What the MCP can answer
The Postscript MCP lets you ask your AI assistant (Claude, ChatGPT, or Codex) plain-English questions about your shop's SMS performance. It pulls the numbers directly from Postscript and answers in the chat. It answers these kinds of questions:
Ask about | What you get | Example |
Shop-wide trends | Attributed revenue, messages sent, or new subscribers as a time series by day, week, month, or year | "Show SMS revenue by week from Sep 1 to Sep 30." |
Campaigns, automations, or popups | Per-entity sends, clicks, CTR, conversion rate, orders, revenue, and unsubscribes. Automations are reported as whole flows. | "Which campaigns had the highest CTR from Sep 1 to Sep 30?" |
Rankings | Top or bottom performers by any measure | "Top 5 campaigns by revenue from Sep 1 to Sep 30." |
Exports | Tables of campaign and automation results you can copy into a spreadsheet | "Give me every campaign from Sep 1 to Sep 30 in one table: IDs, sends, clicks, orders, and revenue." |
How to get the best results
The best questions name a metric, a date range, and a scope. Get those three right and you will rarely need a follow-up.
Name the metric. Say "revenue," "sends," "CTR," "unsub rate," or "orders" rather than "how did we do." If you want several, list them: "sends, revenue, and CTR."
Use explicit dates. Say "June 1 to August 18" rather than "the last 90 days." Explicit dates return the most reliable results.
Say what to compare. "Sep 1–30 vs. Aug 1–31" gets you a side-by-side instead of one number.
Say shop-wide or by entity. "Total SMS revenue" pulls one shop-wide number. "Revenue by campaign" or "which automation drove the most" pulls a per-entity report. These use different revenue definitions (see "Know the numbers" below).
Use the names you know, loosely. Say "the welcome flow" or "the Labor Day campaign." The assistant looks up the exact name first, so you don't need to match it.
Say how you want it delivered. "Give me a 3-bullet summary for my exec update" or "put it in a table" saves a round of edits.
Ask it to verify. Good follow-ups are "Is that the full result, or was anything cut off?" and "Do those rows add up to the total?"
Weak: "How's SMS doing?"
Strong: "Compare SMS revenue, sends, and CTR for Sep 1–30 vs. Aug 1–31, and name the top 3 campaigns by revenue."
Verify the numbers before you share them
Most "wrong number" moments come from a definition mismatch, not bad data. Check these before a figure goes into a deck.
Metric or behavior | What it really means | What to do |
Revenue (two numbers) | Per-campaign and per-automation revenue uses click attribution with a 29-day window. Shop-wide revenue uses a wider definition that includes view attribution, so the shop total is always larger than the sum of the reports. Both are correct. | Use per-entity reports to compare campaigns and flows. Use the shop-wide metric for the account trend. Don't mix them in one comparison. |
CTR | Only available per campaign or automation. There is no shop-wide CTR. | "CTR across all campaigns from [start] to [end]" (the assistant rolls it up from campaign clicks and sends) |
Automation results | Reported per whole flow, not per message step | Use the Postscript app for step-level detail |
A/B tests | Reported as one row, not per variant | Use the Postscript app for variant results |
Names | Reflect the current name. A renamed campaign reports under its new name for all history. | Ask by ID or confirm the current name |
Clicks and revenue on recent sends | Keep rising for 29 days after a send as attribution accrues | Re-run reports on sends less than a month old before finalizing |
Sparse results | Campaigns or automations with zero sends are left out, not shown as zero | Treat missing as zero |
Doesn't match the app | Some figures may differ from what you see in Postscript | Tell us through feedback and we'll investigate with you |
Use cases by role
Retention and SMS marketing
Goal: see what's performing and pull the numbers for your next planning discussion.
"Rank campaigns from [start] to [end] by revenue and show sends, CTR, and unsub rate for each."
"Which campaigns had high CTR but low conversion rate from [start] to [end]?"
"Compare our welcome flow and abandoned cart automation on revenue per send from [start] to [end]."
"Show SMS revenue by day for [Black Friday week dates last year]."
"Which campaigns had the highest unsubscribe rate from Jan 1 to [today]?"
Growth and list building
Goal: track subscriber growth and popup performance.
"Show new subscribers by week from [start] to [end]."
"Show new subscribers by month for the last 12 months, from [start] to [end]."
