Fullerly

Fullerly Board · Marketing

Ads eat 38% of revenue

One saved meeting. The verdict and the plan come first; the full debate is underneath if you want to check the reasoning.

  • A saved meeting, unedited
  • Simulated board
  • Schools of thought from the public record
  • Not advice

The decision

Ad spend is 38% of our revenue. Cut it, keep scaling, or shift the money into retail stores and a new market?

RESHAPE

RESHAPE: switch off only the ad campaigns that fall below your break-even and whose buyers do not reorder, aiming for 25% of revenue, ask the chain for its terms in writing before its 5-week deadline, and leave Malaysia for later.

How the five seats voted

  • RESHAPEThe Go-to-Market OperatorYour one-week pause shows ads drive orders.
  • RESHAPEThe Customer-Obsessed ScalerYour July pause measured the ad tap, not the customer.
  • RESHAPEThe First-Principles EngineerYour 38% is an average of good and bad spend, and the cost floor says some of that money is buying sales you lose money on.
  • GOThe CompounderAt 62% gross margin, every ad dollar has to bring in about SGD 1.61 of sales just to pay for itself.
  • RESHAPEThe ContrarianYour July pause answered the wrong question, and the retail and Malaysia bet asks you to spend cash against figures nobody in this room has.

The plan

  1. This week · You

    Start the 30-day test on 28 October: switch off campaigns under SGD 1.61 of first-order revenue per ad dollar whose buyers did not reorder, worst first, until spend reaches SGD 29,167 a month or that list runs out, and judge the result against the same-season 2025 baseline.

  2. This week · You

    Email the chain asking in writing for the listing fee, the stock they need, their margin, payment terms and whether unsold stock can be returned.

  3. Within 7 days · You

    Add up last year's salaries, rent and other running costs and compare the total with the SGD 378,000 ceiling.

Then decide this yourself

What do your salaries, rent and other running costs add up to in a year?

GO backs one option as it stands, KILL means do nothing or stop, RESHAPE means go ahead only with a named change.

In the app you make the final call in your own words; it goes into a decision journal with a 90-day review date.

Fullerly can do this next

Fullerly Data Live

Market data built by Aye Bot, the company behind Fullerly. Google Trends across 15 countries is on sale per result on the Apify Store.

A separate tool. Nothing from this meeting is sent to it.

See Fullerly Data

Read the full meetingThe brief, five cases with their sources, the cross-examination and the Chair's reasoning

Step 1 · The decision brief

Ad spend is 38% of our revenue. Cut it, keep scaling, or shift the money into retail stores and a new market?

The options

  • Cut ads to 25% of revenue; Keep scaling ads; Move budget into a retail listing and Malaysia

Constraints

cash: about SGD 140,000; time: a retail chain wants an answer in 5 weeks; people: 6 staff

Step 2 · Opening round

Five seats vote and make their case.

Each seat argues independently: it votes GO, KILL or RESHAPE, makes its case, then puts one question to the owner and proposes a 30-day test.

The Go-to-Market Operator

A school of thought, drawn from the public record of Jason Lemkin, Madhavan Ramanujam, Nick Mehta

RESHAPE

The question this school always asksWho pays, how much, and are they still paying next year?

Your one-week pause shows ads drive orders, but it does not show whether those orders came from new customers or from your own customers being sold back to you.

The 30% drop is a blended number, and the real question is whether next year's revenue depends on paying again for customers you already own.

My bias is that I read everything as a subscription funnel, so discount this lens if your product is bought once and never again. Here is what it sees. At 62% gross margin minus 38% on ads, you keep 24% of revenue before salaries and rent, and the success line is above 10% net. A one-week pause cannot tell you who stopped buying. If a good share of that 30% were returning customers, then you are paying advertising rates to reach people whose email you already have. That is the cheapest margin in the business, and cutting it does not slow new growth. Retail and Malaysia add two new unknowns before you have found out who pays you and whether they come back. Learn that first.

The question you must answer

What share of your monthly revenue comes from customers who have bought from you before?

The 30-day test

Starting Monday, split ad reporting into new and returning buyers. For 30 days, exclude past customers from paid ads and send them a reorder offer through email or SMS. Leave the campaigns aimed at new customers unchanged.

