INSIDE
MARKETPLACES
Everything about Ecom & Qcom that operators normally learn the expensive way — unit economics, dark-store math, negotiation. Six chapters. One certificate.
Contents
{{ t.c1t }}
Ten terms and one thinking style — and you stop nodding in meetings and start running them. Every marketplace decision reduces to a funnel: traffic → intent → conversion → units.
SOURCE Redseer (Jan 2026) for Q-commerce orders and GMV; QuickCommerceMap (Mar 2026) for dark stores. S&D% is my own benchmark.
{{ t.h011 }}
Eleven terms that will make you sound like you've done this for years — and, more usefully, let you follow what is actually happening in the room.
SOURCE Definitions are standard. The benchmark ranges are my own, from live client P&Ls — check them against your own category.
{{ t.h012 }}
S&D is the largest cost bucket after COGS and the one brands underestimate most. Two lenses: where the money goes inside S&D, and how the burden shifts across categories.
Where S&D rupees go on a typical ₹100 GMV order, as a share of GMV excl GST.
Same product cost, very different S&D burden.
Fashion's return rate doubles logistics. Electronics carries low referral fees. Category choice is S&D destiny.
S&D% is a structural cost, not a negotiation. You can squeeze 1–2 points through fulfillment mode and courier rates — but the big levers are picking the right category and cutting returns. A 5% drop in returns can unlock 2–3 points of CM1.
{{ t.h013 }}
"S&D" is a bucket. Inside it are seven distinct fees, each with its own formula and its own lever. Most brands optimise two and leave the rest on the table.
SOURCE Platform rate cards, March 2026. Rates move with category, fulfilment mode and account — read your own.
{{ t.h014 }}
Couriers charge on whichever is higher — actual weight or volumetric weight. Large light packages (pillows, storage boxes, baby products) always ship on volumetric.
Volumetric (kg) = (L × B × H cm) / 5,000
{{ t.p014div }}
Chargeable = max(actual, volumetric)
A pillow, 40 × 30 × 20 cm, actual weight 1 kg. Volumetric = 4.8 kg. You pay for 4.8 kg, not 1.
A book, 20 × 15 × 5 cm, actual weight 0.8 kg. Volumetric = 0.3 kg. Actual wins — you pay 0.8 kg.
Lever: right-sizing packaging can cut volumetric by 30–40%. Every centimetre off L, B or H compounds.
{{ t.h015 }}
You will never be 100% right. You will always be far more useful than someone who says "I don't know". Nobody in an interview wants an exact answer — they want to watch you shrink the universe, one honest ratio at a time.
{{ t.scene1 }}
{{ t.scene2 }}
{{ t.h016 }}
Tap a card to reveal. Active recall beats re-reading — answer in your head first, then flip.
{{ card.def }}
{{ q.explain }}
{{ t.c2t }}
One order, stripped to the bone — before offices, salaries and fixed costs enter the room. If a single order makes money on its own, scale is a choice. If it doesn't, scale is just a faster way to burn cash.
SOURCE My own benchmarks on a mid-ticket SKU. Everything sits on GMV including GST except TaCoS, which sits on GMV excluding it.
{{ t.h021 }}
{{ t.p021 }}
{{ t.c2asli }}
{{ t.c2read }}
SOURCE Worked example on a ₹999 order. Cost ratios are my own benchmarks, not platform-published figures. Every percentage is on GMV including GST except TaCoS.
{{ t.h022 }}
The exact Amazon India flow: GMV → GMV excl GST → returns → net revenue → COGS → S&D (referral + logistics + closing) → CM1 → marketing → CM2. S&D% varies by SKU depending on MFN, FBA or Seller Flex.
SOURCE Same ₹999 order as 02.1. Fee ratios from platform rate cards, March 2026; COGS and returns are my own benchmarks.
CM2 = CM1 − (TaCoS% × GMV excl GST). TaCoS sits on GMV excl GST so it measures ad efficiency against the whole business, not just the ad-attributed slice.
{{ t.p028base }}
{{ t.h023 }}
{{ t.h024 }}
Sessions tell you how many people walk into your store. CVR tells you how many actually buy. Two of the most underrated numbers on any marketplace.
{{ t.p024cvr }}
CVR = 1,000 / 10,000 = 10%
ATC rate = 22%
Return rate = 200 / 1,000 = 20%
You're buying traffic that doesn't convert. Fix images, title, bullets and reviews before spending another rupee on ads.
