feels generic. Underneath those requests are constraints that behave like laws—and a small body of management literature has been naming them for years.

Trie was not designed as “another LLM interface.” It was designed as infrastructure for how revenue-facing teams actually move work—from conversation to memory to formatted artifact to the next pursuit. That sounds abstract until you connect it to the authors who studied how business actually behaves under force. Once you do, product decisions that looked like taste become inevitability.

What is business physics?

The phrase is most associated with Edward D. Hess and Jeanne M. Liedtka, both professors at the University of Virginia’s Darden School of Business. In The Physics of Business Growth: Mindsets, System, and Processes (Stanford University Press, 2012), they argue that organic growth is governed by its own natural laws—much as physical laws govern motion. Their central claim: the only certainty is uncertainty. Dominating forces are ambiguity and change; the processes at work involve exploration, invention, and experimentation. Those truths run counter to the stability, predictability, and linearity most firms are built for.

Hess and Liedtka distill growth into a formula—Growth = Mindsets + System + Processes. Mindsets prepare organizations to explore rather than merely execute. A growth system aligns culture, structure, leadership behavior, and incentives so experimentation is possible. Processes cover how teams identify opportunities, run low-cost “learning launches,” and manage a portfolio of growth bets. The book’s first chapter is titled “Fighting the Physics of Growth”—a reminder that ignoring these forces is expensive, not heroic.

Flow, resistance, and competitive dynamics

Other authors extend the metaphor in complementary directions. Adrian Bejan, a Duke engineering professor, formulated the constructal law in 1996: for a flow system to persist, it must evolve configurations that provide easier access to what flows. Bejan and co-author J. Peder Zane applied this in Design in Nature (2012) to biology, technology, and social organization—arguing that hierarchies, networks, and channels emerge to reduce resistance. Richard D’Aveni, in Hypercompetition (1994), described markets where competitive advantage is temporary and constantly recreated through strategic maneuvering—a dynamic equilibrium, not a stable fortress.

These frameworks share a rejection of static planning. Hess and Liedtka focus on how firms grow under uncertainty. Bejan focuses on how systems route work efficiently. D’Aveni focuses on how advantage erodes and must be rebuilt. Together they describe a business environment where context, speed, and compounding matter more than perfect forecasts.

Operational physics: where Trie enters

Growth physics explains why organizations need experimentation systems. Operational physics explains why pursuit teams lose anyway—because the daily path from insight to outbound artifact fights conservation, friction, compounding, and format gravity. You can have the right growth mindset and still rebuild every RFP from scratch if context never enters a durable system.

Energy spent re-explaining an account is energy not spent on positioning. Information left in one rep’s Notion page decays when they rotate. A first draft that ignores brand voice creates rework, which creates deadline pressure, which creates errors—a predictable chain reaction. Outbound work falls toward the format the buyer expects—slides, one-pagers, proposal sections in Google Workspace—regardless of where the words were first generated.

These patterns repeat because they are structural, not cultural. You cannot “culture” your way out of context loss any more than you can policy your way out of gravity. Hess and Liedtka built the case at the organizational level. Trie builds the case at the tooling level—where BD pods, GTM teams, and internet pursuit groups actually live.

Law 1: Context is conserved

In physics, energy is conserved—it transforms, but it does not appear from nowhere. Hess and Liedtka’s growth system depends on the same principle at the organizational level: learning launches only compound if knowledge from one experiment enters the next. In pursuit work, context behaves the same way. The positioning paragraph, the objection from last quarter’s loss, the proof point your lead approved on slide nine, the relationship note from a partner call that never made it to the CRM—all of it is potential energy for the next deck, RFP section, or follow-up. If it was never captured, no model can retrieve it. It simply does not exist in the system.

Generic chat tools violate conservation constantly. Each session starts at zero. You paste fragments, upload a PDF, summarize the account, and hope the model holds state long enough to finish. When output drifts, teams blame the model. Often the real failure is upstream: context lived in fifteen places and never entered a durable corpus.

How Trie respects conservation

Trie’s company brain is a capture layer, not a prompt library. It accumulates decisions, preferences, relationship patterns, and proven language from meetings, Slack, and connected documents. Generation draws from that corpus—warm starts instead of cold prompts. The meeting agent feeds the brain automatically so capture is not a heroic end-of-day habit.

Conservation also explains why Trie invests in Connect: relationships are context stored in human networks. Surfacing people and accounts matched to themes in your compounding IP—past decks, call notes, won pursuits—puts latent energy back into motion before someone opens a blank doc or drafts another cold outreach email.

Prompts are disposable. Context is capital. Tools that treat prompts as the unit of work will always feel fast and deliver slow.

