Est. reading time: 5 minutes

Most brands run SMS as a megaphone, blasting the same discount to the whole list and measuring clicks. The channel’s actual advantage points the other way. People reply to texts, reflexively, in a way they never reply to email, and a reply is a micro-commitment that converts at rates a broadcast click never touches. Postscript is built for exactly this two-way pattern, and programs that adopt it in earnest reach reply rates in the 25 percent range that megaphone programs never see. The gap isn’t a copywriting trick. It’s a framework, and here’s the whole thing.

Build the program around intent, not the calendar

The foundation is segmentation by what people just did and what they’re likely to do next, not by when the marketing calendar says it’s time to send. That means first-party events from Shopify (browse, cart, purchase), subscriber tags (VIP, discount-sensitive, product interest), and lifecycle milestones (day 1, day 30, day 90) as the raw material, so every message can reference a real moment instead of a generic offer.

Then adopt the one rule that changes the channel’s character. Every send invites a reply. “Click to shop” becomes a question that routes the shopper to their next best step, “Need a size rec?”, “Want early access?”, “Ready for 20% off today?”, and the program stops shouting offers and starts collecting small yeses. Structurally, automations do the heavy lifting, welcome, browse and cart abandonment, post-purchase upsell, winback, back-in-stock, while campaigns become genuine events, drops, launches, VIP early access, limited inventory. And the considerate-sending layer isn’t optional garnish, since Smart Sending, quiet hours, and timezone delivery are what keep a high-frequency conversational program compliant and welcome rather than exhausting, a balance we covered in how to use SMS without flooding your customers’ phones.

Design flows that branch on replies

Triggers should map to shopper momentum. In Postscript, fire automations on new subscriber (keyword or checkout opt-in), product view without add-to-cart, cart started without checkout, checkout started without purchase, order delivered, replenishment windows, and winback intervals, pulling event detail like product name, category, price, and inventory into the message so personalization happens without anyone lifting a finger.

Then branch like a decision tree, with replies kept effortless to type, Y/N, 1/2/3, or a single keyword. The pattern in practice: “Want a fit rec? Reply 1 for Slim, 2 for Relaxed, 3 if not sure.” Options 1 and 2 return the size guide and product link, option 3 routes to a short quiz. For incentives, “Ready to claim 20% today? Reply YES for your code” sends a unique code and short link on YES, while a NO diverts to value messaging instead of burning margin on someone who wasn’t asking for a discount.

Two design details separate durable programs from fragile ones. Handle the unexpected, because people text like people, so conditional logic and keyword matching should return a helpful fallback (“Got it! Need a human? Reply HELP”) with helpdesk escalation wired in for the high-intent conversations, returns, fit questions, bulk orders, that a bot should never fumble. And tag every outcome, claimed code, size preference, category interest, so each reply makes the next flow smarter, which is how the list compounds in value rather than just in volume.

Copy that earns the reply

The working formula is hook, value, question. The hook is vivid and specific (“Your size just restocked.” “BFCM early access starts now.”). The value is one clear benefit (“Fit recs in 20 seconds.” “Lock today’s price before it sells out.”). The question is a tap-friendly prompt that signals intent (“Reply SIZE for recs.” “Text DROP for your link.”). Beyond the formula, write like a human and format like a professional, short lines, payoff first, one emoji at most and only on-brand, merge fields for the name and last item viewed, and the exact moment referenced (“Still eyeing the Heather Grey Hoodie?”). The action should be unmistakable, and “Reply YES” beats “Learn more” every time it’s tested.

Prompts worth adapting directly:

• “Want 24-hr early access to tomorrow’s drop? Reply DROP.”
• “Need a gift under $50? Reply 1 for Her, 2 for Him, 3 for Anyone.”
• “Ready for 15% today? Reply YES for your code.”
• “Unsure on size? Reply SIZE for a 30-sec fit quiz.”

They work for the same three reasons. The ask is binary or nearly so, the reply takes two seconds, and the payoff is immediate. When a program’s replies stall despite this structure, the causes are usually diagnosable, and we worked through them in why your Postscript campaigns aren’t getting replies.

Measure in conversations, then codify

For conversational sends, reply rate is the primary metric, with the downstream tracked by branch, click-through, conversion rate, revenue per recipient and per conversation, AOV, and unsubscribes, so you know not just that people replied but which paths made money. Match attribution windows to intent, roughly 72 hours for abandonment flows and 7 days for winback, and run holdout groups on major campaigns, because SMS attribution flatters itself as readily as every other channel’s.

Iterate on three levers in order. Question type first, testing Y/N against numbered options against keywords, with binary and single-keyword formats usually leading. Timing second, within 30 to 60 minutes of the trigger for abandonment, mornings for replenishment, afternoons for drops. Incentive last, and reluctantly, leading with value like fit help and early access, escalating to discounts only after a follow-up goes unanswered, so the program rewards engagement rather than training patience.

Then make the wins permanent. High-performing branches saved as templates, tags standardized for interests and outcomes, segments that update themselves from replies (“Size: M,” “Category: Skincare,” “VIP: Early Access”), A/B splits at the step level, and winning campaign patterns rolled into the always-on flows. That codification step is the difference between a program that had a good quarter and one whose baseline permanently moved. Trigger on intent, branch on replies, ask questions worth answering, and prove the lift with holdouts, and SMS stops being the channel you’re careful not to overuse and becomes the one your customers actually talk back to.