The case for a results-priced consulting firm for the AI era — and the decisions I need the four of us to make together.
I want to put down, in one place, what I think we're actually building — so that when we say yes to this, we're all saying yes to the same thing. The deck is the 10-minute version. This is the version you can sit with. I've tried to make the argument honestly, including the parts that scare me, because the worst outcome isn't that the idea is wrong — it's that we discover we each meant something different six months in.
Here is the whole thing in one breath: AI has collapsed the cost of building and testing toward zero, which quietly broke the economics that traditional consulting was built on. That break is an opening. We can bring genuinely world-class strategy — and the execution to go with it — to Thai businesses that could never afford MBB, and we can charge for results instead of hours because AI lets a tiny team actually do the work. Everything below is the why, the what, and the how of that sentence.
Traditional consulting priced the way it did for a good reason: execution used to be expensive. When building anything real — a product, a campaign, a market test — cost a fortune and took months, it made sense to spend heavily on planning first. Get it right on paper before you bet the money. So the deliverable was a deck, the unit of value was the billable hour, and the leverage was an army of junior analysts. The client was paying for one thing above all: certainty before commitment.
That logic held for fifty years. It doesn't hold anymore.
A landing page, a campaign, a financial model, a market scan, a working prototype — these now take hours, not months, and cost hundreds of baht, not hundreds of thousands. When testing something in the real market is cheaper than perfecting it on a spreadsheet, the optimal strategy flips. It goes from "plan, then build" to "build, test, iterate." The expensive upfront certainty the old model sold is now the expensive path. A live test is simply a better product than a slide that predicts the test.
I don't want to oversell the precision here — the −97% and the 20× are directional, and I've flagged them as low-confidence in our numbers ledger. But you don't need the decimals. The shape of the curve is the whole argument: for the first time, a tiny team can out-execute an incumbent.
This is the part I keep coming back to. AI doesn't make consulting a little more efficient. It breaks the three things the business was built on, all at once:
| The old model | Our model | |
|---|---|---|
| Deliverable | Decks & paper recommendations | Working systems & live results |
| Pricing | Billable hours & retainers | Outcome / revenue-share |
| Leverage | Armies of junior analysts | Small senior team + AI + a compounding library |
Here's why this matters more than it looks: an incumbent cannot copy this without dismantling their own economics. Their cost base, their org chart, and their partner compensation all depend on billable hours and leverage ratios. The moment they price on outcomes and replace analysts with AI, they break their own P&L. That's the kind of opening that doesn't come often — a change that is trivial for us to build around and structurally painful for the people with the brand and the balance sheet.
We are not chasing the Fortune 500; they already have McKinsey. Our wedge is Thai small-to-medium businesses and 2nd/3rd-generation family firms in transition. The incoming generation is often Western-educated, wants to modernize, and can feel the landscape shifting under them. They are exactly the people who need world-class help — and exactly the people who can't get it. They can't justify MBB fees (half a year's revenue for a deck), and the local "consultants" available to them are too often just sales trainers in a nicer suit.
Revenue-share is what makes us fundable for them. The pitch is simple and it disarms the price objection completely: "We ringfence a clear win, we take 5–10% of what we actually generate — you pay because you earned." The risk moves onto us. That's precisely why they say yes, and it's why this only works if we're disciplined about which wins we underwrite (more on that below).
I want to zoom out further than the deck does, because I think we're early to something bigger than "cheaper consulting for SMEs." Step back across the whole arc: humanity moved from agriculture → industrialisation, and we're still living in the industrial chapter — one whose defining feature is that most people must work for someone else to survive. That arrangement was rational for a simple reason: when an employer hires you, in most cases you produce more value than you're paid. The surplus is the employer's profit, and it's why jobs exist at all.
AI quietly breaks that equation. As AI absorbs more of the marginal work, the surplus an average employee generates over their cost stops justifying the wage — and rational firms substitute. I don't say this to be dramatic; I say it because the early signal is already in the data. Over half of Thailand's workforce is already self-employed — the "work for yourself" mode isn't a fringe, it's already the larger half, and the AI transition will push the line further.