Law 2: Friction dissipates momentum

Bejan’s constructal law predicts that flow systems evolve to reduce resistance. Pursuit teams experience the inverse daily: tools that add resistance. Momentum in pursuit work is flow—holding a narrative thread through a discovery call, finishing an RFP section before the deadline window closes, getting a partnership deck out while the conversation is still warm. Friction is anything that forces a stop—switching apps, hunting a case study in Drive, copying paragraphs into slide twelve, restarting a chat because yesterday’s thread is buried.

Industry benchmarks cited in Trie materials suggest roughly forty percent of a knowledge worker’s skilled time goes to non-core desk work: aggregation, rewrites, follow-ups, presentation cleanup. That is not because drafting is impossible. It is because the path from insight to sendable artifact is fragmented. Each handoff is a micro-collision that steals velocity—whether you are responding to an RFP, prepping for a QBR, or drafting an investor update.

Where friction hides

The paste-into-slides tax is the most visible form. Text generation and presentation generation are treated as separate jobs, so strong copy dies in translation—titles rebuilt, proof points re-sourced, layout drift fixed at midnight. Compliance friction is quieter but equally costly: when teams cannot put real briefs into multi-tenant cloud chat, they maintain parallel “safe” and “unsafe” workflows, which doubles administrative load.

Trie reduces friction by collapsing the path. Pursuit artifacts draft and edit in Google Workspace—slides, docs, follow-ups—rather than trapping you in a chat window you must later export. Local-first execution on macOS means sensitive material does not force a detour through a browser tab connected to a shared model provider. Review stays in the artifact, not in an email attachment chain.

Law 3: Knowledge compounds or decays

Hess and Liedtka describe growth as an iterative learning process—a portfolio of small experiments whose lessons accumulate. D’Aveni adds that in hypercompetitive markets, advantage is temporary and must be recreated through know-how and timing. Compounding context is the pursuit-team version of both ideas: yet most teams treat every deal like the first one. They rebuild deck skeletons, re-research the same vertical, re-derive messaging from scratch because institutional memory lived in people who left, chats that expired, or Drive folders nobody indexed.

Decay is the default state. Without a system that captures and reuses, entropy wins: slightly different claims across decks, conflicting proof points, rediscovered objections you already solved twice. At portfolio scale—parallel RFPs, multiple account pursuits, a fundraising roadshow—decay reads as sloppiness even when individual contributors tried their best.

Compounding in practice

When Trie seeds company brain from prior wins, templates, and meeting history, the tenth pursuit starts faster than the first—not because people got smarter, but because the system retained what they learned. A BD team’s next RFP response inherits approved proof points from the last win. A founder’s next investor deck carries objections surfaced on the previous call. Workflow automations extend compounding beyond active pursuits: post-meeting follow-ups, packaged offerings, standing updates for repeat partners. Howdy NYC uses Trie to automate Leapday, a $30k service-based offering, and built custom agents in minutes—operating leverage from accumulated know-how, not more hours.

Compounding is also a continuity play. When employees rotate, org-owned brain keeps frameworks and relationship memory instead of disappearing with personal chat history. For team leads, that is risk reduction. For new hires, it is onboarding that does not begin with archaeology.

Law 4: Format gravity

Format gravity is the pull outbound work exerts toward the channel buyers and partners already recognize. BD teams sell in proposal sections and pitch decks. Startups sell in investor slides and one-pagers. Agencies sell in credentials decks and creative briefs. Internet companies sell in partnership memos and product narratives. No one signs, invests, or agrees to a next meeting based on a chat transcript, no matter how eloquent the paragraphs.

Gravity explains why “AI saved three hours on drafting” so often fails to move pursuit metrics. The words existed in the wrong gravitational field. Someone still had to lift them into Slides or Docs, align visual hierarchy, attach speaker notes, and run the review chain in the format the team actually sends. That lifting work is senior attention spent on conversion, not positioning or relationship work.

Trie treats format as primary, not downstream. Agents produce pitch decks, proposal sections, and follow-ups directly in Google Workspace. Natural-language search pulls relevant case study slides without opening fifteen Drive tabs. The artifact is the production surface—titles tweaked, sections reordered, claims pressure-tested where the buyer will see them.

Law 5: Variance scales with output

Hess and Liedtka warn that firms built for execution struggle with growth because exploration introduces variance by design. Thermodynamics offers a parallel: closed systems tend toward disorder unless energy is applied to maintain structure. Portfolio-level AI output behaves the same way. One rep’s assisted deck looks sharp. Another’s drifts off-brand. A third cites a metric nobody can source. Scaling generation without scaling governance scales embarrassment—especially when multiple pursuits run in parallel.

