How to Build a Dependency Inventory

What Is the Problem?

In an uninventoried business, dependency hides in three costumes.

Dressed as a success statistic

“Eighty percent of our traffic comes from search” was read for years as a line of pride; the same sentence is an exposure report. The publisher asymmetry data is the proof: in one storm, single-gate small publishers lost 60 percent of that traffic while many-doored giants escaped with 22. An eighty percent arriving through one gate is not an asset; it is a debt with an unknown maturity date.

Dressed as scatter

Channel data lives on separate dashboards: search in one place, social in another, the marketplace in a third. On separate panels no single channel looks dangerous; assemble the total table and the concentration appears. Dependency winters in the scatter of the data — merging the table onto one page is the first and cheapest raid on its winter quarters.

Dressed as habit

“We’ve always gotten business from there” declares a habit, not a strategy. Habit grows unquestioned and leans a little harder on the same gate each year — the inventory is the first document that confronts the habit with a number. The confrontation usually produces two feelings at once: surprise at the share, and discomfort that the share had never been questioned.

Why Does It Happen?

The uninventoried state has three causes.

What works is assumed unmeasurable

The measuring reflex is reserved for problems; the working channel gets “don’t touch it” treatment. Yet risk accumulates precisely where things go well — nobody over-leans on a weak channel. The gate that earns the most is the gate that most needs measuring.

Rent and title were never separated

The channel list is kept flat; which channel is the business’s property and which is the platform’s favour goes unmarked. Without the rent-title framework the inventory shows only volume, not risk — the same forty percent is strength on a titled channel and fragility on a rented one.

The bad scenario is not discussed

“What if this gate narrows” counts as pessimism in meetings and goes unasked. An unasked question does not disappear; it postpones its answer to the crisis day. The dismantled-feature case shows the postponement’s invoice: the platform asks the question on its own calendar.

How Is It Done?

The inventory builds in three steps; half a day in total.

Step 1: list the gates, write the two shares

All arrival gates go into one table — search, social, marketplace, maps, advertising, email, direct, referral — and each gate gets two shares: traffic share and revenue share. They are written separately because they diverge: the gate bringing 60 percent of traffic may carry 20 percent of revenue, while a quiet 10-percent gate carries half the income. The share that decides is the revenue share.

Step 2: give every gate an ownership and control note

Each row takes two labels. Ownership: titled (list, members, direct) or rented (algorithm, ranking, feed)? Control: who can change this gate’s rules, and how do we learn of a change — in advance, through a documentation note, or not at all? The combination rented + changes-without-notice is the inventory’s red cell.

Step 3: run the narrowing test, write the thresholds

Each red cell gets one question: if this gate narrows 40 percent within six months, which work stops and which cost runs uncovered? The answers go into the table, and two thresholds go beside them: an alarm when any single rented gate’s revenue share crosses a band (a third is a sensible line for most businesses), and an investment trigger when the titled channels’ combined share falls below its floor. The thresholds are the compass of the many-door arrangement.

How Long, at What Cost?

The inventory’s economics are deliberately small; the resistance lives not in cost but in the confrontation.

Half a day, on existing data

The gate list and the shares compile from existing dashboards in half a day; perfect data is not required — even rough shares make the table discussable. Waiting for the perfect inventory is the politest form of staying uninventoried.

Four times a year, ten minutes each

Once built, maintenance is light: shares refresh quarterly, labels get reread, thresholds get checked. Even one changed row starts a conversation — and the conversation starting before the crisis is the inventory’s entire return.

The yield: decisions change address

The inventory reduces scattered channel debates to one question: which cell does this investment fix? New content, a new channel, a new tool — everything argues against the table. Investment flows not to the shiniest opportunity but to the reddest cell.

The Common Mistake

The inventory gets built; three misreadings drain its value.

Mistaking traffic share for revenue share

The high-volume gate is reflexively crowned “most important”; but unconverting volume is noise in this table, not weight. Sorting always runs on the revenue column — the visibility-selection lesson applies here too.

Mistaking the titled channel for a free one

The title column can slide into “we’re safe” complacency; but an unworked list and an abandoned member base are a titled house with a leaking roof. A titled channel is not riskless — it is the channel whose risk sits in your own hands; the difference is large, and it is maintained with labour.

Mistaking the inventory for the whole job

The table is a diagnosis, not a treatment. Once the red cell is found, two jobs begin: the registered-relationship arrangement and the narrowing drill. The business that stops at diagnosis is the patient who frames the X-ray and skips the treatment.

Frequently Asked Questions