Two things follow, and they're the heart of why I want us moving now:
One — a wave of new owners. A large share of today's employees won't stay employees. Some will choose it; many will be displaced as firms pick AI over headcount. Either way, the flow runs one direction: from job-holder to business-builder. Two — the death of one-size-fits-all. When a bespoke system costs hundreds of baht and an afternoon instead of a six-figure software contract, owners stop buying generic SaaS and start commissioning (or building) tools shaped to their exact business. The services and software landscape gets re-tailored, per-business, almost overnight.
That is our window. This transition generation — the freshly-minted and soon-to-be owners — is exactly who we exist to help: give them the thinking to know what to build, and the tools to build it.
Here's a first, deliberately rough sizing — bottom-up, assumptions on the table, all illustrative until we harden them. The point isn't the decimals; it's the order of magnitude and the fact that the TAM grows as the shift above plays out.
If we're right that millions become owners, the honest second-order question is: what happens when everyone is building a business at once? My read: supply of businesses surges, a lot of them look alike, competition compresses margins, and failure rates climb. That sounds like bad news — but it's actually the deepest version of our demand. In a glut, the scarce thing isn't the ability to start; it's knowing what to build, how to differentiate, and how to actually win. That is precisely what we sell: judgment plus execution plus skin in the game.
So our position in the big picture is deliberate. We don't want to be the firm that helps anyone start anything — that just adds to the glut. We want to be the firm that helps owners pick the right battles and build defensible businesses, and — because we price on outcomes — we are structurally incentivised to back the ones that can actually win. That alignment is our compass when the noise gets loud.
I want to be precise about the category, because it changes everything downstream. We are not a consultancy that uses AI tools. We are a firm that delivers advice and execution on the same invoice, and gets paid on the result. Three things, together:
The Brain — MBB-grade strategy and judgment, the thinking the Fortune 500 pays millions for. The Hands — AI-leveraged execution; we don't hand over a deck, we build the systems and run the test. Skin in the game — pricing tied to the results we create, not the hours we bill. Take any one away and we're just another firm in a crowded category. Together, they're a business no incumbent is structured to be.
On the surface we'll appear to do four different things. Underneath, they feed each other. This is the part I most want us aligned on, because it's tempting to treat the lines as a menu rather than a machine.
10X Claude (฿490), courses, workflow templates. Mass, plug-and-play, near-zero marginal cost.
Engagements whose single job is to drive the client's revenue. Outcome / revenue-share priced.
Chatbots, internal workflows, automations, the data plumbing. Build fee + retainer.
Businesses we build and own; spin off when mature. PromptRent is the first.
The flow: Products give us mass reach, brand, cash float, and — quietly — a filtered list of serious owners (people who buy reveal themselves). Among those buyers we find high-potential businesses for Sales Advisory, where the deep revenue-share wins and reference stories live. Winners then need Systems to scale, which makes them sticky — and here's the non-obvious bit:
Line 3 (systems) builds the dashboards and data pipes that let Line 2 (sales advisory) prove attribution — which is the one thing that makes revenue-share collectible. Owning the measurement layer is our unfair advantage in pricing on outcomes. Everyone can claim they drove the result; we can prove it, because we built the system that counts it.
And every engagement, across all four lines, deposits reusable AI-executable skills into a shared library — so the next project starts further ahead. That library is the asset (Section 5).
Our default is outcome-based pricing wherever a clean, attributable revenue line exists. But I want to be blunt about the risk: revenue-share means we invest our time as capital before we collect. Bad client selection isn't a small mistake — it's months of unpaid work. So we never go pure-contingency without runway (base fee + success fee, always), and we underwrite every rev-share deal like a VC. A client qualifies only if all five of these hold:
| The deal screen — all five must hold | |
|---|---|
| 1 · Ringfence-able win | a specific channel / product / funnel we can isolate and attribute |
| 2 · Clean baseline | we can credibly measure the "before" |
| 3 · Visibility | we see — ideally operate — the numbers; no trust-me reporting |
| 4 · Headroom | the business can actually grow if we do our job (market + margin exist) |
| 5 · Aligned operator | the owner will execute what we hand over |
Fail any one, and it's a fixed-fee project, not revenue-share. I'd like this screen to be a rule we actually hold each other to, not a slide. It is our risk management.
