Deliverability
by
Send hygiene for ecommerce teams: identify the unengaged contacts costing you inbox placement, then apply suppression rules before each send rather than after the damage.
Deliverability fails slowly and then all at once. Sending to people who stopped opening two years ago does not produce an error, it produces a gradually worse sender reputation, and the first visible symptom is usually a revenue drop nobody can explain because the campaign metrics look normal. By then the repair takes months. The instinct to keep sending to the whole list is also the thing destroying the list, because the mailbox providers read engagement as the signal that you are wanted. This skill identifies the contacts costing you placement and applies suppression rules before each send.
What It Does
- Segments the list by engagement recency rather than by subscription date.
- Estimates the placement cost of each unengaged cohort before you send to it.
- Prices the revenue actually earned from the cohort you would be suppressing.
- Detects spam trap patterns and role addresses accumulating on the list.
- Sets a re-engagement path for the cohort still worth trying.
- Applies the suppression rules to the audience before each send.
Use Cases
- Find the cohort quietly costing you inbox placement on every send.
- Suppress a segment on evidence rather than on a round number of days.
- Justify shrinking the list by showing what the suppressed cohort actually earns.
- Catch role addresses and traps before a provider does it for you.
Example Session
My open rates are sliding. What should I suppress before Thursday's send?
Suppress the 24 month cohort now. It is 11% of your list and none of your revenue.
- Cohort | Contacts | Share of list | Revenue, 90 days | Action
- Opened in 90 days | 41,200 | 44% | $186,400 | Send
- Opened in 91-365 days | 23,100 | 25% | $12,700 | Send, reduced cadence
- No open, 12-24 months | 18,600 | 20% | $840 | Re-engagement, then suppress
- No open, 24 months plus | 10,300 | 11% | $0 | Suppress now
The 24 month cohort has produced nothing in three months. Suppressing it drops your list to 82,900 and should lift placement across the rest within two sends. The 12-24 month group gets one re-engagement attempt because $840 is not nothing, and whatever does not open joins the suppression, which takes the unengaged 31% of your list out of the calculation providers make about you.
Why This Beats Prompting It Yourself
List Size vs Inbox Placement: The Hidden Risk
A generic prompt treats a larger list as a larger opportunity and suggests sending more often to revive the quiet part. That is the exact behavior that convinces a mailbox provider you are unwanted. The campaign metrics stay plausible while the engaged part of your list slides toward the promotions tab, and the revenue drop arrives a quarter later with no obvious cause.
When You Still Need The Revenue Behind The Cohort
Suppression is easy to recommend and hard to approve, because nobody wants to delete a third of a list on principle. This skill puts $840 over ninety days next to 18,600 contacts, which turns the argument into arithmetic. The cohort that earns nothing is the one costing you the cohort that earns $186,400.
Known Limitations
Open tracking has been unreliable since mail privacy protection started pre-fetching images, so engagement is read from clicks and orders where available and the output says when opens are carrying more weight than they should. Placement effects take two to three sends to show, so the improvement is never immediate. It applies the suppression rules and does not delete contacts from your platform.
Install It
- Download the ZIP.
It is free and there is no account to create. - Unzip it into your agent's skills folder.
Claude Code reads ~/.claude/skills/, which is hidden by default: the command in the folder block opens it. Other agents scan their own directory, so drop the same folder there instead. - Ask your agent to use it.
Restart the agent if it was already running, then it picks the skill up with no config.
skills folder
Copy
~/.claude/skills/deliverability/
SKILL.md
references/engagement-tiers.md
suppressions/
# one file per send, with the rule applied
# macOS: create the folder and open it in Finder
mkdir -p ~/.claude/skills && open ~/.claude/skills
# Windows: paste in the Explorer address bar
%USERPROFILE%\.claude\skills\Frequently asked questions
How is this different from my ESP's engagement segment?
An engagement segment is a filter on a date. This prices each cohort against the revenue it produces and the placement it costs, so the cut point is argued rather than assumed. It also reads clicks and orders rather than relying on opens, which stopped being trustworthy some time ago.
Which agents does the skill run in?
Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision.
What does it need connected to work?
Send history with clicks, order data joined to contacts, and the current list with subscription dates. It works without opens entirely, which is closer to how deliverability should be measured now anyway.
Will suppressing a third of my list hurt revenue?
On the evidence above, no: that cohort produced $840 over ninety days against $186,400 from the engaged part. The output shows the revenue at risk explicitly so the decision is yours, and a cohort that genuinely earns is never recommended for suppression.
How long before placement improves?
Two to three sends in most cases, longer where reputation has been poor for a while. Nothing improves within one send, and any tool that promises otherwise is describing something other than sender reputation.
What is included with the skill?
The skill itself, the engagement tier reference, the suppression rule format, and the trap detection patterns. It is free and security scanned. We re-publish the ZIP when the tiers change, so download it again if your cohorts start shifting.
Suppression protects the revenue you already have.
Polar joins engagement to lifetime value in one semantic layer, so the segment you hold back is judged on what it is worth, not only on open rate.
<script type="application/ld+json">{"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How is this different from my ESP's engagement segment?","acceptedAnswer":{"@type":"Answer","text":"An engagement segment is a filter on a date. This prices each cohort against the revenue it produces and the placement it costs, so the cut point is argued rather than assumed. It also reads clicks and orders rather than relying on opens, which stopped being trustworthy some time ago."}},{"@type":"Question","name":"Which agents does the skill run in?","acceptedAnswer":{"@type":"Answer","text":"Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision."}},{"@type":"Question","name":"What does it need connected to work?","acceptedAnswer":{"@type":"Answer","text":"Send history with clicks, order data joined to contacts, and the current list with subscription dates. It works without opens entirely, which is closer to how deliverability should be measured now anyway."}},{"@type":"Question","name":"Will suppressing a third of my list hurt revenue?","acceptedAnswer":{"@type":"Answer","text":"On the evidence above, no: that cohort produced $840 over ninety days against $186,400 from the engaged part. The output shows the revenue at risk explicitly so the decision is yours, and a cohort that genuinely earns is never recommended for suppression."}},{"@type":"Question","name":"How long before placement improves?","acceptedAnswer":{"@type":"Answer","text":"Two to three sends in most cases, longer where reputation has been poor for a while. Nothing improves within one send, and any tool that promises otherwise is describing something other than sender reputation."}},{"@type":"Question","name":"What is included with the skill?","acceptedAnswer":{"@type":"Answer","text":"The skill itself, the engagement tier reference, the suppression rule format, and the trap detection patterns. It is free and security scanned. We re-publish the ZIP when the tiers change, so download it again if your cohorts start shifting."}}]}</script>