Trie’s answer is not to remove human judgment—it is to make judgment fast and traceable. Structured review tiers: automated checks for unsourced metrics and off-brand phrasing, a short subject-matter pass by the deal owner, a single send authority for outbound approval. When context and format live in one system, leads review diffs instead of rewriting from fear.

Boundary conditions are not bugs

NDAs, regulatory constraints, and leadership sign-off are boundary conditions—hard edges the system must respect. Architecture that pretends they do not exist gets adopted in shadow workflows or not at all. Local-first storage, org-owned company brain, and explicit review before send are Trie’s way of working within those boundaries instead of asking teams to choose between speed and safety.

How the laws map to Trie’s surfaces

Trie’s product map is easier to read once you see it as physics infrastructure:

Meeting agent → capture

Calls are high-context events. Letting them evaporate into personal notes violates conservation. The meeting agent joins, transcribes, summarizes, and drafts post-call documents that feed company brain—turning conversational energy into reusable corpus before anyone closes the laptop.

Company brain → compound

The brain is where decay loses. Decisions, voice, proof points, and relationship patterns accumulate across pursuits so marginal cost per outbound artifact falls over time instead of resetting every Monday.

Google Workspace deliverables → reduce friction and respect gravity

Output lands where review chains and buyers already live. No export step. No second production pass that steals evenings.

Connect → activate latent context

Relationships and accounts matched to compounding IP convert dormant memory into pursuit action—who to call, what you last discussed, why this opportunity is credible now.

Workflows → leverage

Configured automations multiply output without multiplying hours—during active pursuits, between pursuits, and for productized offerings that run on their own cadence.

Trie Desktop (local-first) → boundary conditions

Sensitive briefs and proprietary methodology stay on your machine by default. Team sync through Trie Cloud when collaboration helps—but the architecture acknowledges why many teams paused on cloud AI in the first place.

Who feels these laws most?

The physics are universal, but some teams hit the walls first:

BD and pursuit teams responding to RFPs at volume—where every response should start warm but usually starts from zero. Startup GTM and founder-led sales—where one person wears product, positioning, and pipeline hats and cannot afford rework. Agencies and internet companies running parallel new-business pursuits—where variance across decks becomes a brand problem overnight. Lean teams between one and twenty-five people—where there is no production department to absorb formatting and research rework.

The pattern that matters: structured outbound work, sensitive material, repeat pursuits, and small enough teams that modest productivity gains justify real software spend immediately.

What happens when you ignore the laws

Teams that ignore business physics usually follow a recognizable arc. Month one: excitement at draft speed. Month two: paste tax and sourcing hunts. Month three: leadership clamp-down after an off-brand send. Month four: “AI is helpful for brainstorming” while the team still rebuilds decks manually. The tooling did not fail. The workflow fought the constraints.

Trie’s outcomes framing matches what teams report when architecture aligns with physics: over seventy percent of the deterministic workflow between first draft and final artifact can be meaningfully automated when context and format are handled natively; teams who once averaged four weeks from first draft to sendable deck compress that last mile; pursuit teams run more parallel responses without adding headcount.

Those numbers are not magic. They are what you get when you stop treating AI as a paragraph generator and start treating institutional knowledge as an asset that obeys compound rules.

Designing with physics, not against it

If you are evaluating AI tooling for a BD, GTM, or pursuit team, ask physics questions instead of feature questions:

Where does context enter the system, and does it stay org-owned? How many handoffs sit between first draft and outbound send? Does output land in the format your review chain uses? What happens to knowledge when someone leaves? Can you trace a claim on slide seven to a captured source without a scavenger hunt? Does the architecture respect your boundary conditions, or force shadow usage?

Tools that score well on those questions tend to feel boring in demos and indispensable in pursuits. Tools that score poorly feel thrilling for ten minutes and expensive for ten months.

A rollout that respects compound dynamics

Physics-friendly adoption is incremental—big-bang “AI transformation” memos rarely compound because they skip capture. Hess and Liedtka’s learning launches offer a useful parallel: start small, test hypotheses, let lessons feed the growth system. Trie rollout follows the same spirit at the tooling layer.

Phase 1: Seed conservation

Connect a bounded corpus: three to five prior wins, active positioning, the Drive folder for one live pursuit. Goal: one warm start that proves context transfers—whether that pursuit is an RFP, a partnership deck, or a fundraising round.

Phase 2: Remove one friction point

Automate the artifact your team hates most—post-call summaries, executive one-pagers, credential sections, or a standard pitch skeleton. Measure time from meeting end to shareable doc, not message count.

Phase 3: Add review discipline

Define send authority and what “sourced” means. Publish a one-page checklist. Review should take minutes because context and format already live together.

