How FOCUSADS decides where the next pound goes
A plan is the output of a decision system, not a template. It reads your economics, your buyer and your market, removes what cannot work, ranks what can and funds only what can learn something at your budget. Then it checks the decision again every day.
Re-evaluated daily · market data checked 3 November 2026
Seven moves from brief to budget
Every brief passes through the same sequence. Nothing reaches a plan without clearing the move before it.
1
Read
Your market, industry, buyer, goal, budget and economics, plus the stack you already run.
2
Qualify
Turn the brief into a decision frame: what an outcome may cost, how long the buyer takes, what the team can absorb.
3
Gate
Remove every option that cannot legally, practically or economically run for this brief.
4
Rank
Order what is left by fit to this buyer, this market and this budget.
5
Allocate
Give each funded line a role and enough money to learn. Leave the rest named and explained.
6
Govern
A strategist approves the plan before it becomes a quote. Nothing moves in your accounts without a recorded approval.
7
Re-evaluate
Market facts, delivery data and confirmed outcomes flow back in. The ranking is refreshed every day.
Six stages, one accountable record
The same six stages run the website plan, the onboarding and the monthly review. Each has an input, a decision, an output, an owner and the evidence that closes it.
Stage 1
Define the business outcome
- Input
- Goal, value of an outcome, margin, sales cycle, capacity
- Decision
- What a result may cost and what counts as one
- Output
- A permitted cost per outcome and a result definition
- Owner
- You and your strategist
- Closed when
- The limit is written into the plan and approved
Stage 2
Understand the buyer
- Input
- Industry, sub-industry, segment, trigger, decision maker
- Decision
- Where the buyer is in the decision and what moves them
- Output
- A segment with its trigger and its proof of purchase
- Owner
- Strategist
- Closed when
- The segment is named on the plan
Stage 3
Evaluate eligible media
- Input
- Country registry, formats, policy, account access, minimum test sizes
- Decision
- Which channels can run, and what job each one does
- Output
- Funded lines with roles, plus every exclusion with its reason
- Owner
- FOCUSADS Superintelligence, reviewed by the strategist
- Closed when
- Every funded seller has a dated check
Stage 4
Prepare conversion and measurement
- Input
- Landing pages, forms, consent, tracking, CRM, offline outcomes
- Decision
- How a result will be counted and how sure that count can be
- Output
- A measurement plan and the fixes that come first
- Owner
- Delivery team with your developer
- Closed when
- Test events reach your CRM or order system
Stage 5
Launch and verify
- Input
- Creative, links, accounts, events, budgets, permissions
- Decision
- Whether every line is safe to spend on
- Output
- A launch checklist signed off line by line
- Owner
- Delivery team
- Closed when
- First data matches the plan on spend and events
Stage 6
Review and reallocate
- Input
- Confirmed outcomes, cost, volume, market changes
- Decision
- Scale, hold or stop each line
- Output
- A new allocation with its reason, waiting for your approval
- Owner
- Strategist and you
- Closed when
- The decision log records who approved what and why
Six signal families
Each family is built from many signals, kept current by research and by delivery data. The families are public. Their weights, thresholds and interactions are not, because they are the product.
Market reality
Which sellers exist and sell ads in the country, their share of local search and attention, platform policy, currency and price level, consent regime and the date each fact was checked.
Buyer fit
Intent stage, the trigger that starts a purchase, who decides, how long it takes and where comparison happens. Drawn from 2,176 market and segment combinations across 169 sub-industries and 68 buyer segments.
Commercial economics
The value of an outcome, margin, lead-to-sale rate, lifetime value horizon and team capacity. Economics limit a channel before any platform does.
Execution readiness
Feed quality, landing pages, tracking, creative and marketplace listings. A strong channel that cannot run well this month waits.
Evidence quality
How each fact was verified, how recently and how confidently. Old or thin evidence lowers confidence and is shown on the plan.
Risk and compliance
Restricted categories, sector rules, platform eligibility for the advertiser and the country, and the consent model the plan must respect.
Hard gates before any ranking
Ranking only starts once every option has cleared 14 hard rules. A channel that fails one is removed, whatever its potential, and the plan says which gate removed it.
