Which channels produced installed, billing accounts?

Circuits, ports and installations take weeks or months. By the time the account bills, the campaign that produced it is two reporting cycles in the past.

Matching on the account rather than a session credits the campaign whenever billing starts, which is what makes the marketing budget comparable to the revenue it produced.

Agents, VARs and referral partners bring a large share of business connectivity revenue, and their contribution is usually accepted rather than measured.

Added to the lead file with each partner as the source, the deal registration list ranks on installed, billing revenue next to direct campaigns — which may be the first time the two have been comparable.

Connectivity revenue is sold in thirty-six month terms and lost in month nine. A customer outgrows the service, moves premises, is acquired, or leaves after a bad outage, and whatever the contract said, the billing stops. Marketing is judged on the order booked, which records the promise. The finance team lives with the months actually billed. Those two numbers can differ by a third, and the difference is not spread evenly across the channels that produced the accounts. A provider reading order value alone is measuring how well it sells rather than what its marketing brought in, and the two diverge most on the cheapest accounts.

Exporting billed months rather than order value settles it without anybody having to define churn. The report credits each month of billing to the channel that produced the account, so a source whose customers stay for four years accumulates against it while a source whose customers leave after the first invoice simply stops accumulating. No status flag is required and nothing is inferred; the money either arrived in a given month or it did not, and the report shows what arrived.

This is usually the measurement that separates two channels a provider has been treating as interchangeable. Price-led comparison traffic and a partner introduction can look identical on order value and behave nothing alike afterwards. Once the ranking is on revenue billed, the cheap channel that produced a long tail of short-lived accounts falls, and the expensive one that produced a handful of accounts still billing three years later climbs past it and stays there. Nothing about the campaigns changed between those two readings. Only the quantity being ranked changed, which is usually enough to move a budget.

An account that lands with twenty seats and one circuit rarely stays there. New offices, more handsets, a second site, an upgrade from a broadband tail to a leased line, extra call paths added before a busy season: the base grows quietly and the growth is handled by an account manager or a self-service portal. None of it produces an inquiry, so none of it appears anywhere near the campaign that won the customer in the first place. By the third year the original order can be a minority of what that account bills each month.

Because the match is held on the account contact rather than on the original order, every later billing line credits the original channel. That turns the ranking from a measure of how many orders a channel produced into a measure of how much billing it is responsible for today. The smallest-looking acquisition channel may turn out to have brought in the accounts that have grown the most since they landed. That is a different question from which channel produced the most orders, and it is the one a growth plan is built on.

A product or service column makes the growth legible rather than merely larger. Mapped as Product / service, it separates connectivity from voice from managed services, showing the revenue in each line and how much of it traces to a channel. It does not rank channels within each line, but it does show whether the attributed revenue is concentrated in one product or spread across several, and a single blended figure hides that completely from whoever is setting the budget. Adding the column costs one field in the export and changes what the report is able to say.

The revenue side exists because you invoice. An account contact's email or phone, an amount and a date, exported by month from the billing platform, is the whole file. It is the same report used to reconcile revenue, so nobody has to build anything, and a completed year can be uploaded on the first afternoon because those records were kept for accounting rather than for measurement and have been sitting there since. Nobody has to reconstruct anything, no month needs reprocessing to be included, and a period that closed two years ago uploads exactly like the current one.

The demand side is where the gap usually is. Paid search exports easily. What does not is everything else: deal registrations from agents and resellers, inquiries taken by a switchboard and typed into a spreadsheet, responses to an outbound campaign, registrations from a vendor-run webinar. Each of those is a list of people with contact details, each goes into the lead file tagged with its source, and each ranks on billed revenue alongside the campaigns that were easy to export. The sources hardest to export can be the ones carrying the most revenue, which is why the file is worth assembling properly rather than quickly.

Nothing goes near the network, the billing platform or the ordering system. There is no integration, no credential and no standing access for anyone's security review to work through, which for a business whose own customers audit their suppliers is worth stating plainly rather than leaving implied. Two files go up, a report comes back, and the data sits inside one workspace that can be emptied on request without a support ticket. For most providers the whole exercise is a spreadsheet job, done once and then repeated monthly in a few minutes.