Which channels produced paid orders?

A company that orders branded goods once will order for events, onboarding and campaigns for years. The first order is a sample of what the account is worth.

Because the match is on the account, every reorder credits the channel that produced it — and in this category that routinely multiplies the apparent value of a channel by five or more.

Order volume concentrates around conference seasons and year-end gifting, and a channel that looks strong in November can be irrelevant in February.

Running the same period against the same period a year earlier — which is possible because there is no attribution window — separates a channel's real performance from the calendar's.

Once a client moves from ad hoc orders to a branded program — a company store, an onboarding kit, a recurring uniform allocation — the relationship stops being a series of purchases and becomes an operating arrangement. Orders arrive continuously and in small amounts, often placed by dozens of different people inside the client organization, and the original buying contact may never place another order personally. The revenue is larger and steadier than anything the account produced before, and it is the hardest of all to attribute, because the rows now carry the names of recipients rather than the buyer.

The practical answer is to export program revenue against the account's original buying contact rather than against whoever placed each individual order. That is a decision about how the export is built, not about the matching, and it is the difference between a program crediting the channel that won it and a program dissolving into hundreds of unmatched rows. Aggregating by account and month is usually the cleanest form: one row, the buying contact, the total, the period. It also keeps the row count sensible when a single client generates thousands of small orders a year.

Programs are also where the value of an acquisition becomes obvious. A first order of a few hundred dollars that becomes a program billing every month for four years is not a small sale followed by luck; it is what the channel that produced the account actually bought. Ranking on program revenue rather than first-order value changes which channels look worth funding, and it changes the case for the account management effort that converts an order into a program. That second finding is often the more valuable, because it is about something the business directly controls.

Branded merchandise is bought by marketing managers, HR coordinators and executive assistants, which are among the most mobile roles in any organization. A buyer who enjoyed working with you moves to a new employer and reorders there within a year, and that new account arrives looking like a fresh acquisition from nowhere. Every report in the category treats it that way, which means the channel that produced the original relationship is credited with one account when it produced two. The effect compounds across a decade in a way that nothing currently measures.

Where the contact detail follows the person — a mobile number usually does, a work email does not — the new account links back to the original source automatically. Where it does not, the revenue sits in Direct / Unknown, which is the honest outcome and a standing reminder to capture more than one contact detail per buyer. Files carrying both a phone number and an email give each sale two chances to match instead of one. It costs nothing to include both columns, and the improvement shows up in the very next report rather than after a process change.

This also affects what should be kept. A retention window that purges old imports on a schedule is good practice and worth setting, but the gap this match exploits can run to several years, so a very short window removes the history that makes a buyer's move visible at all. The right setting is a business decision balancing data minimization against how far back a useful comparison reaches. Most distributors settle on something long enough to hold two or three buying cycles and no longer, and the purge then runs without anybody remembering it.

Every distributor carries unbilled work. A client asks for a spec sample, a virtual proof, three artwork variations and pricing across four quantities, and then goes quiet for a month. None of it is invoiced, all of it costs studio and account time, and the volume varies enormously by where the inquiry came from. An inquiry source producing constant proof requests and few orders is expensive in a way no cost-per-lead figure records, and the cost lands on a team that does not control the media budget and is never asked about it.

Ranking sources on paid order revenue rather than on inquiries produced makes the difference legible immediately. A channel producing forty inquiries and three orders and a channel producing eight inquiries and six orders look identical on volume and nothing alike on revenue. Putting the quote or inquiry list in the lead file and the paid orders in the sales file shows each source's inquiry count beside its orders and revenue, which is the close rate in all but name. That figure is what should be taken into a renewal conversation about any paid listing or directory, and it is one nobody currently has.

There is a caveat about seasonality that is specific to this trade. Close rates fall sharply in peak season, when buyers are quoting several distributors at once for the same conference or year-end program, and rise off-peak when a buyer has already decided who to use. A source evaluated only across a peak looks worse than it is, and one evaluated only off-peak looks better. Comparing the same window across two years removes the effect, and it is available immediately because the orders were recorded for invoicing rather than for tracking.