Returning-customer revenue falls by less than the ad money you stopped spending on those customers, and new-customer orders hold at their prior 30-day level. If returning revenue falls by more than the money saved, put that spend back and stop.

Simulated seat. The words above are generated and speak for no one named here.

The Customer-Obsessed Scaler

A school of thought, drawn from the public record of Jeff Bezos, Jack Ma, Bret Taylor

RESHAPE

The question this school always asksWhat does the customer get, and can you do it at 100x the volume?

Your July pause measured the ad tap, not the customer, so you cannot yet tell whether 38% buys you customers or rents them.

A 38% ad line is cheap if it buys customers who return on their own, and ruinous if it pays for every order twice.

Ask what the customer gets. Your sales fell 30% the week ads stopped, which means paid clicks carry a lot of your revenue. It does not tell you whether those people come back unpaid. If they do, scaling is right and the margin arrives later. If they don't, you are buying the same customer again each month, and no amount of volume fixes that. Malaysia with 6 staff and SGD 140,000 is spreading thin, not scaling. The retail shelf is the more interesting door, because it reaches customers who never see your ads. It can probably be reversed if the terms allow an exit, so press the chain for numbers now. Own this school's bias: it will happily fund growth past the point of proof. That is why the cutting rule comes first.

The question you must answer

What did your salaries, rent and other running costs add up to last month? Without that number, nobody in this room can tell you whether 10% net profit is within reach.

The 30-day test

From Monday, tag every order for 30 days as first-time or repeat, and record which ad channel it came from. Work out each channel's ad cost against the first-order revenue it brings. Email the chain this week asking for the listing fee, the stock they need and their margin, in writing.

Cut any channel where ad cost is above 62% of the first-order revenue it brings and those buyers do not reorder within 30 days. Pass if, after the cuts, total ad spend moves toward 25% of revenue and weekly revenue holds at or above your pre-test weeks. If revenue drops when the cuts land, stop cutting and return to the ads question with that data.

Simulated seat. The words above are generated and speak for no one named here.

The First-Principles Engineer

A school of thought, drawn from the public record of Elon Musk, Jensen Huang, Andy Grove, Morris Chang

RESHAPE

The question this school always asksWhat does physics or the cost floor say, and where is the bottleneck?

Your 38% is an average of good and bad spend, and the cost floor says some of that money is buying sales you lose money on.

Your July pause measured what all your ads do together, but the decision depends on what the last dollar of ads does, and those are different numbers.

Start with the arithmetic. At a 62% gross margin, every ad dollar has to bring back about SGD 1.61 in sales just to break even. Across the whole budget you get about SGD 2.63 per ad dollar, which looks healthy. But that is an average. Some campaigns probably return well above that, and others probably fall below the floor and lose you money on every sale. The July pause cut everything at once, so it could not show you which was which. Delete the losers and keep the winners. That is how you get from 38% toward 25% without the 30% drop. Malaysia and a retail chain are new bottlenecks you would be adding with six people, before you have fixed the one you already have.

The question you must answer

What do salaries, rent and your other running costs add up to each year? Without that number, nobody in this room can tell you whether 10% net profit is within reach.

The 30-day test

On Monday, rank every campaign and channel by revenue per dollar spent over the last 90 days. Switch off the bottom group until total spend falls to about 25% of revenue. Hold that for 30 days and compare it with the previous 30.

Lost revenue is less than 1.61 times the ad money saved, and total revenue is flat or up. If it falls more than that, restore the cut campaigns and stop.

Simulated seat. The words above are generated and speak for no one named here.

The Compounder

A school of thought, drawn from the public record of Warren Buffett, Charlie Munger, Mark Leonard, John Malone

GO

The question this school always asksWhat is it worth, and what does it compound into over 20 years?

At 62% gross margin, every ad dollar has to bring in about SGD 1.61 of sales just to pay for itself. You cannot tell which of your SGD 532,000 a year clears that bar, so cut the worst spend first and keep the rest.

The July pause measured what all your ads together are worth, not what the last dollar is worth, and the last dollar is where the waste sits.