I don't want us buying growth. I want a motion that compounds on its own work. Here's the loop I have in mind, and it's the operational version of the flywheel: products (10X Claude and what comes after) give us cheap, wide reach and surface the serious owners; some of those become advisory clients, where we earn margin and, just as importantly, learn the real problems Thai SMEs face at close range; we then distil what we learn back into products — better, deeper tools — which pull the next, larger wave of leads. Each turn makes the next cheaper and sharper.
Think 10X Claude, but a generation further. Not a prompt pack — a downloadable, deploy-it-yourself system built for an owner with low technical ability: e.g. a complete CRM-and-sales workflow they can stand up in a day, pre-wired with the thinking we'd otherwise sell in an engagement. The product does two jobs at once. It closes the affordability gap — an SME that could never pay for a consulting project can buy the productised version of it — and it acts as a "thought buddy" that helps an owner go from blank page to a working prototype with us in their corner.
That's the bridge that makes the whole firm cohere: advisory is where we discover what's worth knowing; product is how we make that knowledge affordable to everyone else. The high-touch work funds and informs the low-touch work; the low-touch work feeds the funnel for the high-touch work. Neither cannibalises the other — they're two speeds of the same engine.
Paid leads stop the moment you stop paying. A product-led loop does the opposite — every engagement makes the product better, every product release widens the funnel, and the cost-per-qualified-lead falls over time. We're building an owned channel and a compounding asset, not renting attention.
The whole model rests on one operational claim: that a small, senior team — with AI doing the analyst work — can deliver what used to take a floor of consultants. That leverage isn't a nice-to-have; it is the business model. People do the judgment, the relationships, and the last-mile quality bar. AI does the research, the modelling, the first-draft decks and content. If we can't make that real for ourselves, we have no business selling it to anyone.
Between us we already have the right raw mix — strategy, build, and operator instincts. I'd rather we stay deliberately flexible on who owns what for now, and let the roles settle as the work shows us where each of us is strongest, instead of forcing ourselves into boxes on day one. The one thing I won't be flexible on is the thing money can't buy: a shared picture of where this goes. Misaligned founders is the number one killer at this stage — that's the entire reason this memo exists.
If someone asks "what stops a smarter, faster team from doing this to you?" — this is the answer. Every engagement deposits reusable, AI-executable skills into a firm library. Each new project starts further ahead, at lower marginal cost than the last. It compounds. (10X Claude's skill system is the working prototype of exactly this.)
Stacked on top of the library are three more defences: the measurement infrastructure that makes our outcome-pricing provable, Thai-context trust (local nuance plus a brand SMEs believe), and founder pedigree (2× BCG + a real builder + operator credibility). The library is the one that compounds; the others are why we win the first ten clients while it's still small.
Here's the trap I'm most worried about, and it's a quiet one. Four lines × four part-time-ish founders = the classic way to spread thin and do none of it well. The discipline I'm proposing for the next two quarters is deliberately boring:
Products run semi-autonomously (cash + brand + lead-gen) — systematized so they don't eat founder time. Sales Advisory gets the full focus: we go deep on one flagship (RK) and make it the undeniable reference case, because the rev-share model is unproven until exactly one win is documented end-to-end. Systems we take only opportunistically, when it strengthens a Line-2 client. Ventures are parked. We revisit when Lines 1–2 throw off the time and cash to justify it.
One documented revenue-share win (RK), and a product engine that runs on a few hours of founder time a week. Those two things — and only those two — unlock partner conviction, fundraising-of-attention, and the next ten clients. Everything else is a distraction until they're done.
You asked the right question, and I think it's the most important one in this memo: when a new consulting firm starts, why do the overwhelming majority drift into being a no-name shop "optimising Facebook ads" that ends up selling courses — while a tiny few become BCG or Bain? I've thought about this hard. It isn't luck, and it isn't a single trick. It's that the great firms built a self-reinforcing system on purpose, and the rest built a service business by accident. I want us to understand the system on Day 1, because almost every early decision either feeds it or starves it.
