Agency or In-House? The Data Rejects the Dilemma

What Is on the Agenda?

Three measurements arrived at the decision table together, and together they changed the question’s shape.

Both curves, one peak

The advertisers’ association’s fifteen-year series carried in-house ownership from 42 to 82 percent — while, in the same survey, external agency use stands at a record 92 percent and the work splits 61-39 on average. The dilemma’s absence from the data is this set’s opening number: everyone already lives in both worlds; the difference belongs to those who draw the border deliberately.

The budget refuses to wait

Marketing budgets fell from 11 percent of revenue to 7.7 and flatlined in the “era of less”; agency, talent and technology allocations are being trimmed together. Abundance could hide a bad division of labour for years; scarcity invoices every misplaced job — the decision left the category of postponable preference and entered the category of efficiency necessity.

Scarcity strikes both doors

The talent measurements documented a symmetry: the biggest deficit, data and analytics, runs at 84 percent on both sides. Building inside is ownership of the shortage, not escape from it; giving outside is delegation, not exemption. No scarcity-free option exists — the choice is between two ways of managing the same scarce well.

The decision lands on ground the series prepared

This set’s question finds the earlier sets’ concepts ready: the rent-title distinction supplied the partner relationship’s ownership language, the delegation ruler supplied the division’s logic, and the decision line pre-asked “who holds the threshold”. On that ground, the agency decision stops being ideology and becomes engineering.

Why Is the Question Being Asked Now?

The dilemma is old. Four forces sharpened it into today’s question.

Tools ended the doing monopoly

Production work cheapening under tools made the outside’s classic promise — “we’ll do it” — ordinary, and widened the inside zone. The same force sharpened the core: when everyone can do, difference pools where doing can’t reach — voice, relationship, judgement. The border question is the tool wave’s natural offspring.

The asymmetry premium was erased

The old agency model priced an information gap: the agency knows, the client pays. The client partly learned, knowledge grew abundant, and the premium fell out of the budget. The one surviving invoice is for what the business cannot produce alone — which is no longer production but mastery and arrangement: hence the relationship’s rewritten contract.

The dependency lessons generalised

What the platform side taught — separating rent from title, measuring exposure — crossed to the partner side: the business whose knowledge, access and decisions pile up outside lives separation day as a hostage negotiation. Partner selection is no longer a procurement decision; it is a dependency-management decision.

The undefined hybrid got expensive

Everyone de facto runs a mixed model; in most, the border is unwritten — work drifts by capacity, habit and the latest pitch. The unwritten hybrid sums both models’ costs and splits their benefits; the written one does the reverse. That is why the question is current: the mixed model became fate; its quality remains a choice.

What Is Wrong?

Four maxims circulate at the table, and each drags the decision onto the wrong axis.

“Agencies are expensive, inside is cheap”

The inside shows up as salary; tools, management, learning time and idle capacity stay invisible — while the outside arrives as an invoice and therefore feels dear. Full-cost accounting usually yields two surprises: the inside dearer than assumed, the outside — with a handover — more instructive than assumed. Price is the tiebreaker after five criteria; never the first question.

“For control, everything must be inside”

The feeling of control bloats the core: every job gets declared “important”, the team sinks into routine, and deep-specialism work stays amateur. Control’s real address is not the work’s address but ownership of the four rooms — with voice, deed, decision and border inside, the edge can race outside in peace. Whoever holds everything deepens in nothing.

“Expertise lives outside; let’s mind our own trade”

The inverse maxim rents out the core: voice judgement at the agency, data in the agency’s account, thresholds set by the agency — the business becomes a guest in its own marketing. That model was comfortable in the fat years; with the asymmetry premium erased it proved both expensive and fragile. Expertise can be rented; ownership cannot.

“Find the right agency and the problem’s solved”

Partner quality is necessary and insufficient: the best agency, facing a client with no named core and no drawn ruler, can only produce a well-meaning drift. The sequence runs the other way — the business draws its own map first, then shops for a partner to fit it. The test serves whoever holds a map; in maphandless hands, the question list becomes a pitch tournament.

The Real Mechanism

The decision is made by a four-instrument arrangement; the instruments install in sequence and run together.

