How to Set Up SMS Fallback When Live Chat Isn't Available

Sending a text message requires explicit consent for the specific purpose, and treating a chat conversation as implied permission is not sufficient.

Explicit rather than implied

The visitor should actively agree to be contacted by text, with the purpose stated, rather than having a number field appear and be filled in.

This is both a compliance requirement in most jurisdictions and the behaviour that avoids complaints from people who did not expect a message.

Stating the purpose narrowly

Consent to continue a support conversation is not consent to marketing, and using it as such is the fastest way to generate complaints and regulatory attention.

Keeping the purpose narrow also improves opt-in rates, since visitors are considerably more willing when the scope is clear.

Opt-out and record keeping

Include opt-out instructions in the first message and honour them immediately, since this is both required and the behaviour customers expect.

Recording when and how consent was given matters if it is ever questioned, and it is trivial to capture at the point of agreement.

The point of the fallback is continuity, so the customer should never have to explain anything twice because the channel changed.

Carrying history across

Everything discussed in chat must be attached to the SMS conversation, so an agent replying sees the full exchange rather than an isolated text message.

Without this the fallback moves the conversation and loses its content, which produces exactly the repetition it was meant to prevent.

Threading replies back

A reply arriving by SMS should attach to the existing conversation rather than creating a new one, which requires matching on the number and the recent context.

Failing to thread produces duplicate records and the possibility of two agents responding to the same customer independently.

Handling the return to chat

Customers who move to SMS sometimes return to the website later, and recognising them means the conversation continues rather than restarting.

This requires identity matching across channels, which is the same mechanism message threading depends on generally.

SMS has different constraints from chat, and responses written for a chat window read badly as text messages.

Length and structure

Long responses split across multiple messages, arrive out of order occasionally and are considerably harder to read than the same content in a chat window.

Keeping SMS replies short and putting detail behind a link works better than compressing a full answer into a text.

No rich content

Quick reply buttons, cards and formatting do not exist in SMS, which means flows relying on them need a text equivalent rather than degrading badly.

Numbered options are the usual substitute and work acceptably, provided the list is short.

Response time expectations

SMS carries an expectation of a reasonably prompt reply but not an immediate one, which is exactly why it suits the out of hours case.

Stating when a reply will arrive in the first message sets that expectation correctly rather than leaving it to assumption.

SMS costs per message and the costs accumulate quickly, which affects where the fallback should and should not be offered.

Per-message pricing

Every message in both directions carries a cost, which makes a long SMS conversation considerably more expensive than the equivalent chat exchange.

This is an argument for using SMS to re-establish contact and moving back to a richer channel where possible, rather than conducting the whole conversation by text.

International variation

Costs and deliverability vary enormously by country, and a fallback that works well domestically may be expensive or unreliable elsewhere.

Restricting the offer to countries where it works properly is better than offering it universally and delivering inconsistently.

Comparing with email

Email costs nothing per message and is read considerably less reliably, which makes the choice between them a trade-off rather than an obvious answer.

Offering both and letting the customer choose is frequently the right approach, particularly where the enquiry is not urgent.

SMS delivery is not guaranteed, and a fallback that silently fails is worse than not offering one.

Validating the number

Checking the number format and country code at capture prevents the most common failure, which is a message sent to an address that cannot receive it.

Confirming the number back to the visitor before sending catches typos while they are still on the page to correct them.

Monitoring delivery

Delivery receipts should be monitored and a failure should trigger an alternative route rather than the conversation simply ending.

A customer who agreed to SMS and received nothing has been failed more visibly than one who was never offered it.

Offering a second route

Capturing an email alongside the number, or offering it as an alternative on failure, provides a fallback for the fallback at essentially no cost.

This matters most out of hours, when nobody is monitoring for failures until the following morning.

Chatboq offers SMS continuation at defined trigger points, captures consent explicitly and threads replies into the original conversation.

Trigger points configured

The offer appears outside coverage hours, during long waits and on inactivity, rather than being a permanent alternative to chat.

Restricting it to these moments keeps SMS costs proportionate and avoids collecting numbers from visitors who did not need the fallback.

Consent and threading handled

Explicit opt-in with a stated purpose is captured and recorded, and opt-out instructions are included in the first message automatically.

Replies thread into the existing conversation using number matching, so agents see one continuous exchange with full history attached.

Measuring the fallback

Track offer acceptance rate, delivery success, reply rate and whether conversations continued by SMS reached resolution.

Comparing resolution for SMS-continued conversations against those abandoned at the same trigger point shows what the fallback actually recovered.

Deciding who replies

SMS conversations need an owner, since a reply arriving into an unmonitored queue is worse than the abandonment the fallback was meant to prevent.

Routing SMS replies into the same queues as chat, rather than a separate inbox, is what prevents them being overlooked.