What is a dollar of ads worth? Not the average dollar, the last one. Your July week showed that ads as a whole earn their keep. It said nothing about the final third of the budget, and that is usually where returns fade and nobody notices. Cutting by channel and campaign, worst first, keeps the dollars that work and drops the ones that don't. That is the See's Candies test in reverse: find the spend that needs more capital every year to stand still. Retail and Malaysia may be fine businesses someday. Today you cannot name the chain's margin or the up-front cost, and a deal you can't price is one to walk past. If the chain wants you, it will still want you once you know your numbers.

The question you must answer

What do salaries, rent and all other running costs add up to in a year? Without that figure, no one in this room can say whether 10% net profit is within reach.

The 30-day test

On Monday, rank every ad channel and campaign by revenue per ad dollar over the last 90 days. Switch off campaigns from the bottom up until spend runs at 25% of revenue. Hold that for 30 days and track weekly revenue, ad spend, and gross profit after ads in SGD.

Gross profit after ads is higher in SGD than in the prior 30 days, and revenue falls no more than 5%. If you fail either, turn the last campaigns you cut back on.

Simulated seat. The words above are generated and speak for no one named here.

The Contrarian

A school of thought, drawn from the public record of Charlie Munger, Peter Thiel, Alex Karp, Kazuo Inamori

RESHAPE

The question this school always asksWhat is everyone here getting wrong?

Your July pause answered the wrong question, and the retail and Malaysia bet asks you to spend cash against figures nobody in this room has.

Turning everything off for a week measures what all your ads do on average, while the decision turns on what the last third of spend does.

Everyone reads the 30% drop as proof that every ad dollar earns its keep. It proves only that some ads matter a great deal. Switch everything off and you lose the good campaigns with the bad ones, so the drop tells you nothing about the bottom slice you would actually cut. Invert it: how does this fail? It fails if you keep feeding campaigns that would not be missed, or if you hand a chain your stock and margin before you know what they cost. Moving from 38% to 25% means cutting about a third of ad spend. At a 62% gross margin, that cut pays as long as revenue falls by less than about a fifth. That is a wide margin for error, which is the only kind worth having.

The question you must answer

What do your salaries, rent and other running costs add up to in a year? Without that figure, nobody can tell you whether 10% net profit is within reach under any option.

The 30-day test

Monday, rank every campaign by sales it can be traced to per dollar spent, switch off the lowest-ranked ones until about a third of ad spend is gone, and compare each week's revenue with the same weeks before.

Revenue falls less than 21% over 30 days (the point where saving 13 points of ad spend stops paying at a 62% gross margin), or restore the cut campaigns.

Simulated seat. The words above are generated and speak for no one named here.

Step 3 · Cross-examination

Each seat takes on another.

Each seat cross-examines another seat's case. Then each one says whether it holds or changes its vote. In this meeting, no seat changed its vote.

  1. The Go-to-Market Operator questions The Compounder

    You say cut the worst campaigns first, judged on whether each ad dollar returns SGD 1.61 in sales. That measures the first order, not the customer. A campaign that looks weak on the first order may be the one bringing in people who come back next month without being paid for again. Cut it and you have dropped your best customers to save on your cheapest sales. So before you cut anything, check which campaigns bring buyers who order again unpaid. On the chain, "walk past" is the wrong call. You have five weeks, and that deadline gives you leverage. Ask them in writing for the listing fee, the stock they need, their margin and the exit terms. If they cannot price it, then walk. Skipping that question means saying no to a deal you never priced.

    Holds RESHAPEbecause nothing here shows whether ad-bought customers keep paying, and that is still what decides between cutting and scaling.

  2. The Customer-Obsessed Scaler questions The Compounder

    You say a deal you can't price is one to walk past, and that the chain will still want you later. Nobody here knows that. Walking past is the one move that might not reverse. Asking the chain for its fee, its stock order and its margin costs you one email and keeps the door open, so send it this week. Your worst-first cut is right in spirit, but it ranks campaigns by the ad platform's own report. That report claims credit for loyal customers who would have bought anyway. Then you cut the wrong campaigns with confidence. Start from the customer instead: who is new, who is returning, and who comes back without being paid for. Then you know what each ad dollar actually buys.