The product works. Scale ads aggressively and push more sessions into it.
Your listing converts but nobody finds it. Invest in ads, keywords and deals to drive traffic.
SOURCE Formulae from the seller portals. Benchmarks are my own, measured on established listings.
{{ t.h025 }}
Two different ratings, constantly confused. Seller rating protects your account. Product rating drives your sales. You need both above threshold — track weekly.
SOURCE Amazon and Flipkart seller documentation, current to March 2026.
{{ t.h026 }}
SOV is the best single proxy for visibility on any marketplace. If you're not showing up, you're not selling.
SOURCE Definitions from the platform ad consoles. How to read them is my own.
SOV moves with competition, bids and stock. A weekly cadence catches drops early.
Track your top 10–15 revenue keywords individually. 8% at category level can still mean 0% on a high-intent term.
SOV above SOM means you're overspending on visibility that isn't converting. SOV below SOM means organic momentum — protect it.
Amazon Brand Analytics, Helium 10 Market Tracker, DataHawk, PIPPO for Qcom. Most Qcom platforms don't expose SOV natively yet.
{{ t.h027 }}
{{ t.p027 }}
SOURCE My own operating heuristics.
{{ t.h028 }}
The key difference from Ecom: there is no referral/fulfillment split. The platform takes one large commission upfront, and your S&D shrinks to the logistics cost of getting stock to the dark store.
Blinkit has the widest band — top FMCG brands negotiate down to 25%, new or niche brands pay 40%. Fifteen points of commission is your entire CM2. Where you land inside the band is decided by negotiation, not a spreadsheet: velocity, offtake share and exclusivity are your only chips.
SOURCE Worked example on a ₹450 order. Commission bands from live rate cards, March 2026; other ratios are my own.
COGS {{ cogsStr }} · S&D {{ sndStr }} · marketing {{ mktgStr }}
SOURCE GMV is divided by (1 + GST%), never reduced by it. Every cost percentage — returns, COGS, S&D and TaCoS — is on GMV excluding GST, the base platforms charge on; only the CM2% returned is expressed against GMV including GST, to match the % GMV column in 02.1 and 02.2. The defaults reproduce the ₹999 order in 02.2. The chapter tables round every line to the rupee so the column adds up on the page; this works to the paisa, so it can read a rupee higher.
{{ q.explain }}
{{ t.c3t }}
Same product, same price — the fulfillment mode alone swings S&D by 6–8 points and conversion by roughly 18%. MFN, FBA, Seller Flex and IXD are not logistics jargon; they are four different P&Ls for one SKU.
{{ t.h031 }}
{{ t.p031 }}
Instead of 100 units each to FC-Bangalore, FC-Delhi and FC-Mumbai, you send 300 units to IXD-Gurgaon and Amazon's network splits and forwards on its own demand forecast. Best for pan-India demand with limited warehouse manpower.
{{ t.c3asli }}
SOURCE Platform fee schedules, March 2026. The cost ranges are my own.
S&D% is not fixed. It changes per SKU with fulfillment type, weight and dimensions, and shipping zone. Always calculate S&D at SKU level — never as a blanket percentage.
{{ t.h032 }}
Who owns the inventory decides who owns the price — and therefore who owns the margin.
SOURCE Platform seller and vendor agreements. “Best for” is my own judgement.
{{ t.h033 }}
Regular marketplace selling. A normal Seller Central account, open to all brands.
Upgraded seller tier with better SLAs, offered to high-GMV sellers with a track record.
Amazon sends a PO, you ship direct to customer. No FC needed — good for heavy or bulky items.
Amazon's mechanism to increase FC inventory — sometimes with advance payment to accelerate supply.
{{ t.h034 }}
Two very different machines inside the same group.
{{ t.h035 }}
Seller Central is you selling to the customer. Vendor Central is you selling to Amazon — they raise a PO, you invoice them, they own pricing, storage, delivery, returns and customer service. Lower commission, far fewer operational moving parts. The trade: you give up control of price and you live on their payment terms.
1P buys you calm. It also costs you the price lever, the ad-data granularity and the ability to run your own promos. Most healthy brands end up hybrid — hero SKUs on 3P where the margin and control matter, long-tail and bulky SKUs on 1P where operations were eating the margin anyway.