Gate 1
Availability
The seller must sell the format in your country, to an advertiser like you, through a route we can run.
Gate 2
Category
9 restricted categories are checked against each platform’s policy before a line is considered.
Gate 3
Viability
A line needs enough budget to learn. Below that, money is concentrated rather than spread thin.
Gate 4
Readiness
No feed, no listing, no app or no verified offline partner means the line waits for the fix.
Gate 5
Capacity
The plan never buys more leads than your team can follow up.
Gate 6
Consent
Measurement follows the privacy regime of each market. We plan for 11 regimes today.
Allocation discipline
Ranking decides the order. Allocation decides who gets money, and how much, so each line can prove or disprove itself inside the test window.
Roles, not a list
Each funded line is primary, supporting or a test. The role sets its share and what success looks like.
Local sellers compete
Verified local search engines, marketplaces and publishers compete for the same budget as the global platforms, on the same terms.
Concentrate first
Smaller budgets fund fewer channels. Breadth comes with budget, never before it.
Every pound accounted for
Media, fees, production and tracking appear as separate lines. The allocated total always equals the permitted media.
Re-evaluated every day
Markets move. A platform changes its policy, a local seller opens a format, a channel’s cost per outcome drifts. Each day the system works through the same loop for every market, channel and sector we operate in.
- 1
Refresh the facts
Market records that are due are checked again on the sellers’ own pages. Anything older than its validity window drops out of new plans.
- 2
Re-run active plans
Every live plan is ranked again against today’s facts and the latest confirmed outcomes.
- 3
Surface the change
When the best allocation moves, your strategist sees the proposed change and the reason before anything happens.
- 4
Feed the campaigns
Approved signals reach your campaigns in real time through the FOCUSADS integration in your own accounts.
- 5
Record it
Every proposal, approval and change lands in the decision log you can read.
What we publish, and what stays private
Clients see the full reasoning behind their own plan: every funded line, every exclusion, every assumption and every change. The public site explains the system. It does not publish the weights, thresholds or rule texts, for two reasons.
They are the product
Years of market research and delivery data sit in how the signals are combined. Publishing the recipe would hand it to anyone who wants to copy it.
They change
The model is tuned as markets and outcomes change. A published number would be out of date the day after it was printed.
Nothing about your plan is hidden from you. If a strategist overrides a recommendation, the plan says so and why.
Tested before it is trusted
We stress the model the way a sceptical buyer would. In 1,920 runs across 120 briefs with the inputs disturbed, the top recommendation stayed the same in 95.7% of cases and the funded set in 95.1%. 92 automated tests check that every hard rule fires when it should, on every release.
- Top choice stable
- 95.7%
- Funded set stable
- 95.1%
Where the method stops
- It plans; it does not forecast sales. A ranking is a decision about where to test first.
- It is as good as the economics you give it. Missing figures are replaced by typical values, and the plan says which.
- Lift from offline media is proven with geo tests and holdouts, not with attribution we cannot defend.
- A strategist reviews every plan before it becomes a quote. The system narrows the options; people make the decision.
Methodology questions
Why don’t you publish your scoring weights? The way signals are combined is what makes the plan better than a template, and it is tuned continuously. Clients see the full reasoning for their own plan, including every exclusion and assumption.
How often is a plan re-evaluated? Every day. Market facts that are due are re-checked, active plans are re-ranked, and any proposed change waits for approval before it reaches your accounts.
Does a higher-ranked channel always get more money? No. Ranking sets the order; allocation sets the money. A line needs a role and enough budget to learn, so a smaller budget funds fewer channels.
Do local search engines and publishers really compete with Google and Meta? Yes. Verified local sellers compete for the same budget on the same terms. In markets such as South Korea or Czechia that changes the plan.
Can a strategist overrule the system? Yes, with a written reason that appears on the plan and in the decision log.
Is the plan a forecast? No. It is a decision about where to put the next budget and how to tell whether it worked.
See the method on your own brief
Five questions get you a first allocation. The full planner adds your economics and shows every reason.