Strip away the mystique and I see five things, and they reinforce each other:
1 · They sell trust to the top, not tactics to the middle. MBB sells risk reduction on the decisions that matter most, to the people accountable for them. When the cost of being wrong is enormous, the buyer is not price-sensitive — and prestige is part of the product. The Facebook-ad shop sells a commodity task to a junior buyer who churns on price. Same industry, opposite businesses.
2 · They build proprietary IP that compounds. The experience curve, the growth-share matrix, benchmark databases — owned frameworks and data that make each engagement faster and more credible than the last. The no-name firm re-does every project from scratch and owns nothing afterward.
3 · They are ferociously selective — and selectivity creates the brand. Saying no, holding a brutal quality bar, and protecting "one firm" consistency is why the brand means something. Scarcity and standards are not the by-product of being elite; they are the cause of it.
4 · They price on value, never on cost. They simply don't compete on being cheaper. Once you compete on price you've conceded you're a commodity — and commodities race to zero, which is exactly the road that ends in selling courses.
5 · They run a talent-and-alumni flywheel. Elite output builds the brand, the brand attracts elite people, and alumni become the executives who hire the firm a decade later. The relationship outlives the project.
We can't copy MBB's cost structure, and we shouldn't want to. But the system translates almost perfectly to what we're building — and AI lets us reach it without their overhead:
Our version of "selling trust, not tactics" is revenue-share: the purest possible promise that we own the outcome with you. Our proprietary IP is the playbook/skill library plus Thai-SME benchmark data nobody else is accumulating. Our selectivity is the five-point deal screen — it's not bureaucracy, it's brand-building. Our "never compete on price" is productising the cheap tier so we widen access through products while keeping advisory premium and outcome-aligned — that's how we avoid the course-mill trap while still being affordable. And our flywheel is the GTM loop in Section 4.
Deliberately loose on tactics, clear on the game. Three horizons:
| Horizon | The goal of the phase | What we must not do |
|---|---|---|
| Now → ~6 mo Earn the right | One undeniable, documented outcome (RK). Set our quality bar and start hoarding IP from every engagement. Define the standard of "our work." | Don't chase logos or revenue breadth. Don't take a deal that fails the screen just to eat. |
| ~6–18 mo Turn the loop on | 5–10 documented wins; product loop running; own a narrow category ("the firm that does X for Thai SMEs"). Reference stories sell the next ten. | Don't broaden the category before we own a narrow one. Don't let products become a discount on our trust. |
| 18 mo + Compound & expand | Benchmark-data moat; talent flywheel begins; expand lines/segments/geography from a position of brand, not need. | Don't dilute the bar to scale headcount. Growth follows the brand, never the reverse. |
If we hold the direction — trust over tactics, IP that compounds, selectivity as brand, value over price, a loop that feeds itself — the specific tactics can flex as the market teaches us. That's the balance I want: obvious about the game, flexible about the moves. And it has to be shared across the four of us from Day 1, because the first time we're tempted by an easy, off-strategy deal, this is the section that has to win the argument.
Here's the north star, and then what I'm actually asking of each of you. The north star is simple enough to say in one line and big enough to spend years on: world-class help, finally within reach of the businesses that drive Thailand — and, eventually, beyond it. From ฿490 products, to revenue-share advisory, to helping Thai SMEs expand abroad.
The ask is this: read this through, come with where you agree, where you don't, and what you'd change — and let's spend our next session aligning, not admiring the problem. I don't need us to agree on everything today. I need us to agree on the same thing, deliberately, before we pour the next six months into it. If we get the alignment right, I think the window in front of us is genuinely rare. If we get it wrong, no amount of AI leverage will save us from ourselves.
Let's go build it — together, and on purpose.
— Weerapat
Confidential · Founding Vision · 2026. Figures and their confidence levels are tracked in the Control Room (_control_room/index.html); illustrative numbers are marked as such and should be hardened before any external use. Companion deck: Firm_Vision_Deck.pptx.