    Holds RESHAPEbecause nothing here shows whether the 38% buys new customers or rents back ones you already have.

  3. The First-Principles Engineer questions The Compounder

    Your SGD 1.61 bar is the right test for the last ad dollar, and I agree with cutting worst first. But "a deal you can't price is one to walk past" treats ignorance as a fixed cost. It isn't. Getting the chain's listing fee, stock requirement and margin costs one meeting this week. The deadline is five weeks away, and the chain will not wait while you learn your numbers at your own pace. More important: today every customer arrives through a paid tap, and July showed what happens when that tap closes. That single channel is the bottleneck. A shelf reaches buyers without paying for each click, so it may be the only route to cheaper customers. Price it before you refuse it. Malaysia, yes, walk past that.

    Holds RESHAPEthe board agrees the spend needs cutting by channel, and the open retail question needs pricing, not dismissal.

  4. The Compounder questions The Customer-Obsessed Scaler

    You call the retail shelf the more interesting door and say it can probably be reversed. "Probably" is doing a lot of work there. Nobody here knows the listing fee, the stock the chain wants, or the margin it takes. You have about SGD 140,000 in cash. Stock sitting on a chain's shelf is cash you cannot pull back quickly, and the chain set the five-week clock, not you. A deadline plus missing terms is how a permanent loss starts. Ask the chain for its fee, stock and margin in writing. If it will not give them, the answer is no and it costs you nothing. Meanwhile the money already in play is your SGD 532,000 a year of ads. Some of it falls below the SGD 1.61 floor. Fix that first.

    Holds GObecause cutting the spend that falls below the floor is still the only move where you know the numbers and can reverse it.

  5. The Contrarian questions The Compounder

    Seat D, "if the chain wants you, it will still want you" is a guess dressed as patience. The chain set the five weeks, not you. Walking past an unpriced deal is sound. Refusing even to ask the price is an error of omission with good manners. Ask the chain this week for the listing fee, the stock they need, their margin and the exit terms. Asking is free, and so is saying no. Second, "cut the worst spend first" assumes you can already see which campaigns clear the SGD 1.61 bar. If your tracking can't show that, you will cut by gut and call it discipline. Third, this whole room is arguing over the 24% while nobody knows your salaries and rent. Your net margin could already be above 10%, or far below it. Get that number before anyone touches a dial.

    Holds RESHAPEbecause the room has converged on the right caution but the wrong timetable, and the fix is to get the retail terms and your real cost base now, not to switch to GO or KILL.

Step 4 · The Chair's reasoning

All five seats back cutting the weakest ad spend first rather than a flat trim, and The Compounder's GO on Option 1 is the same move as the four RESHAPE votes, so the room is closer to unanimous than the tally shows. Cross-examination added two guards: ask the chain for its price before refusing it, and keep any campaign whose buyers come back and reorder without being paid for again.

The vote

VoteOpeningAfter cross-examination
GO11
KILL00
RESHAPE44

Where they agree

  • The July pause measured all ads together, not the weakest campaigns you would actually cut
  • Cut campaign by campaign, worst first, against SGD 1.61 of first-order revenue per ad dollar, the break-even at your 62% gross margin
  • No retail cash until the chain puts its fee, stock and margin in writing, and Malaysia waits this quarter

The live split

You must choose whether to cut weak campaigns now on first-order returns, or hold spend at 38% for 30 days until you can see which campaigns bring buyers who reorder without being paid for again.

Against your constraints: The chain's deadline falls 5 weeks from today, after the ad test ends on 27 November and before your 31 December decision date, so your answer to the chain has to rest on its written terms, not on the ad test.

Bring your own decision

Fullerly Board is in private testing. Ask for early access and bring the decision on your desk.

Ask for early accessHow the Board works →

Each seat is a labelled simulation of a way of reasoning, drawn from the public record of several operators. No seat is any one person, and none of them took part. Real quotes appear only in the boxes marked "From the public record", attributed and dated, kept apart from the simulation. Fullerly Board is in private testing.

Not investment, legal, financial, tax or medical advice. No person whose public words appear is affiliated with Fullerly Board or has endorsed it. Fullerly Board is provided by Aye Bot Pte. Ltd. (UEN 202549226G), Singapore.