Sign in

Blog · Board and management reporting · Telecoms and connectivity

Order-to-activation time per customer: the provisioning delays that precede churn

How a business telecoms provider measures the days from a service order to activation per customer and per site from the order and provisioning exports, the norm per service type from its own history, the customers whose recent orders took twice the norm, the correlation with contract non-renewal from the provider's own outcomes, and the list that reaches the account manager while the customer is still installing.

The short answerOrder-to-activation is the days from a signed service order to the service being live, per order, from the order and provisioning exports. The norm is the provider's own median per service type. Customers whose recent orders ran well past the norm are listed per account manager, with the orders and the stage where the delay sits. From the provider's own history, customers with a long-delayed activation in the year before renewal non-renew at a higher rate, and the report states that rate beside the list, as an association.

A business telecoms customer orders three new sites and the third takes eleven weeks to go live. The account manager finds out at renewal, when the customer tells them. The order and provisioning exports knew in week four. This guide sets out order-to-activation per customer, the norm per service type, the delayed list, and the association with non-renewal.

The measures

Per order:

Days to activation = live date − order date Delay ratio = days to activation ÷ norm for the service type Stage split = days in each provisioning stage

Per customer:

Recent orders over a stated delay ratio, and their stage split

Per service type:

Norm = median days to activation, trailing year, from completed orders

The rows you need

  • Orders: order, customer, site, service type, order date, account manager.
  • Provisioning: order, stage, stage entered date.
  • Contract outcomes: customer, renewal date, renewed or not.

Customer identifiers only.

The assertion

every live order has stage dates in sequence, and a live date after its order date

Orders with stages out of sequence, or live with no order date, fail it and are excluded from the norm with a count.

A worked list

Customer Order Service Days Norm Ratio Delay in stage Account manager
2207 O-8821 Leased line 77 35 2.2 Wholesale carrier: 51 days AM-04
2207 O-8834 Leased line 62 35 1.8 Wholesale carrier: 40 days AM-04
4471 O-8902 Broadband 31 12 2.6 Customer site access: 19 days AM-11
9034 O-8917 SIP 9 8 1.1

Customer 2207 has two leased lines running twice the norm, both stuck at the wholesale carrier. The account manager can tell the customer that today, with a date, and escalate the carrier. Customer 4471's delay is at their own site, which is a different conversation.

The association

Renewing customers, last year Non-renewal rate
Had an order over 2× norm in the prior year 23%
Did not 8%

Stated as an association. It is why the list is worked.

Where it goes wrong

Provisioning measured as an average. Thirty-five days on average, and one customer at seventy-seven.

Stage split missing. Every delay is the provider's fault, or nobody's.

List after activation. The customer already has the story.

Association read as cause. It is a signal for a call, not a proof.

Every week, delayed orders per account manager

Mapped once, the orders, provisioning and outcomes produce the norms, the delayed list with its stage split and the association every week. Covirage builds this from the exports as they are. The telecoms page describes the setup, and the renewal calendar guide covers the renewal the delayed install will otherwise surface at.

Questions people ask

Where does the delay sit?

Provisioning exports carry stage timestamps: order accepted, survey, circuit ordered from the wholesale carrier, installed, tested, live. The split by stage per delayed order says whether the delay is the provider's, the wholesale carrier's or the customer's site, and each is a different conversation.

How is the churn association measured?

Customers renewing in a period, split by whether they had an order over a stated multiple of the norm in the prior year, and their non-renewal rate in each group. From the provider's own contract outcomes. It is an association from observational data, stated as one, and it is usually large enough to change behaviour.

What does the account manager do with it?

Calls the customer before the customer calls them. A delayed install that the account manager knows about and explains is an irritation; one the customer discovers alone is the story they tell at renewal.