How to pull your cold email analytics
How-to guide Updated 12 July 2026
How do you pull cold email analytics with AI?
Ask Claude for the numbers and it reads them straight from HotHawk into the chat: reply rate, positive reply rate and bounces, per campaign or across the workspace. Open tracking is off by design, so what you get back is honest interest and deliverability, not a pixel-inflated open rate that means nothing.
No credit card required.
Before you start
You need a HotHawk account and the HotHawk MCP server connected to Claude. It takes a couple of minutes and no code. Full setup is on the MCP server page.
Step by step
-
1
You ask Claude
Give me the reply rate and positive reply rate for my SaaS Q3 campaign.
Claude callsmcp__hothawk__workspaces_analytics_listworkspaceId current workspace scope SaaS Q3Claude reads the numbers straight into the chat: how many have sent, the reply rate, and the positive reply rate, so you see interest and not just activity.
-
2
You ask Claude
Now show me bounces and how each campaign compares across the whole workspace.
Claude callsmcp__hothawk__workspaces_analytics_listworkspaceId current workspace breakdown by campaignIt lists every campaign side by side with sends, replies, positive replies and bounce rate, so a deliverability problem shows up before it spreads.
-
3
You ask Claude
Which campaign is my best and which one should I fix first?
Claude callsmcp__hothawk__workspaces_analytics_listworkspaceId current workspace sort positive reply rateClaude ranks them for you and calls out the laggard, so you get a read and a next move, not another table to stare at.
Most cold email dashboards flatter you
A 60 percent open rate looks great in a screenshot. It also tells you almost nothing, because the pixel that measures it fires on preview panes, security scanners and image caches, and half the time it never fires at all. You end up optimising a number that does not connect to a single booked call.
HotHawk does not report opens, because it does not track them. Open and link tracking are off by design, both because a tracking pixel can work against your sending and because opens are the wrong thing to chase. What you get instead is the honest set: reply rate, positive reply rate and bounces.
Ask in plain language, get the real figures
Run it from Claude and you skip the reporting screen entirely. Ask how a campaign did and it reads the numbers back. Ask it to compare every campaign in the workspace and it lays them side by side. Ask which one to fix first and it ranks them and points at the laggard. Same data the app shows, read to you in the chat where you are already working.
Every one of those calls hits the same endpoint, workspaces_analytics_list, and it is read only. Claude is reporting, not changing anything, so you can pull the numbers as often as you like without touching a live campaign or its sending.
Numbers you can act on
The point of honest analytics is that the next move is obvious. A high positive reply rate means the play is working and is worth cloning onto a new list. A creeping bounce rate means a list needs cleaning or a mailbox needs a look, before it drags the rest of your sending down with it. Neither of those reads comes out of an open rate.
So the loop is short. Pull the numbers, see what is winning and what is slipping, then act, whether that is an A/Z test on the weak campaign or a relaunch of the strong one. The data is honest, so the decision is easy.
The HotHawk tools behind this
Every action here is a real endpoint. Claude reaches it over the MCP server, and you can hit the same endpoints directly from the REST API if you would rather build it into your own stack.
-
mcp__hothawk__workspaces_analytics_listRead sends, reply rate, positive replies and bounces per campaign or workspace
Frequently asked
Which numbers can Claude actually pull?
Sends, reply rate, positive reply rate and bounce rate, per campaign or across the whole workspace. Ask in plain language, "how did Q3 do" or "compare all my campaigns", and it reads the real figures back into the chat.
Why is there no open rate or click rate in here?
Because open and link tracking are off by design. A tracking pixel is one of the things that can hurt your sending, and open rates are easy to inflate and tell you nothing about interest. So HotHawk reports on replies and positive replies, the numbers that actually reflect whether the campaign is working.
What counts as a positive reply?
A reply that shows real interest rather than a bounce, an out of office or a hard no. It is the closest number you have to "this is turning into pipeline", which is why it is worth watching more than raw reply volume.
Do I need to write code to pull my analytics?
No. You just ask Claude over the MCP server and it reads the figures back. The same analytics endpoint is in the REST API if you want to pipe the numbers into a dashboard or your own reporting stack.
Can reading analytics from Claude change my campaign or my sending?
No. Pulling analytics is read only. Claude reports the numbers back and nothing about your sending changes, so warmup and mailbox rotation keep running exactly as they were while you look at the results.