Instrument one: the four-room core

The untransferables are named up front: the brand’s voice and judgement, the deed to the customer relationship, the decision line’s authority and the right to draw the border. The measure is irreplaceability, not importance — and as the border clarifies, a paradox starts working: the business with a clear core opens outward more, not less, because it knows what it is giving.

Instrument two: the five-criteria ruler, three zones

Every job scores on five questions — repetition, brand proximity, skill scarcity, speed need, permanence value — and falls into one of three zones: inside, outside, bridge. The ruler’s most valuable output is the third zone, invisible in the dilemma frame: work done outside today and moved inside tomorrow. Once written, the “who should do it” negotiation disappears; every new job passes through the rule.

Instrument three: the tested partner, the three-clause contract

The partner is chosen by a seven-question test — the central distinction: selling hours, or building capability? — and the relationship is written in three clauses: ownership (every asset in the business’s name), handover (at least one job a year moves inside), reading (a decision page, not a report). The test applies to running relationships as much as new ones: most relationships deserve amendment before termination.

Instrument four: the ninety-day handover, the grow-together ladder

The bridge zone operates through the three-phase handover: shadow, together, solo — with every handover’s permanent output being the handbook: transfer to the institution, not to a person. The capacity the handover frees gets priced at the ladder’s higher rung: strategy, builds, hard problems. The relationship doesn’t narrow; it climbs — the mechanical description of sheltered growth.

Who Is Affected, and How?

Four instruments, four tables, four different opening moves.

The founder-led SME

The core usually sits correctly in the founder’s instinct; the risk is it staying unwritten as scale grows. The opening move is the four-rooms document — half a page, half an hour; afterwards, every external offer hits that half page. The founder’s hours are the firm’s scarcest asset, and the ruler reserves them for work nobody else could take.

The business that gave marketing away entirely

The riskiest profile: voice, data and decisions accumulated outside over years. The way back is staged, not panicked — deed first (accounts and data into the business’s name), decision next (the reading table moves inside), production handovers last. A heart is reclaimed from a prosthesis by degrees; a move made while the relationship is warm is routine, made when it sours, a negotiation.

The business that does everything inside

The opposite profile: control pride keeps the edge inside too, and the team drowns in volume. Its homework is the courage to shrink the core: every job outside the four rooms opens to outside competition while the inside deepens in the irreplaceable. This profile’s first ruler carries the most rows — and brings the most relief.

The agency and the consultant

The framework is the agency’s mirror: the model eyeing the client’s core is indispensable short-term and a cut-list line long-term; the model strengthening the core, carrying the edge and pricing the handover is the new contract’s winner. Answering “the day you leave, what remains?” with a list is the era’s strongest sales page.

Decision Order

The arrangement builds in four steps; the sequence runs core-to-bridge and cannot be skipped.

One: name the core

The four-rooms document is written and the current state audited: which room is partly outside today? Skip this step and the ruler feeds the most valuable assets into the score table — but the core is not scored; it is protected. The border is drawn by principle before any bargaining begins.

Two: draw the ruler

Last quarter’s jobs are scored on the five criteria, the three zones emerge, and the overflow and annual-review rules are written. The table holds three voices — work knowledge, budget knowledge and, where possible, the partner itself; a partner dodging the ruler is the test’s first result.

Three: test the partner, clause the contract

Seven questions asked, answers read against proof, three clauses written. The running relationship passes through the same gate: with a partner who welcomes the clauses, the relationship deepens; with one who resists, amendment is tried first — change before separation.

Four: run the bridge by handover

The most settled job in the bridge zone is picked, the ninety days begin, the handbook gets written, the memorandum signed — and the freed capacity prices the ladder’s next rung together. One or two handovers a year is the healthy tempo: muscle starts with light weights and grows by routine.

Where to Start?

The opening month splits into four finishable weeks.

Week one: four rooms and the deed audit

The half-page core document plus a one-table ownership audit: accounts, data, access — in whose name? Every asset not in the business’s name gets a moving date — routine while the relationship is warm, a negotiation once it isn’t.

Week two: the ruler’s first drawing

A half-day shared table: jobs, scores, three zones. The first drawing’s two classic surprises — heart work living outside, routine living inside — are not accusations but the first two moving orders.