{{ t.h036 }}
Your dashboard shows GMV. Your bank shows something else. The gap lives in one report, and knowing the exact click-path is the difference between a guess and a number you can defend. On Amazon Seller Central: Payments → Reports Repository → Settlement (V2). Every fee, every reimbursement, every adjustment, per order.
Charged on a heavier slab than your carton actually is. Common, recoverable, and nobody checks.
Reimbursements are not automatic in every case. Reconcile inbound vs received.
Customer refunded, unit never came back. Chase it in the same cycle, not next quarter.
Seller Hub → Payments → Settlement reports. Same discipline, different column names.
{{ q.explain }}
{{ t.c4t }}
Qcom hands you impulse demand at premium AOV and takes its share before your P&L even starts. The platform buys your stock and owns the customer, so your levers shrink to three: the commission you negotiate, the availability you maintain, and the velocity you earn.
SOURCE QuickCommerceMap platform scrape, March 2026. Blinkit reported ~2,100 by April. The 95% OSA floor is my own.
{{ t.h041 }}
{{ t.p041 }}
TARGET 10–30 MINUTES · DARK STORE RADIUS ~2 KM
{{ t.h042 }}
Share of dark stores showing 'In Stock'. Measured at SKU × store level, never in aggregate.
TARGET >95%When DS% sits below OSA you have picking errors or SKU mismatches.
TARGET >92%Low fill rate leads to stockout, rank drops and dead organic. A vicious cycle.
TARGET >90%Ads perform 3–5× better in peak windows. Set +30–50% bid multipliers there.
CUT BIDS OFF-PEAK{{ t.h043 }}
Qcom runs on a different metric language. Platforms track inventory at dark-store level — FE (frontend, the stores) and BE (backend, the mother warehouse).
SOURCE Formulae from the platform seller portals. Targets are my own.
Listing % (are you even on the shelf?) → OSA (is the shelf in stock?) → Weighted OSA (in stock where it matters?) → DOI (will it stay in stock?) → PSL (what did getting this wrong cost?). Most brands jump straight to PSL without fixing anything upstream.
Platform-owned central warehouse. Bulk stock before fan-out to dark stores.
~2 km radius neighbourhood hubs. Pick-pack happens here. This is 'the shelf'.
The order. Ten-minute delivery. What every upstream metric is optimising for.
{{ t.h044 }}
Three rails, three different bets. Pick where your category actually moves.
SOURCE Dark stores — QuickCommerceMap, Mar 2026. Market share — Datum Intelligence via Reuters, Jan 2026. AOV — 2026 analyst projections on a GOV/orders basis, not cart value; Zepto has none published on that basis.
{{ t.h045 }}
{{ t.p045 }}
There is no national listing. Stores are grouped into city clusters, and your SKU is either in a cluster's assortment or invisible in it.
Backend is the city warehouse. Frontend is the dark-store shelf. Only frontend stock is sellable — BE inventory that never transfers may as well not exist.
A SKU out of stock across most of a cluster loses visibility, and every rupee of ad spend pointed at it is spent buying traffic to an empty shelf.
Spend resolves at SKU × city × keyword, not nationally. A blended national bid is three good cities subsidising ten bad ones.
SOURCE My own operating checklist.
Nobody loses Blinkit on price. They lose it on availability — stock sitting in the backend while the ads keep running against an empty frontend. Fix the transfer before you touch the bid.
{{ t.h046 }}
Most brands treat data as a reporting tool. The best treat it as a decision engine. Below: PIPPO and the CrepDog Crew stack — live dashboards built to stop brand managers deciding from stale spreadsheets.
Every platform on one screen, each with its own commission, COGS, S&D, CM1 and CM2. "Which platform is actually working" stopped being a week of spreadsheet work.
GMV excl GST → commission → COGS → S&D → CM1 → ads → CM2, with gross and net RoAS per platform and a scale-up or optimise call on each.
D2C, retail stores, marketplace and personal shopper side by side — GMV, orders, AOV, return rate and category mix against the previous period.
GMV to EBITDA with every leak named — returns, GST, COGS, S&D, payments, marketing, fixed costs — plus the GST position and what it costs to move a box.
LIVE DASHBOARDS BUILT AND RUN BY SAHIL · ALL FIGURES REDACTED FOR CONFIDENTIALITY
GMV, CM2, TaCoS and OSA in one view. Decisions at a glance.
