How the leader track works
Six sessions, no code. The artifact you build across them is a one-page data roadmap for an enabler that a chief executive could fund. Tonight is the business model, because every choice after it follows from who pays for what. Session 2 turns that into the sequencing rule: margin comes from hours removed. Session 3 is Pasar Live, a new live channel nobody knew how to sell on, and how one table got eight teams reading the same numbers. Session 4 is what the marketplaces let you keep, and why the customer is the one thing you cannot. Session 5 is three products built from one data estate. Session 6 is the metric registry, because GMV, take rate and active brands are the numbers everybody quotes and nobody defines. The running example is Serai Commerce, invented for these pages: 48 brands, three marketplaces, four markets, 220 people.
★ What you walk out with tonight. A money map of your own business, or one you advise: every fee line, who pays it, what behaviour it rewards, and the five places your people spend the most hours. Plus one sentence you can say in any planning meeting and defend: "we do not own the goods and we do not own the buyer, so our margin is the work we stop doing."
Four businesses, one order
6 min liveFollow one hundred dollars of lipstick from a buyer's phone to four different companies. The reseller, the marketplace, the brand and the enabler can all say "we sell lipstick online" in a pitch, and they make money in four completely different ways. The difference is what each one owns.
LiveWhat an enabler does, in the order a brand experiences it4 min▶
A brand that wants to sell in Southeast Asia has a practical problem: three or four marketplaces, each with its own seller centre, campaign calendar, content format, livestream tooling and customer service inbox, in four or five languages. Building that team in every market takes years. An enabler rents it out.
In the order the brand meets it: the enabler opens and decorates the stores, loads and localises the catalogue, produces the content (listing images, short videos, cutdowns), runs the livestreams with its own hosts or the brand's, buys and manages the ads, brokers creators and affiliates, answers the buyers in platform chat, and often handles fulfilment. Every one of those is a job done by people, every week, for every brand.
What it does not do is buy the stock (the brand owns the goods) or own the relationship with the buyer (the marketplace does, and session 4 is about how completely). That absence is the most important fact on this page.
Self-studyThe hybrids, and why they do not change the rule3 min▶
Real enablers blur the columns. Some buy stock for a few brands and resell it, which puts a reseller inside the enabler. Some run a creator network that signs hosts and takes a cut of their commissions. Some own a small house brand. Each hybrid is a different business with a different margin sitting inside the same company, and the useful discipline is to keep them apart on the page: a reseller line is paid for owning stock, the service line is paid for hours, and the creator line is paid for brokering.
The rule on this page holds for the service line, which is usually the largest one. When a hybrid line grows big enough to change the company's economics, it deserves its own money map rather than a footnote on this one.
Olist, in its own words. The public Olist dataset that the analyst track uses comes from a Brazilian company that describes itself as connecting "small businesses from all over Brazil to channels without hassle and with a single contract": a merchant signs once, and Olist lists and sells through the big marketplaces for them. That is the enabler shape, one step from the brand side, and it is why its order data is a fair stand-in for Serai's on the bench.
Three fee lines, three payers
6 min liveSerai is paid three ways, and each fee line rewards a different behaviour. The mistake is to read them as one revenue number. They behave differently when volume moves, they are paid by different people, and they point a roadmap in different directions.
LiveThe three lines, and what each one rewards4 min▶
| Fee line | Who pays | How it moves | What it rewards |
|---|---|---|---|
| Service fee (retainer) | The brand, monthly | Fixed per brand; moves with the number of brands | Doing the agreed work at the lowest cost in hours |
| GMV share | The brand, as a share of its sales | Moves with the brand's sales on the stores Serai runs | Growing the brand's GMV, whoever's effort grew it |
| Creator cut (the network arm) | Out of the creator's commission, or a brand's campaign fee | Moves with how many creators sell, and how well | Matching the right creator to the right product |
On Serai's constructed books the first two lines are almost equal: 48 brands at $6,000 a month is $864,000 a quarter, and 2 percent of $45 million of quarterly GMV is $900,000. They look like one revenue number of $1.76 million. They are not. Double the brands and the first doubles while each brand's GMV stays the same. Double one brand's GMV and only the second moves, by $2 on every extra $100.
LiveWhat "MCN" means, and why it is a different business inside the same building3 min▶
A multi-channel network signs creators and hosts, finds them brand deals, and takes a share of what they earn. On a social-commerce marketplace the creator is paid a commission on the orders they drive, and the network's cut comes out of it. Serai's network arm keeps 15 percent of the commissions it brokers (constructed), so on the $8 creator commission in the figure it keeps $1.20.
The network line has its own economics: its cost is recruiting and coaching creators, its asset is the roster, and its data is creator performance. Session 3 and analyst session 6 are where that data earns its keep, because "which creator actually sells" is the question brands ask most and platforms answer least.
Self-studyTake rate, and why an enabler's is not a marketplace's3 min▶
"Take rate" is revenue divided by GMV. For a marketplace it is a clean idea: fees earned over the value of goods sold through it. For an enabler it is ambiguous, because two of its three lines are not a share of GMV at all. Serai's per-order lines are 3.2 percent of an order; add the retainer and the quarterly take rate is ($864,000 + $900,000) / $45,000,000 = 3.9 percent, before the creator line. Put the retainer in or leave it out and the number moves by almost half.
Neither version is wrong. Quoting one without saying which is wrong, and session 6 is about writing it down once. The general mechanics of GMV and its equation are taught in the ecommerce metrics course; this track only adds the enabler's twist on top.
