Sign in

Blog · Forecast and pipeline · Telecoms and connectivity

Sites near their capacity: the upgrade signal in usage data before the customer complains

How a business telecoms provider reads its usage data per site against each site's contracted capacity: peak utilisation against the circuit's bandwidth, the sites above a stated share of capacity for a stated number of days, the customers with several such sites, the upgrade value at the next tier, the sites whose usage fell to nothing and may be closing, and why the upgrade list beats the fault ticket as the moment to sell more.

The short answerEvery connectivity circuit has a contracted bandwidth and a usage record. Peak utilisation per site per day against that bandwidth, from the usage export, finds the sites running above a stated share of capacity on most days, which are the sites whose users are about to complain and whose customers are about to be quoted by a competitor. The upgrade list is those sites, ranked by the value at the next tier, per customer and per account manager. The same data finds sites whose usage fell to nothing, which are closing, and that is the estate reconciliation's other half.

A telecoms provider learns a customer's site has outgrown its circuit when the fault ticket arrives, and learns the customer has been quoted elsewhere when the upgrade order does not. The usage export knew the site was at capacity a quarter before the ticket. This guide sets out utilisation against contracted capacity, the upgrade list, the value at the next tier, and the sites going quiet.

The measures

Per site, per day:

Peak utilisation = peak throughput ÷ contracted bandwidth

Per site, per month:

Days above threshold; trend against the prior quarter Near capacity if days above threshold ≥ a stated count Upgrade value = next tier's monthly price − current, annualised Quiet if usage ≈ 0 for a stated number of days with a live circuit

Per customer, per account manager: near-capacity sites and upgrade value; quiet sites.

The rows you need

  • Usage export: site, circuit, day, peak throughput.
  • Contract register: site, circuit, contracted bandwidth, product, price, next tier price.
  • Account assignments: customer, account manager.

Customer and site identifiers only.

The assertion

every live circuit has usage data for the period, or is listed as unmeasured

A live circuit with no usage data is a monitoring gap before it is a quiet site.

A worked list

Threshold: peak above 75 percent on 10 or more days a month.

Customer Site Product Contracted Days above 75% Trend Next tier Upgrade value/yr Account manager
2207 S-14 Ethernet 100 100 Mb 22 Up from 9 200 Mb $4,800 AM-04
2207 S-22 Ethernet 100 100 Mb 18 Up from 11 200 Mb $4,800 AM-04
4471 S-03 Broadband 80 80 Mb 26 Flat Ethernet 100 $7,200 AM-11
9034 S-41 Ethernet 1G 1 Gb 2 Fine

Customer 2207 has two sites that hit capacity most working days and the count doubled in a quarter. The account manager's call is this week, with the usage graph, before the fault ticket and before the competitor.

Quiet sites

Customer Site Days at zero Contract end Reading
1187 S-08 34 14 months Closed or migrated; contract conversation

Where it goes wrong

Utilisation averaged. A site at forty percent average is at ninety at three o'clock.

Single peak used. One backup job flags every site. Days above threshold, not one day.

Fault ticket as the trigger. A quarter late and a competitor ahead.

Quiet sites ignored. The customer cancels a circuit the provider could have repurposed.

Every month, capacity and the list

Mapped once, the usage export, the contract register and the assignments produce utilisation per site, the near-capacity list with upgrade value and the quiet list per account manager every month. Covirage builds this from the exports as they are. The telecoms page describes the setup, and the customer estate guide covers what the quiet sites do to the estate.

Questions people ask

What share of capacity is the threshold?

Where the customer's applications start to degrade, which for most business circuits is a peak sustained above seventy to eighty percent on a stated number of days a month. The threshold is stated per product type, and the report shows the days above it, not a single peak.

Why is this better than waiting for the fault ticket?

Because the fault ticket is the customer's experience of the problem, and by then they have had it for weeks and mentioned it to a competitor's rep. The usage data shows the site approaching capacity a quarter earlier, and an account manager who calls then is selling a solution rather than answering a complaint.

What about sites whose usage dropped?

A site at zero usage for a month with a live circuit is either closed or migrated. It is on the estate reconciliation list, and it is a conversation about the contract before the customer asks to cancel it.