Replay Live AI-Powered Calling and Texting in HubSpot with Trent Chapman + Brett Spusta. Watch now

Why One Emoji Can Triple Your SMS Cost (Segments Explained)

Admin 6 min read

Carriers do not bill a text message by the message. They bill it by the segment. A plain-text message fits 160 characters in one segment. Add a single emoji and the whole message switches to a different character set that fits only 70 characters per segment. That is how a 150-character text that cost one segment can cost three after one smiley face, with not one extra word.

Below is how segments work, which characters flip a message, what that does to a broadcast, and how to see the count before you send.

What is an SMS segment?

A segment is the unit a text message travels and bills in. The SMS standard limits each one to a fixed number of bytes, so a message longer than one segment is split into several. The phone stitches them back together, and your customer sees one message. Your bill sees several.

How many characters fit in a segment depends on the encoding, which is the character set the message is sent in. There are two:

  1. GSM-7, the standard SMS alphabet. It covers English letters, numbers, common punctuation and a short list of accented letters.
  2. UCS-2, which covers almost every character in every language, emoji included. It takes more space per character, so fewer fit.
EncodingFits in one segmentPer segment once the message splitsUsed when
GSM-7160 characters153 charactersEvery character is in the GSM-7 alphabet
UCS-270 characters67 charactersAny one character is outside it

The per-segment number drops once a message splits because each part carries a small header that tells the phone how to reassemble it.

Why does one emoji change the cost of the whole message?

Because encoding is all or nothing. A message is sent in GSM-7 only if every single character is in the GSM-7 alphabet. One character outside it, anywhere in the message, sends the entire message as UCS-2.

Most emoji also count as two characters in UCS-2, not one. So one emoji does two things at once: it shrinks every segment from 160 characters to 70 (or 67), and it adds a character on top.

Here is what that looks like on real message lengths:

MessagePlain textSame text plus one emoji
60 characters1 segment1 segment
120 characters1 segment2 segments
150 characters1 segment3 segments
300 characters2 segments5 segments

A short message survives an emoji. A message near the 160-character line does not. That is exactly the length most marketing texts land at, because people write until they hit the limit.

Which characters switch a text to UCS-2?

Emoji are the obvious one. The others are easy to miss, because they look almost the same as characters that are safe:

  1. Curly quotes and apostrophes (“ ” ‘ ’), which word processors and phones insert automatically in place of straight ones (" ').
  2. The long dash (—) and the single-character ellipsis (…), both common in copy written in a document first.
  3. Some accented letters. GSM-7 includes é, è, à, ü and ñ, but not á, í, ó, ú or ê. A first name like José stays GSM-7. María does not.
  4. Hidden characters pasted along with copy, such as non-breaking spaces.

A few characters are in GSM-7 but cost two places each: the euro sign (€), square and curly brackets, the caret (^), the tilde (~), the backslash and the pipe (|). They do not switch the encoding, but they use up the 160 faster.

Some sending platforms quietly swap common look-alikes, such as curly quotes, back to their plain versions before a message leaves. Emoji have no plain version to swap to, so they always switch the message.

How do merge fields change the segment count?

A merge field fills in a different value for every recipient, so the same template can be one segment for one contact and two for the next. A long company name can push a message past 160 characters. An accented first name can switch one contact's text to UCS-2 while everyone else's stays plain.

That means the count you see while writing is an estimate for a broadcast, not a promise. Leave some room under the limit when a template carries merge fields, and test it with your longest real values.

What does this cost on a broadcast?

Multiply it out. A 150-character promotion sent to 2,000 contacts is 2,000 segments. The same promotion with one emoji is 6,000 segments. Nothing about the offer changed, and nobody reading it would notice the difference in the text itself.

On YetiConnect, automated and broadcast texts draw from marketing credits, and one credit equals one SMS segment (see pricing). So for our customers, the segment count is the credit count. The same math applies on any platform that bills by segment, which is most of them, because it mirrors how the carriers handle the message.

How do you keep a text to one segment?

  1. Write it where you send it. Typing the message in the texting tool avoids the curly quotes and long dashes a document adds.
  2. Paste as plain text when you do draft elsewhere, and replace any curly quotes with straight ones.
  3. Decide on emoji deliberately. If the message is under about 65 characters, an emoji costs nothing extra. Near 160, it can triple the cost.
  4. Leave room for merge fields. Aim for about 140 characters in a template that fills in names or company names.
  5. Look at the count before you send. Any tool that bills by segment should show you the segment count while you write.

How do you see the segment count before you send in YetiConnect?

The YetiConnect broadcast editor counts as you type. Under the message it shows the character count, the per-segment limit and the number of segments, for example "152 / 67 chars · 3 segments". The line turns amber once the message runs past one segment, and a UCS-2 tag appears the moment a character switches the encoding, so you can find the emoji or curly quote and decide whether it is worth it. For broadcasts on a credit plan, the editor also estimates the credits the whole send will use.

Texts and replies log to the HubSpot contact record, so the conversation lives where your team already works. If you are setting up broadcasts and automated texts in HubSpot, our HubSpot SMS Playbook has templates and workflow recipes written to stay short.

Want a second set of eyes on what your texts cost today? Book 30 minutes with our team.

Frequently Asked Questions

A plain-text message in the GSM-7 alphabet fits 160 characters in one segment. Once it runs longer, it splits into segments of 153 characters each. If any character is outside GSM-7, such as an emoji, the message is sent as UCS-2 and fits 70 characters in one segment, or 67 per segment once it splits.
Yes, on any platform that bills by segment. One emoji switches the whole message to UCS-2, which fits fewer characters per segment, and most emoji count as two characters. A 150-character message goes from one segment to three.
Yes. Curly quotes, the long dash and the single-character ellipsis are outside the GSM-7 alphabet and switch a message to UCS-2. Some sending platforms swap curly quotes back to straight ones before sending, but it is safer to type straight quotes.
No. Modern phones reassemble the segments and show one message. The split only shows up in how the message is billed.
Automated and broadcast texts draw from marketing credits, and one credit equals one SMS segment. The broadcast editor shows the segment count as you type and estimates the credits for the whole send.

See YetiConnect in Action

Book a personalized demo and see how YetiConnect can help your team close more deals.

Book a Demo

Related Articles

SMS Strategy

5 Texts That Fill a Healthcare Practice's Schedule

Healthcare and wellness practices lose real revenue to patients who inquire, get a plan, and never book. Here are 5 simple texts, sent inside HubSpot, that turn stalled consultations into booked appointments, with the recovery math behind each one.

Admin 6 min read
SMS Strategy

5 Texts That Close More Home Services Estimates

Home services teams lose real revenue to estimates that go quiet. Here are 5 simple texts, sent inside HubSpot, that restart the conversation and recover unsigned quotes, with the recovery math behind each one.

Admin 7 min read