Not legal advice: This article covers general compliance awareness and marketing practice, not legal advice. Israeli communications and consumer-protection law changes over time and applies differently case by case — verify your specific obligations with a qualified lawyer before you rely on anything here.
An SMS lands on a lock screen. It doesn't sit in a Promotions tab, and it doesn't need an open WhatsApp session to arrive — which is exactly why it's usually the channel a business reaches for last, as the message that has to get through when the others haven't. That also makes it the channel where getting the details wrong is the most visible: the wrong sender ID reads like a scam, a Hebrew message that silently splits into three billed segments looks like a billing mistake, and a marketing text sent without consent is judged under a specific Israeli law with its own, frequently misquoted, rules.
We checked three things that usually get discussed separately, if they get discussed accurately at all: which Israeli law actually governs a marketing SMS — and where a second, newer law leaves a gap nobody's fully closed — the technical reason a Hebrew text can cost more than an English one of the same length, and the real, sourced research on whether an SMS's wording changes anything, next to the vaguer claims that circulate about SMS engagement in general.
Key takeaways
- A marketing SMS in Israel is governed by the same law as marketing email — Section 30A of the Communications Law, added by Amendment 40 (2008) — which requires opt-in consent, clear sender identification and a working opt-out, with a statutory damages ceiling of ₪1,000 per unlawful message. A separate, newer law — the 2020 "Do Not Call Me" registry — was written for phone calls, and its application to text messages is, in one Israeli law firm's own words, "a matter for future interpretation by courts."
- Since January 3, 2021, sending marketing SMS from a real phone number has required a one-time Ministry of Communications check to prove you own that number. An alphanumeric sender name — your business name instead of a number — is explicitly exempt from that check, a gap the regulation's own drafters acknowledged rather than closed.
- Hebrew text can't use the 160-character GSM-7 encoding an English SMS uses. Any Hebrew character forces UCS-2 encoding, which caps a single message at 70 characters and each segment of a longer, split message at 67 — so a 140-character message costs 1 billed segment in English and 3 in Hebrew, without a single word changing.
- "SMS has a 98% open rate" is repeated on dozens of marketing blogs and traced reliably to none of them: standard SMS has no native read-receipt mechanism at all, so no platform can actually measure whether a text was read the way an email pixel measures an open.
- Klaviyo's own disclosed 2026 benchmark data, drawn from over 183,000 of its customers, puts a "good" SMS campaign click-through rate at roughly 8.9%–14.5%, with automated flow messages averaging near 10% — a real, dated, sourced figure, considerably less dramatic than the 98% folklore and the one worth planning around.
- The strongest controlled evidence we found wasn't about open rates at all: two randomized trials at a London NHS hospital trust, covering just under 20,000 patients between them, found a specific-cost SMS reminder cut missed appointments from 11.1% to 8.4% — and in the follow-up trial, every attempt to add more to that winning message made it perform worse, not better.
What law actually governs a marketing SMS in Israel — and is it the same one that covers phone calls?
Yes to the first part, and no to the second. A marketing SMS in Israel falls under Section 30A of the Communications Law (Bezeq and Broadcasting), 1982 — the statute added by Amendment 40, passed May 27, 2008 and in force since December 1, 2008, and covered in more depth in our email compliance guide. SMS is one of the channels the law names explicitly, alongside email, fax and automated dialing systems: no prior opt-in consent, no advertising message, regardless of channel. The same statutory damages ceiling applies — up to ₪1,000 per unlawfully sent message, awardable by a court without the recipient having to prove real harm, though it's a ceiling a judge may reduce, not a fixed fee owed automatically.
Where it gets less settled is a second, newer law that sounds like it should cover the same ground but was written for something else. Amendment 61 to the Consumer Protection Law, enacted November 24, 2020, created Israel's "Do Not Call Me" registry (מרשם "אל תתקשרו אליי"), run by the Consumer Protection and Fair Trade Authority, with its regulations approved in mid-2022 and enforcement live since January 1, 2023. It bars telemarketing contact — including calls placed over VoIP, WhatsApp or Skype — to any phone number listed in the registry. What it doesn't clearly do is name text messages one way or the other. One Israeli law firm's own analysis of the amendment put it plainly: because the law never explicitly limits itself to voice calls, marketing text messages sent through phone-number-based channels may also be covered — and that's "likely a matter for future interpretation by courts," not a settled question today.
