Decision Library
Spend where the next dollar changes what customers do
Marketing budgets are decided on dashboards that each claim the same customer. A channel can report excellent returns while mostly collecting buyers who would have arrived anyway. A Decision Room compares scaling, shifting, pausing and changing the offer on the return each one actually causes.
The options for a budget
Scale the channel that looks best. Shift money to an untested channel. Pause spend for a period and measure what falls. Fix the offer or the landing page before buying more traffic. Each is a real option, and pausing deserves a fair hearing because it is often the cheapest experiment a team can run.
Measure what spend causes
The dependable evidence comes from holdout tests: a share of the audience, or a set of regions, sees no ads for a period, and the difference in sales between the two groups shows what the spend caused. Platform reports count conversions that followed an ad, including many that would have happened without it, so they usually overstate a channel's contribution.
Once incremental customers are known, two numbers carry most decisions: the cost to win one incremental customer, and the months of gross margin it takes to earn that cost back.
The common mistake
Scaling a channel after one strong quarter without asking what made it strong. A single creative, a seasonal peak or a discount may have done the work. When the budget doubles, returns fall as the channel reaches less interested buyers, and the team learns that the first results were the best ones it would see.
How the room argues a channel
Each option gets a Champion, the pause included. The Inspector checks what each reported figure actually proves, marking platform conversions as partly supported until a holdout or a revenue reconciliation backs them. The Contrarian presses on the story everyone in the room finds pleasing, and the Grandfather asks what a deep discount teaches customers to expect next year.
The engine works the payback sums from your own figures. The Verdict names the leading option, the measurement it rests on, and the test result that would reverse it, and the marketing lead records the Human Decision.
What to weigh
Incremental return
Sales the spend causes, measured against a group that saw no ads.
Payback period
Months of gross margin needed to recover the cost of an incremental customer.
Evidence quality
Whether the measurement design can actually support the claim being made.
Scale headroom
How far spend can rise before returns fall as the audience widens.
Brand effect
What the message, offer or frequency teaches customers about price and value.
Where it goes wrong
Double-counted buyers
A channel reports customers that another channel or word of mouth actually brought in.
Discount habit
Customers learn to wait for the promotion, lowering full-price sales afterward.
Audience fatigue
The same people see the same ad until response falls and cost per sale climbs.
Paid search, measured honestly
A subscription business spends $40,000 a month on paid search, and the platform reports 400 new customers, or $100 each. A four-week holdout across a matched set of regions shows only 35 percent of those customers were caused by the ads: 140 a month. The true cost per incremental customer is about $286. Each customer yields $90 a month in gross margin, so payback takes a little over three months rather than the five weeks the dashboard implied. With weights of incremental return 35, payback 25, evidence quality 20, headroom 10 and brand effect 10, the room favors holding spend flat and moving a fifth of it into a landing page test. The Verdict names what would favor scaling: a second holdout showing an incremental share above 50 percent.
Questions
- What if we have never run a holdout test?
- Then the Inspector marks the platform figures as partly supported, the Verdict carries lower confidence, and the conditions it names will usually include running one. A regional holdout of two to four weeks is often the cheapest evidence a marketing team can buy.
- Who sets the budget in the end?
- Your team does. The Verdict sets out which option the evidence favors and why, and the choice you make is kept beside it in the Decision Passport.
Bring Argumont a Decision worth challenging.
Start with a paid 60 day pilot on your own real Decisions.