Week three: two questions and the test

Two questions for the current or candidate partner: the day you leave, what remains in our hands — and what will you teach us this quarter? The full test follows the answers, and the three-clause amendment goes on the table. One meeting reveals which contract the relationship lives under.

Week four: the first handover decision

The first candidate is picked from the bridge zone, the ninety-day calendar written with the partner, the receiving name and learning hours reserved. The month ends holding not a preference but a running arrangement’s first quarter plan.

What Not to Do?

The build season’s four traps sit symmetrically at both poles.

Deciding wholesale

“Everything inside from now on” and “give it all to the agency” are the same error: surrendering a job-level decision to a wholesale ideology. Decisions are made job by job — that is what the ruler is for; wholesale verdicts are the confession of rulerlessness.

Opening the core to bargaining

Ceding voice judgement in a busy season, data ownership for an attractive offer, threshold authority for convenience — each looks small, and together they make the business a guest. The four rooms are outside negotiation; temporary support is taken, ownership is not transferred, and temporariness is proven by a return date.

Living a handoverless hybrid

Inside and outside run side by side for years while the bridge never operates: nothing moves, no handbook gets written, the business doesn’t learn. A handoverless hybrid pays two rents at once — the arrangement’s compound return travels only across the bridge.

Casting the partner as enemy, the relationship as zero-sum

“The agency is milking us” and “the client wants to shrink us” are two bottles of one poison. The new contract’s arithmetic is not zero-sum: handover matures the client, and the matured client brings higher work. What sits at the table is not a rival but a ladder companion.

What to Watch?

The arrangement’s health reads on four gauges.

The room audit result

One quarterly question: how many of the four rooms are fully inside and in the business’s name? The target is four, and staying at four — any room’s door drifting outward is the whole arrangement’s alarm.

The handover counter and the handbook shelf

Jobs internalised per year, and the handbooks written. A zero counter means the bridge is rotting; an empty shelf means handovers are going to people — both trigger the same question: which handover is next, and who receives it?

Cost per permanence

The external spend’s annual reading: the invoice divided by the capability it left behind. The business learning more each year on the same budget has the right partner; where learning reads zero, the invoice is rent — and the era of less does not forgive rent.

Ruler freshness

The last scoring’s date and the last zone movements. A ruler older than a year has reverted to habit — tools move the scores, scale adds rooms; a live ruler is a live border.

How Does This Period End?

Read the close through three separations.

The unwritten hybrid loses to the written one

The mixed model is everyone’s fate; its quality is a choice. The business with border, ruler and handover in writing strengthens every year on the same budget, while its unwritten neighbour keeps summing both models’ costs and splitting their benefits — the gap compounds, and the era of less invoices it fast.

The agency market splits in two

Hour-sellers will melt on the cut lists; capability-builders will rise as their categories’ consultants. The dividing line will show in one question: the day you leave, what remains in your client’s hands? The longer list wins — because that list is the new contract itself.

The memorandum shelf enters the valuation

Handbooks and handover memoranda join the documents a valuation reads: the buyer asks whom the marketing depends on, and a full shelf answers “no one”. A learning relationship’s ledger is a sellable asset; a rented relationship’s invoice is only ever an expense.

The big question shrinks into period design

“Agency or in-house” dissolves, in the arrangement-running business, into calendar questions: which job crosses the bridge this year, what sits on the ladder’s next rung? The debate’s disappearance is not neglect but graduation — the question’s shrinking is the keystone finishing its work.

A Solid Digital Foundation

The arrangement stands on four stones, each outliving any partner and any tool.

The four-rooms document

Voice, deed, decision, border — a half-page constitution. Its brevity is its power: known by heart, hit by every offer.

The zoned ruler

Five criteria, three zones, an overflow rule, an annual date. The standing answer key to “who should do it” — and the place where budget talk rises to zone strategy.

The three-clause contract and the test file

Ownership, handover, reading — the relationship’s law; seven questions and their proofs — the relationship’s measure. Goodwill ages; clauses remain.

The memoranda and the handbook shelf

Each year’s record of what was learned: the business’s growth ledger, separation day’s insurance and the valuation table’s favourite shelf. Every job sent outside is measured by one thing: what did it leave behind? — all four stones serve that single question.

Frequently Asked Questions