| Section 30A ("Spam Law", Amendment 40) | Do Not Call Me registry (Amendment 61) | |
|---|---|---|
| What it regulates | Sending a specific advertising message: email, SMS, fax, autodialer | Initiating telemarketing contact to a listed phone number |
| Covers SMS? | Yes, named explicitly in the statute | Unclear — written for calls, application to text is undecided by Israeli courts |
| Consent model | Opt-in required before the first message | Opt-out — a business may call unless the number is registered |
| Penalty | Up to ₪1,000 per message, court-awarded, no proof of harm needed | ₪53,070 per violation for a corporation, ₪29,490 otherwise (up to 1.5x for violations affecting many consumers) — the 2026-indexed figure confirmed in the law's own consolidated text; an older, since-superseded indexed value (₪46,080) still circulates in some compliance guides |
| In force since | December 1, 2008 | January 1, 2023 |
The practical takeaway isn't to wait for a court to settle the ambiguity — it's to not need the answer. Section 30A's opt-in requirement for SMS is unambiguous and has applied since 2008; a business that already has real consent for every marketing text it sends is compliant regardless of how the Do Not Call registry's edge case eventually gets resolved.
What actually has to be in a marketing SMS, given that it has no subject line?
Section 30A's disclosure rules were written with a subject line in mind — the "פרסומת" (advertisement) label our email compliance guide covers has to open the subject of a marketing email. SMS doesn't have one, so the same underlying requirement lands inside the message body instead: Israeli guidance for SMS senders — echoed by every major local SMS gateway's own compliance page — describes the same three elements applied to a subject-line-free format: the message has to clearly identify the sending business, it can't be misleading about what it is, and it has to offer a real, working way to opt out, sent only to someone who gave prior consent in the first place.
None of that is optional styling. Leaving out the business name or the opt-out line is itself a violation under the same statute — independent of whether the recipient had opted in to begin with.
- Name the business sending the message, not just a generic short code or unfamiliar number.
- Include a real, working opt-out — Israeli SMS platforms standardly implement this as a short link rather than a keyword reply, since a link works identically however the message is read.
- Don't disguise a promotion as something else — the same rule the WhatsApp template categories enforce automatically, Section 30A enforces by statute for SMS.
- Send only to a number that gave affirmative, provable consent — the burden of proof sits with the sender, not the recipient, if it's ever challenged.
Do you need to verify your sender number before sending — and does a business name skip that?
Yes, and yes — but not for the reason most senders assume. Separately from Section 30A's consent rules, Israel's Ministry of Communications introduced a sender-identification regulation that took effect January 3, 2021, aimed specifically at SMS number spoofing: the practice of forging a sender ID so a text appears to come from a contact already in the recipient's phone, a bank, or a trusted brand, used to run phishing and malware campaigns. If your SMS sender ID is a real mobile or landline number, the rule requires a one-time verification before your first send — an SMS with a confirmation link for a mobile number, or a callback to the line itself for a landline — to prove you actually own that number.
Alphanumeric sender IDs — a business name shown instead of a number — and calls or texts sent from a withheld number are explicitly exempt from this verification. That's not an oversight the coverage of the rule missed: officials who worked on the regulation said as much themselves when it was announced, acknowledging in public comments reported at the time that the decision "does not address SMS messages where the [sender ID] has text" — text-based impersonation was, and remains, outside what this specific rule closes. One outlet's own headline on the announcement called it "better than nothing, but nowhere near sufficient."
| Sender ID type | One-time verification required? | What it protects against |
|---|---|---|
| A real mobile number you own | Yes — confirm via a code sent to that number | Someone else spoofing your number as a sender ID |
| A real landline number you own | Yes — confirm via a callback to that line | Someone else spoofing your number as a sender ID |
| An alphanumeric sender name (your business name) | No — explicitly exempt | Nothing — regulators flagged this as an open gap, not a closed one |
| A withheld / hidden number | No — explicitly exempt | Nothing — same acknowledged gap |
oToK's SMS feature reflects the same two real paths the regulation describes: verify a mobile number you own with a one-time code, or register an alphanumeric sender name shown in its place — set as a workspace default and overridable per campaign. Given the acknowledged gap above, a verified real number is the stronger signal to a recipient who's learned to be wary of unfamiliar text senders; an alphanumeric name is convenient, not verified.
Why does a Hebrew SMS cost more segments than an English one of the same length?