Even GMV is a choice. Sea Limited's annual report defines Shopee's GMV as the value of orders on the marketplace and says its calculation "includes shipping and other charges", and it counts an order "regardless of whether the transaction is settled or if the item is returned". That is a legitimate definition for a marketplace. An enabler paid 2 percent of GMV has to decide whether its fee is on that number, or on settled sales net of returns, and the two can differ by a lot on a campaign day.
It sells hours, and gives away the knowing
5 min liveTurn the money map over and look at the cost side. Serai has no stock, no warehouse it owns, no factory. Its cost is 220 people's weeks. Everything it earns has to cover those hours, and everything it learns while working them goes out of the door inside the fee.
LiveWhy the drain is the lever3 min▶
Look at who controls each pipe. The retainer is negotiated once a year and mostly set by what competitors charge. The GMV share moves with the brand's sales, which depend on the brand's product, price and the marketplace's campaigns as much as on anything Serai does. The creator cut depends on creators. The one pipe Serai controls week to week is the drain.
Forty listing images in five languages is not an invention, it is repetition. A reply to "where is my parcel" is not judgement, it is a lookup. Every such hour that a machine takes over lands straight in the bucket, and it scales: a new brand then needs fewer new hours. That is the argument session 2 turns into a sequencing rule, and it is the reason a data roadmap for an enabler should not start where most data roadmaps start.
LiveThe leak: the knowing, given away inside the fee2 min▶
Every quarterly business review Serai writes for a brand contains knowledge nobody else has: which creator sold in which category, what went wrong in the first week of a new store, which live slot was worth paying for, how the same SKU behaves on three platforms. Today that goes out inside the retainer, as a deck, for free.
Session 5 is about pricing that leak, carefully: not by charging brands for conversations they expect, but by building products on what the estate knows, sold first to buyers who are not competitors.
Self-studyThree questions that tell you which business you are really in3 min▶
- If every brand's GMV fell 20 percent tomorrow, what happens to your revenue? If the answer is "not much this quarter", your retainer line dominates and you are a services business. Your margin is in the drain.
- If you signed ten new brands, how many people would you hire? If the answer is proportional, the drain scales with revenue and nothing you have built is leverage yet.
- What did your last three brand reviews tell the brand that it could not have learned anywhere else? Whatever that is, it is the product session 5 builds.
Your money map, on one page
★ 20 min · everyone buildsNo data needed. Your finance lead would recognise every number you write, which is the point: this map is the shared ground the rest of the track stands on. If you do not run an enabler, do it for Serai using the figures on this page, or for any agency or services firm you know.
List every fee line, and keep hybrids apart. Write each way money arrives on its own row. If one line is really a different business (a stock-holding resale arm, a house brand), give it a row and a note, not a merge. Most people find three or four.
Name who pays each one, and when. Monthly, per sale, per campaign, per creator payout. The timing matters later: a line paid per sale reacts to a bad campaign within a week; a retainer reacts at renewal.
Write what each line rewards in one phrase. This is the column that surprises people. A retainer rewards doing the agreed work in fewer hours. A GMV share rewards the brand's growth whoever caused it. Say out loud which line your incentives actually follow today, because a team paid on one and managed on another is the most common source of friction in the building.
Rank your five biggest hour sinks. People in the team times the hours a week they work on it. Round hard; order of magnitude is enough. For Serai, content (40 people x 30 hours = 1,200) and customer service (24 x 38 = 912) lead, which is typical of any enabler with a content studio.
Write the one sentence, and test it on the three questions. Fill in the brackets and read it to someone from finance. If they disagree with a word, the map is wrong somewhere and that disagreement is worth more than the map.
Before session 2
25 min- Finish the money map and ask your finance lead to correct the split between retainer and GMV-linked revenue for last quarter. Bring the two numbers.
- Replace the hour-sink estimates with the real rota for your two biggest teams. The corrections are usually upward for the team that feels most manual.
- Pick the last brand review you sent. Underline every sentence the brand could not have learned without you. Count them; that count is session 5's raw material.
- Write your enabler's take rate twice: with the retainer inside it and without. Note which one your last board pack used, and whether it said so.
Three questions
Clear all three and Leader session 1 gets stamped in your passport.
1 · What separates an enabler from the other three business models on the same order?
The reseller and the brand earn goods margin, the marketplace earns fees for owning the buyer and the matching. The enabler owns neither, so its revenue is fees for effort and its cost is hours. B is exactly what it does not have, which is session 4.
2 · Serai signs twelve new brands and the existing brands' sales stay flat. Which fee line moves?
The retainer moves with the number of brands. The new brands' GMV share will arrive as they sell, but nothing about the existing brands' sales changed. Reading the lines as one revenue number hides exactly this.
3 · Which pipe in Serai's bucket does Serai control week to week?
Retainers are set at renewal against competitors, GMV depends on the brand's product and the marketplace's campaigns, creator earnings depend on creators. The hours are Serai's, every week, and every hour removed lands as margin. Session 2 builds on this.
What this session teaches, and where it came from
Every fact on this page, and its verification tier, is recorded in the course's source map.
Leader session 1 cheat sheet · pin this
You have the money map and the drain. Session 2 turns it into a sequencing rule: on a service-fee model, the use case that removes the most hours pays first, and the one every data roadmap reaches for first, pricing optimisation, mostly pays the brand. You will see it computed on ten use cases and one assumptions block, including the setting where the rule stops holding.