Joining sales with COGS to compute CM1/CM2; aggregating by SKU, city and time.
Demand forecasting for PO planning, anomaly detection, automated bid changes.
P&L templates, fee calculators, what-if scenarios for pricing calls.
{{ t.h047 }}
Qcom is closer to Vendor Central than to a marketplace. The platform buys: it drops a PO, you confirm what you can actually serve, you book a slot on their shipment booking portal, and you land the units at the right warehouse in that window. Miss the window and the PO ages out — and your fill rate, the number your category manager is judged on, drops with it.
Units received ÷ units ordered. Below 85% and your next PO gets smaller on its own.
Landing inside the booked slot. Late trucks get turned away and re-slotted days later.
One wrong pack config and the whole line item gets rejected. Fix it at packing, not at the gate.
Book against your real production and inward TAT, not your best-ever week.
{{ t.h048 }}
If you are a D2C brand going into Qcom, this is the one thing to get right before ads, before pricing, before content: supply efficiency. Qcom does not sell what isn't in the dark store closest to the customer. Your listing can be perfect and your ad spend live, and a stock-out in 40% of dark stores just quietly deletes 40% of your demand.
Of all live store–SKU combinations, what share is actually buyable right now. The single number that predicts your Qcom month.
The share of a city's dark stores that carry your SKU at all. Coverage is distribution; OSA is whether that distribution is working.
70% coverage at 80% availability is 56% of the city addressable. Read them together or you will over-read your own growth.
Dark-store-level inventory mapping is genuinely painful — hundreds of stores, inconsistent naming across platforms, no clean feed. Do it anyway. Until you can see stock per store per SKU, every conversation about Qcom growth is a conversation about averages, and averages hide exactly the stores that are killing you.
Normalise every platform's store IDs to your own cluster names once. Everything downstream depends on it.
100 units means nothing. 3 days of cover on a fast store means a stock-out this week.
Spend on a store that cannot fulfil is pure TaCoS with zero revenue attached.
A small set of stores carries most of the volume. Availability there is worth more than coverage everywhere.
{{ t.h049 }}
What this looked like in practice at Frido. Every platform reports in its own dialect — its own SKU codes, its own campaign names, its own date grain. Until that is reconciled, nobody in the room is arguing about the business; they are arguing about whose export is right.
Nine exports, three versions of the truth, decisions taken on last month's numbers.
One dashboard, SKU-level CM2, and a decision you can make in the meeting you're already in.
It is the boring foundation. Skip it and every join, every forecast and every CM2 number inherits the mess.
Demand is not flat across a day — especially on Qcom. Spend where the hours are, starve the rest.
{{ t.h0410 }}
{{ t.p0410 }}
SOURCE Flipkart Minutes build rate and 2026 target — UBS and Bernstein. Total dark stores — Bernstein. BB Now share — Datum Intelligence, Jan 2026.
The three-player mental model is already out of date. Bernstein puts the total across all players above 6,000 dark stores. Assortment, case packs and PO discipline now have to be run across five or six rails, not three — and the newest ones are where the commission bands are still soft.
{{ t.c5t }}
Channels behave like a portfolio — one funds, one grows, one teaches, one drains. The mistake is running every channel to the same target. The move is assigning each channel its role.
{{ t.h051 }}
{{ t.p051 }}
Fund them properly. They are where the next Cash Cow comes from.
TARGET STATE FOR QCOMFeed until they turn Star, or cut honestly. Under-investing here is the classic error.
QCOM · 140% GROWTH, THIN CM2Milk them. Don't over-invest chasing growth that isn't there.
AMAZON + FLIPKART · FUNDS EVERYTHINGBe honest about them. Either it's too early or it's over — say which.
OFFLINE GT · TOO EARLY{{ t.c5asli }}
{{ t.h052 }}
Every strategy that worked had one thing in common: walking in knowing more about the other side's business than they did. Not tactics. Not jargon. Preparation.
Studied their city rollout, onboarding flow and category gaps — knew more than most people inside.
Referenced their specific expansion gap and exactly how Frido fills it as a category anchor.
Sent a one-page note on how the sub-category metrics would improve. Made the CM's job easy.
Meeting inside 48 hours. Partnership framework the same week.