This is a fixed technical constraint of the global SMS standard, not an Israel-specific rule — but it lands harder on an Israeli, Hebrew-first business than almost anywhere else. A standard SMS is encoded one of two ways. GSM-7 packs each character into 7 bits and fits 160 characters into a single message — but its character set is built for Latin-based alphabets: it has no Hebrew letters in it at all. The moment a message contains a Hebrew character (or, just as easily, an emoji), the whole message switches to UCS-2 encoding, which needs 16 bits per character and caps a single message at 70 characters instead of 160.
The gap widens further once a message needs to split across multiple segments. Each concatenated segment reserves a few bytes for a header that tells the receiving phone how to reassemble the pieces in order — 7 characters' worth for GSM-7, leaving 153 usable characters per segment, and more, proportionally, for UCS-2, leaving just 67 usable characters per segment. Put a number on it: a message of exactly 140 characters fits in a single billed segment in English. The identical-length message in Hebrew needs three — 67 characters into the first two segments still leaves 6 characters unaccounted for, which is what actually forces the third.
Single message (no split)
Each segment, once the message splits
The opt-out line eats into the same budget. oToK's SMS composer counts segments live as you type, opt-out link included, specifically so what's on screen is what the send actually costs — and its default opt-out link is kept deliberately short so a Hebrew message has a real chance of staying inside one segment.
Is the "98% SMS open rate" real — and what does the actual disclosed data show instead?
It's repeated constantly, almost always attributed to Gartner, and we couldn't locate a dated Gartner report anywhere that the number actually traces back to — despite checking dozens of the marketing pages that cite it. That's worth pausing on for a more basic reason than a missing citation: standard SMS, unlike email, has no native read-receipt mechanism at all. A delivery report confirms a message reached the handset; nothing in the global SMS protocol tells the sender whether the recipient actually looked at it, the way an email's tracking pixel fires when the message is opened. "Open rate" isn't a number any SMS platform can honestly measure the way it measures an email open — which means a precise-sounding 98% figure is, at best, describing something else entirely.
What is real, dated and disclosed is considerably less dramatic. Klaviyo's own 2026 SMS marketing benchmark page, built from data across more than 183,000 of its customers, puts a "good" SMS campaign click-through rate in the 8.9%–14.5% range, with automated flow messages (triggered by a specific action, like a cart or a browse) averaging close to 10% — roughly double a typical campaign send — and its own reported top decile for one of its tracked categories, apparel and accessories, clearing 16.65%. Separately, a 45% SMS response rate gets cited from a platform's self-reported analysis of 100,000 texts — but the figure circulates with inconsistent attribution across the sites that repeat it, tracing to different named sources depending on which page you land on, which is the same pattern worth treating with real caution the moment a precise number can't settle on where it actually came from.
Click rate isn't a perfect substitute for a true open rate, but it's the honest one: it only counts something an SMS platform can actually observe — a link being tapped — rather than inferring a read that the protocol has no way to confirm.
Does the wording of an SMS reminder actually change whether someone shows up?
Here the evidence is unusually strong, and it comes from healthcare rather than marketing. Hallsworth, Berry, Sanders, Sallis, King, Vlaev and Darzi published "Stating Appointment Costs in SMS Reminders Reduces Missed Hospital Appointments: Findings from Two Randomised Controlled Trials" in PLOS ONE in September 2015, run across outpatient appointments at Barts Health NHS Trust in London. The first trial, covering 10,111 participants between November 2013 and January 2014, tested the hospital's existing SMS reminder against a version that stated the specific cost of a missed appointment to the health system. The existing message produced an 11.1% did-not-attend (DNA) rate; the specific-cost version produced 8.4% — a statistically significant drop (odds ratio 0.74, 95% CI 0.61–0.89, p<0.01).
The second trial, covering a further 9,848 participants between March and May 2014, is the more interesting result. Rather than stopping there, the researchers made the winning specific-cost message the new baseline (8.2% DNA) and tested three ways of building on it: adding a general (rather than specific) cost figure, adding empathetic language, and adding an option to hear a recorded version of the message. Every single addition performed worse than the plain specific-cost baseline it was built from — general costs rose to 9.9% (p=0.046), empathy rose to 10.7% (p<0.01), and the recording option rose to 9.6% (not statistically significant, p=0.07) — none of them an improvement, all in the direction of more missed appointments, not fewer.
The lesson generalizes past hospital appointments: once a short, specific message is measurably working, the instinct to make it warmer, longer or more detailed isn't a safe default — in this study, every version of "more" measurably underperformed the plain version it was added to. That's a convenient rule for a channel that also bills you by the character.
