SMS campaign budgets: count encoded segments before approving the copy

A recipient can see one text message while the sender is charged for several underlying SMS segments. Estimate the campaign using the final encoded text and the provider's applicable segment rules. Multiplying the number of recipients by a headline segment rate can materially understate spending when the approved copy crosses a size or encoding boundary.

Twilio documents different SMS capacities for GSM-7 and UCS-2 encoding, with smaller per-segment capacities for concatenated messages, and states that its SMS billing is per segment. It also documents route-specific exceptions, including certain toll-free traffic. Use the actual channel and sender configuration for the campaign rather than treating one character limit as universal.

The budget should cover the text actually sent, including expanded names, links and any required template footer. A plain draft in a document may differ from the content produced by the sending system. Preserve an approved rendered example and the provider's segment preview alongside the cost estimate.

Assume a standard SMS route uses the documented 160-unit single-segment GSM-7 capacity and a 67-unit capacity for multipart UCS-2 messages. A hypothetical message containing 145 basic GSM-7 characters fits in one segment. Assume replacing one character with a non-GSM character makes the final text use UCS-2 and the provider's preview counts 145 encoded units.

That second version exceeds the seventy-unit single-segment UCS-2 capacity. Under the assumed multipart limit, 145 units require three segments because two segments hold only 134 units. The visible message is almost the same length, yet the counted SMS units have tripled. Verify the exact final string with the provider's current calculator or equivalent preview.

For 50,000 recipients, one charged attempt per recipient and an invented rate of $0.008 per segment, the one-segment version costs $400. The three-segment version costs $1,200 before any other applicable charges. The $800 difference comes from segmentation under these assumptions; it is not a quoted Twilio price or a prediction for every sender and destination.

A template can fit in one segment for one recipient and require more for another. Names, reference numbers, product descriptions and substituted links can vary. Test the real allowed field lengths and relevant language versions rather than approving only the shortest sample used during design.

For budgeting, group recipients by supported segment count and applicable route. If the expected population includes different templates or languages, preserve those groups. An average segment count can summarize the result, but it should be calculated from the expected mix instead of assumed to be one because the campaign has one name.

Review automated character substitutions with the communications owner. A cheaper encoding is not useful if the revised text changes a person's name, obscures the message or removes information the approved template requires. The financial task is to make the cost visible while preserving the intended communication.

Confirm the relevant sender, destination, route and contractual rate for each population. Identify whether the quote includes carrier or other pass-through charges, fixed sender costs and any minimum commitments. Do not apply a domestic example rate to an international population without evidence that the same price applies.

Some costs may be per segment, some per message or event, and some fixed for a period. Put each on its actual unit. A flat monthly sender charge should not be multiplied by recipients, while a per-segment charge should not be multiplied only by campaign count.

Keep discounts and credits separate from the gross usage calculation. A temporary credit may reduce the current cash bill without reducing the underlying cost of a recurring campaign once it expires. The decision to repeat the send should use the terms expected to apply at that time.

A retry or a second authorized reminder can add usage even though it reaches the same person about the same business event. Identify which attempts the application can create and when they are permitted. Count those expected events separately from unique recipients so the budget reflects the real sending plan.

Do not infer chargeability solely from a delivery-status label. Review the provider's applicable billing rules and the usage records associated with the message. A technical submission, a segment charge and a delivered recipient experience are related but distinct events, and a refund may be a later adjustment rather than a different original count.

Keep implementation changes within the approved communication behavior. A cost investigation should not disable a necessary message or silently create extra reminders. If the application is generating unintended repeated attempts, identify and correct that confirmed cause, then preserve the evidence needed to review the associated bill.

The approval packet should identify the audience count, template version, personalization rules, expected segment distribution, charging rates and any known additional attempts. Include a realistic high-usage case where uncertain fields can push messages over a boundary. That makes the spending consequence of a last-minute copy change reviewable.

Recheck the segment preview after changes to punctuation, links, localization or footer text. A previous approval should not be treated as evidence for a different rendered message. Keep the content owner's approval connected to the version that the sending system will actually use.

For a recurring notification, retain a small set of representative boundary examples as an operational check. The useful examples are the variants that could change encoding or segment count, not hundreds of identical short messages. They help the team notice a material cost change without slowing routine communication work.

After sending, compare the actual charged segments and adjustments with the planned recipient and attempt populations. Explain differences through encoding, expanded text, destination mix or additional sends. A message-count report alone cannot reproduce a segment-based invoice.

Assess the campaign's business outcome using the appropriate delivered or response population, clearly labeled. Segment count is a spending measure, not a count of people reached or conversations started. Keeping those measures distinct allows the next campaign to improve its economics without mistaking fewer billed units for a better customer result.