The lesson: a CM gets 100 cold messages a week. Everyone asks. Almost nobody studies. Be the one who studied.
Travel neck pillows were everywhere — except the good ones. Walked into a shop and asked: "do you have a Frido travel neck pillow?" They didn't. Within five minutes the store manager was curious. Within two weeks the distribution conversation had closed.
Growth needed a price lever, and every rupee of discount came straight out of CM2. Onboarding POP UPI moved the incentive off the brand's P&L entirely: POP funds the reward in coins, the listing holds its price, and the customer still sees a better deal.
The lesson: the cheapest discount is the one somebody else funds. Before you touch price — the weakest lever you own, and the only one that comes straight out of CM2 — find out who else will pay for the incentive.
SOURCE My own, from category-manager conversations across Ecom and Q-commerce.
{{ t.h053 }}
A CM's job is category growth — not helping your brand. Show them how your brand grows their category.
SOURCE My own, from category-manager conversations across Ecom and Q-commerce.
{{ t.h054 }}
{{ t.p054 }}
Make the first offer. Ask for 500 bps (5 percentage points) off commission — you'll settle at 200–300 and still win.
"I'll commit ₹2 Cr GMV if you give me hero banner plus net-30." Much harder to reject outright.
Hit CMs before quarter-end when they need GMV, or before BBD when they need catalogue depth.
"Your category grew 28%, the platform grew 12%." Hard to argue with their own numbers.
Whoever speaks first after the ask almost always concedes. Just wait.
Offer things that cost little but feel big — an exclusive one-week launch window.
Sometimes the whole strategy is being physically present where the opportunity is.
{{ t.h055 }}
In this order. Skipping to ads is the most expensive mistake in the list.
SOURCE My own, from client work at Frido and CrepDogCrew.
{{ t.h056 }}
A handful of days carry a disproportionate share of the year. Event days routinely run close to 2× BAU — business as usual — and the brands that win them were not deciding anything in event week. They locked it a month earlier.
SOURCE Platform seller calendars and my own run-up plans. Dates shift year to year — check the current calendar.
Nobody loses a sale event during the sale. They lose it at T−21, when the deal never got submitted, or at T−30, when the PO was cut for a normal month. Event week is execution — the decisions were all made a month earlier.
{{ t.c6t }}
You can make money on every single order and still run out of cash. Marketplaces pay you late. Your supplier does not wait that long. Profit is an opinion. Cash is a fact.
SOURCE CCC band and self-funding rate are my own, across client P&Ls. TCS is Section 52; TDS is Section 194-O.
{{ t.h061 }}
{{ t.p061 }}
SOURCE Illustrative. The three numbers below are the ones you would pull from your own settlement reports.
How long stock sits before it sells. Long on marketplaces because you ship stock in before anyone orders it.
How long the platform takes to pay after the sale. Amazon is fast. Myntra can be months.
How long your supplier lets you wait. The cheapest number to improve, and the one nobody negotiates.
DIO + DSO − DPO. That is your cash conversion cycle.
Every day you cut off that 78 is cash back in your account. On a ₹5 Cr a month business, taking 20 days out frees up about ₹3.3 Cr — without selling one extra unit. Most brands chase margin points for a year and never look at this.
{{ t.h062 }}
{{ t.p062 }}
SOURCE Platform seller agreements and my own settlement reports. Terms vary by account.
On margin it depends on the category. Qcom commission bands run 20–40%, above the referral fee most categories pay on Amazon, and the two worked P&Ls here land at 15.1% CM2 on Amazon (02.2) against 10% on Qcom (02.8). An individual FMCG SKU can still read better on Qcom, because one commission replaces Amazon’s stacked referral, fulfillment and closing fees. What does not vary is the cash: 30–60 days from GRN with deductions taken at source, against 7–14 on Amazon 3P. You are lending the platform your working capital at 0% to sit on a shelf you do not own. That is worth saying out loud in the commission negotiation — almost nobody does.
{{ t.h063 }}
{{ t.p063 }}
You get that ₹43.60 back as input credit — later, and only if the platform filed its invoice properly. So the “24% S&D” in your P&L is really 28.3% of the cash going out this month.
TCS (1%) and TDS (1%) are deducted before you see the money. You get them back when you file and offset. On ₹5 Cr a month that is ₹10 L sitting still, permanently.