So when should a business reach for SMS instead of WhatsApp or email — and what should it set up first?
SMS's real edge isn't engagement — it's reach that doesn't depend on an app, a session, or a connection. It doesn't need WhatsApp's 24-hour window to be open, it isn't sorted into an inbox tab, and it works on a phone with no data connection and no app installed, which is why it fits best as the channel behind the other two rather than the first one reached for. That's the same logic oToK's own SMS feature is built around: the same audiences already built for email and WhatsApp campaigns, a Send SMS step inside the automation builder with the same variables and conditions as any other step, and — specifically — a fallback that fires when a WhatsApp message can't be delivered, including once the 24-hour session window has closed and only a fallback channel can still reach that contact.
Consent is also unified rather than per-channel: opting out of SMS closes WhatsApp for that contact too, on the reasoning that someone asking a business to stop means both, not just the channel the opt-out link happened to arrive on. Booking reminders are treated as the one deliberate transactional exception, and billing runs on prepaid per-segment credits — refunded automatically if a message doesn't go out — or through a connected Extra Mobile account for a business that already has one.
- 1Build one opt-in flow that names the business and satisfies Section 30A regardless of what the Do Not Call registry's SMS ambiguity ever resolves to — that's the law with no gray area, so it's the one worth actually relying on.
- 2Verify your sender number the day you set it up, not the day a send silently fails — a real mobile or landline sender ID needs the Ministry's one-time proof-of-ownership check before it can send.
- 3Choose deliberately between a verified real number and an alphanumeric business name, and treat the fact that alphanumeric skips verification as a reason to be more careful about who inside the business can set one, not less.
- 4Draft the Hebrew version of any SMS first, not last, and check its segment count before sending — a message that fits comfortably in English can silently jump from one billed segment to three once it's written in Hebrew.
- 5Once a message is measurably working, resist adding to it — the reminder-wording research above found every attempt to build on a winning message made it perform worse, not better.
- 6Reserve SMS for the moment something else won't reach the person — a closed WhatsApp session, a card about to expire, a booking about to be missed — rather than making it the default first channel for everything.
None of this treats SMS as a replacement for WhatsApp or email — it's the narrowest, most reliable of the three, which is exactly what makes it the wrong channel to run everything through and the right one to have wired in as a fallback.
Frequently asked questions
- Does Israel's Spam Law (Section 30A) apply to SMS the same way it applies to email?
- Yes — SMS is named explicitly in the statute alongside email, fax and autodialers. It requires prior opt-in consent, clear identification of the sending business, and a working opt-out, with a statutory damages ceiling of ₪1,000 per unlawful message. Because SMS has no subject line, the disclosure requirements that apply to an email's subject line land inside the SMS body instead.
- Does Israel's "Do Not Call Me" registry cover marketing SMS?
- It's genuinely unclear. Amendment 61 (2020) was written for telemarketing calls — including calls over VoIP, WhatsApp or Skype — and doesn't explicitly name text messages either way. One Israeli law firm's own analysis calls its application to SMS "a matter for future interpretation by courts," not a settled question. The safer position is to treat Section 30A's clear opt-in requirement as the one that applies regardless.
- Do I need to verify my SMS sender number in Israel?
- Yes, if you're sending from a real mobile or landline number — a Ministry of Communications rule in force since January 3, 2021 requires a one-time verification to prove you own that number, aimed at stopping SMS spoofing. An alphanumeric sender name (your business name shown instead of a number) or a withheld number is explicitly exempt from that check — a gap officials themselves acknowledged when the rule was announced.
- Why does my Hebrew SMS split into more segments than I expected?
- Hebrew characters force UCS-2 encoding, which caps a single message at 70 characters (and 67 per segment once it splits into more than one), versus 160 characters (153 per segment) for GSM-7-encoded Latin text. A 140-character message needs 1 billed segment in English and 3 in Hebrew — the same rule applies to any emoji in an otherwise-English message.
- Is it true that SMS has a 98% open rate?
- That figure is widely repeated, usually attributed to Gartner, and we couldn't trace it to any dated report. It's also close to unmeasurable: standard SMS has no native read-receipt mechanism, only delivery confirmation, so no platform can actually observe whether a text was read. Klaviyo's own disclosed 2026 data (183,000+ customers) is the more honest number: a "good" campaign click-through rate of roughly 8.9%–14.5%.