"How did my popups perform from [start] to [end]?"
"Rank my popups by conversion rate from [start] to [end]."
Executives
Goal: a fast, trustworthy read on how SMS is doing
"Give me SMS revenue by month from [start] to [end], with month-over-month change."
"Summarize [last quarter dates] vs. [prior quarter dates] in 3 bullets: revenue, sends, and top campaign."
"Compare campaign revenue vs. automation revenue from [start] to [end]."
"Compare Jan 1–[today] to the same period last year: sends, revenue, and revenue per send, split by campaigns vs. automations."
Finance
Goal: tie SMS to revenue and track send volume. Note: Cost and spend data aren't available through the connector, so use the Postscript app or your billing export for those.
"How many messages did we send each month from [start] to [end]?"
"Revenue per send by month for campaigns and automations from [start] to [end]."
"Orders and revenue from campaigns and automations for [quarter dates]."
Before a revenue figure goes into a board or finance report, confirm which revenue definition it uses (see "Know the numbers" above).
Ecommerce and merchandising
Goal: see how promotions performed over SMS. Note: The connector can't show which products people bought.
"How did the Labor Day sale campaign do on orders and revenue?"
"Show daily SMS revenue for [two weeks around your last product launch]."
"Which campaigns drove the most orders from [start] to [end]?"
CX and support
Goal: spot messaging that creates friction or complaints.
"Which automations had the highest unsub rate from [start] to [end]?"
"Show unsubscribes per campaign from [start] to [end], sorted highest first."
"How many messages did we send per day from [start] to [end]?" (check for over-sending)
Agencies and partners
Goal: produce client-ready performance reporting fast.
"Build a monthly SMS report for [month] vs. [prior month]: revenue, sends, CTR, top 5 campaigns, top 3 automations, and subscribers."
"Which automations are enabled right now, and what did each send from [start] to [end]?"
What not to ask it right now
The Postscript 1.0 MCP reads analytics for your shop only. It can't change anything in your account, and some data isn't available yet.
Cost or spend. The connector has no cost data. Billing questions need the Postscript app or your billing export.
Individual messages inside an automation. It reports each automation as a whole. It cannot break a flow into its steps or show the copy of each message.
Message copy or creative analysis. Message text is not available yet.
Segment or audience breakdowns. Performance by segment is not available yet.
A/B test variants. It reports a test as one row, not per variant.
Deliverability or carrier detail. Not available yet.
What products people bought. Order contents are not available yet.
Open-ended strategy advice. The assistant sees your numbers, but it does not know Postscript products or best practices. Advice like "should I turn this off?" or "where should I move budget?" is not reliable yet. Bring those questions to your Postscript team. Use the connector for the numbers behind the discussion.
How to submit feedback
Feedback goes straight to our team, and the more you send, the faster this improves.
Just tell the assistant in plain words (typing or voice dictation both work):
"That revenue number doesn't match what I see in Postscript."
"That report was exactly what I needed."
"Give feedback: I wish this could do X."
The assistant attaches your comment to the answer it refers to, so we see exactly which result you meant.
Good to know:
Feedback is a note, not a fix. It doesn't correct a number or re-run a query. If a number looks wrong, tell us and we will investigate further.
The assistant only submits feedback when you ask it to. It will not send anything on its own.
Cheat sheet
Copy a pattern, then swap in your metric, dates, and scope.
You want | Prompt pattern |
A weekly check-in | "SMS sends, revenue, and CTR for [this week's dates] vs. [prior week's dates]." |
A monthly trend | "SMS revenue by month from [start] to [end]." |
Top performers | "Top 5 campaigns by revenue from [start] to [end], with sends and CTR." |
Underperformers | "Sort my campaigns by revenue per send from [start] to [end] and show the bottom five." |
One named thing | "How did the welcome flow perform from [start] to [end]?" |
List growth | "New subscribers by month from [start] to [end]." |
Popup performance | "How did my popups perform from [start] to [end]?" |
An exec-ready summary | "Summarize SMS results for [month] in 3 bullets for an exec update." |
A full export | "Every campaign and automation from [start] to [end] in one table: IDs, sends, clicks, orders, and revenue." |
A check on completeness | "Is that the full result, or was anything cut off? Do those rows add up to the total?" |
Flag a bad answer | "That number looks wrong. Give feedback." |