Selling FMCG, food or cheap apparel? You charge 5% or 12% GST but pay 18% on fees, freight and ads. Credit piles up faster than you can use it and you are stuck claiming a refund for months.
One on accrual for the board, one on cash for you. The gap between them is your GST float — and it grows fastest exactly when you are growing fastest.
SOURCE CGST Act, Sections 52 and 194-O, and the platform fee schedules. Confirm with your own CA before modelling.
{{ t.h064 }}
{{ t.p064 }}
Amazon storage fees double after 180 days. Pull slow movers at 120 days, not 180 — by 180 you are paying rent on a mistake.
You fund roughly 2× normal stock a month early, get paid afterwards on the platform’s terms, then eat the returns wave. An event is a cash drain for about six weeks before it is a gain.
Manufactured, shipped, and sitting in a warehouse instead of a dark store. No sale, no receivable, all cost. This is what a bad OSA number actually costs you.
You book the sale this month and give the money back next month. At 30% fashion returns, a big growth month sends a bill that arrives after you have already spent the cash on the next PO.
Add up every rupee you cannot touch right now: GST credit not yet claimed, TCS and TDS ledgers, the Amazon reserve, stock in transit, stock in the backend. Most founders have never added it up, and the number genuinely shocks them. That is your blocked capital. Track it monthly.
{{ t.h065 }}
{{ t.p065 }}
15% CM2, 90-day cycle → 15 ÷ 3 = 5% a month.
Anything above that is somebody else’s money, whether you have raised it yet or not.
Brand B earns less than half the margin per order and still grows faster without raising a rupee.
Days beat points. Past a certain size, taking 30 days out of your cycle does more for you than adding 5 points of margin — and it is usually easier. Nobody works on days, because days never show up on the P&L.
{{ t.h066 }}
{{ t.p066 }}
DIO {{ dioStr }} + DSO {{ dsoStr }} − DPO {{ dpoStr }} · CM2 throws off {{ cm2CashStr }} / mo
SOURCE Cash stuck = monthly GMV × cycle ÷ 30. Safe growth = CM2% ÷ (cycle ÷ 30). GMV and CM2% are both on GMV including GST.
{{ t.h067 }}
{{ t.p067 }}
SOURCE Indicative market rates, March 2026. Your actual rate depends on vintage, collateral and filing history.
Compare the interest rate to how hard your cash works, not to your margin. If a SKU sells through 4 times a year at 15% CM2, the cash in it earns about 60% a year. So 30% money makes you richer and 70% money makes you poorer. Most founders compare the fee to their 15% margin, panic, and turn down money that would have paid for itself.
Nobody’s board deck has a cash slide until the month they need one. Every founder in this spot says the same thing: “but we’re profitable.” You were. The money was just somewhere else — in a dark store, in a warehouse, in a tax ledger, in an invoice dated 45 days out. CM2 tells you the business works. Your cash cycle tells you whether you will still own it when it does.
{{ q.explain }}
You made it through.
Now prove it.
Three timed sections, 23 questions, 52 marks — 23 minutes end to end. You cannot see a section before the one above it closes, and you cannot go back.
DIRECT 5×2 · MONEY MATH 6×3 · SITUATIONAL 12×2
Below 75% there is no certificate. Pass or fail, the next sitting is a full month away — nobody grinds this out in an afternoon.
Knight and up are invited in. Kings also get the certificate printed and posted, with a goodie.
"If you read this seriously, you now think like a marketplace strategist." Go build. Go grow.
{{ t.wallh }}
{{ t.wallp }}
FREE FOREVER · NO CARD · ALREADY HAVE ONE? LOG IN
Sahil Panwar
Strategy & Operations (Frido, CrepDogCrew) · ex-Founder, DravyBrands
Sahil's strategy journey started at 18 — not in a classroom, but in the deep end. He launched his first venture and scaled it to ₹40 lakh in revenue in three months. Fast growth, real learning, and a very real failure right after.
He rebuilt with DravyBrands, bought his own office at 21 and built a team of 12.
Then came Frido — deep immersion in CM2 optimisation, Qcom hyperlocal dynamics, channel onboarding (Slikk, Ajio, airport retail) and data infrastructure. PIPPO was born in that period. This playbook exists because he wished something like it had existed when he started.
Connect on LinkedIn learn@insidemarketplaces.com