By the second pursuit, teams often report they stopped opening generic chat for that workflow—not because they were mandated to, but because ignoring conservation and gravity finally felt slower than respecting them.

Why this framing matters now

AI agents can finally operate computers reliably enough to take on knowledge work—but most pursuit teams do not have engineering resources to build privacy-conscious, format-native tooling themselves. They are already paying the physics bill in tabs, rework, and token spend that duplicates research agents already ran yesterday. Hess and Liedtka showed that growth requires systems, not slogans. Bejan showed that flow finds the path of least resistance. Trie productizes what those insights imply for daily pursuit work.

Capture once, compound across pursuits, draft where buyers live, review with traceability, execute locally where sensitivity demands it. That is not a metaphor borrowed from a pitch deck. It is the operational layer beneath the growth frameworks Hess, Liedtka, Bejan, and D’Aveni spent careers describing—and the difference between AI that generates text and AI that accelerates pipeline.

Related topics worth exploring next: the last mile in AI-generated work, building institutional knowledge that compounds, local-first AI for confidential pursuit work, turning meeting notes into sendable decks and proposals, Hess and Liedtka on growth mindsets and learning launches. Each connects to the same core challenge—turning AI speed into client-ready quality without losing brand, facts, or judgment.

Frequently asked questions

What is business physics?

In management literature, the term is most associated with Edward Hess and Jeanne Liedtka’s The Physics of Business Growth (2012), which argues that organic growth obeys natural laws—chiefly that uncertainty is the only certainty—and requires mindsets, internal systems, and processes for experimentation. Adrian Bejan’s constructal law and Richard D’Aveni’s hypercompetition theory extend the metaphor to flow efficiency and temporary advantage. Trie applies operational versions of these ideas to pursuit work: conserved context, reduced friction, compounding knowledge, format gravity, and governed variance.

How is Trie’s “business physics” related to Hess and Liedtka’s book?

Hess and Liedtka focus on organizational growth capability—how firms build systems for learning launches and portfolio management under uncertainty. Trie focuses on the operational layer beneath that: how BD, GTM, and pursuit teams capture context, move work through tools, and ship formatted output without fighting conservation, friction, and decay every week. The book explains why static planning fails; Trie is infrastructure for the teams doing the exploring.

Who are the main authors in the business physics tradition?

Edward D. Hess and Jeanne M. Liedtka (The Physics of Business Growth, 2012; related works include Hess’s Grow to Greatness and Liedtka’s Designing for Growth). Adrian Bejan and J. Peder Zane (Design in Nature, 2012) on constructal law and flow systems. Richard D’Aveni (Hypercompetition, 1994) on temporary advantage and dynamic competition. Trie draws on all three lineages while building product for pursuit teams, not replacing the underlying management research.

How is business physics different from “best practices”?

Best practices are recommendations teams can ignore. Business physics—whether in Hess and Liedtka’s growth framing or Trie’s operational framing—describes forces you pay for whether or not you name them: rework, context loss, pipeline drag, compliance risk. Architecture that ignores them fails predictably.

Does Trie replace human judgment?

No. Judgment is the work buyers and partners pay attention to—positioning, relationship fit, strategic framing. Trie automates the deterministic layer between insight and formatted artifact: research pulls, reformatting, first passes, follow-through—so humans spend attention on story, scope, and send authority.

Is Trie only for consulting and professional services?

No. The same physics apply to BD teams, startup GTM, agencies, and internet companies running partnership or new-business pursuits. Any team that ships structured outbound work—decks, proposals, one-pagers, follow-ups—benefits from conserved context and format-native output.

Why local-first if context compounds in the cloud?

Both matter. Local-first respects boundary conditions for sensitive material on your machine. Trie Cloud syncs team workspaces when collaboration helps. The split is physics: keep confidential corpus where NDAs require it; compound org memory where the team needs shared access.

What is format gravity?

The tendency for outbound work to pull toward established channels—usually Google Slides and Docs—regardless of where text was first generated. Tools that ignore format gravity export friction to the user.

How long until compounding is visible?

Many teams feel less friction on the first pursuit once company brain is seeded from prior work. Compounding accelerates over weeks as meetings, wins, and workflows feed context back in—similar to how the tenth RFP response starts faster than the first once templates and proof points exist.

Trie is built for teams that ship client-facing deliverables at scale. If you want AI that respects how your team actually runs pursuits, start with conserved context, Workspace-native output, and architecture that compounds instead of resets. If you are tired of re-prompting from zero and pasting into slides at midnight, start with a workflow that keeps company context, review, and presentation output in one place.

Download Trie for Mac at trie.dev, or talk to the founders if you want help mapping your first capture-and-deliver workflow. The laws do not bend—but your tooling can finally stop fighting them.