Campaign and conversation analytics
The Analytics page answers three questions: is outreach working, which offers and buyers are moving, and where a conversation is waiting on you. Open it from Analytics in the menu, next to Campaigns. Each campaign, offer and company page shows its own slice of the same numbers.
The short version: replies are the result that counts. Opens are a hint, and are shown lighter for that reason.
Conversations
A wholesale campaign isn't a newsletter. Each buyer you email gets one thread, and every follow-up, counter-offer and reply stays in that thread. So the thing we measure is the conversation: one thread with one buyer.
Every rate divides by conversations, not by emails. If you send a buyer three follow-ups, that is still one conversation, so chasing never makes a campaign look worse.
A conversation counts in the period it started. A thread that began in September and got a reply in October is a September conversation.
Who counts as a reply
A reply is a message in the thread from someone outside your company. These never count:
- Your own team. A colleague writing in the thread is coordination, not a buyer reply. Your company's domains are set in Settings → Email → Company email domains. When the list is empty, we use the domains of your team's logins and connected mailboxes, but never public ones like gmail.com.
- Bounces. Mail that didn't arrive can't be answered.
- Auto-replies and "remove me" replies. They are shown in the thread, but they are not interest.
If a buyer copies a colleague, or a colleague joins later, that person is a participant. Participants don't count as extra recipients, so they never dilute a rate. A new person joining from the buyer's side often means someone with buying power is now involved.
Positive replies
When you classify a reply, the classes Interested, Question and Counter-offer count as positive. Not now, Not interested, Objection, Auto-reply, Unsubscribe and Other don't. A conversation is positive when its latest classified reply is positive, so a buyer who asked a question and then said no is not positive.
Classify replies from the email page, or ask Claude to work through the unclassified ones with the connector.
The metric dictionary
| Metric | What it means |
|---|---|
| Recipients | Distinct people you emailed in the campaign (To). CC'd colleagues count as participants, not recipients. |
| Conversations | Distinct email threads the campaign started. |
| Touches | Your sent messages in a conversation: the first email, follow-ups and replies. |
| Delivered | Conversations whose emails did not hard-bounce. A temporary delay (a 4.x.x code) is not a bounce. |
| Opened | Conversations with at least one open by a person. Apple prefetches, scanners, repeats and re-views after a reply are left out. A signal, not an exact count. |
| Engaged | Conversations where a person clicked a content link, such as an offer or deal page. Login and unsubscribe links don't count. |
| Replied | Conversations with at least one reply from someone outside your company. Bounces and auto-replies don't count. |
| Positive | Replied conversations whose latest classified reply is Interested, Question or Counter-offer. |
| Reply rate | Replied conversations divided by delivered conversations, always shown with the count. |
| Touches to reply | How many of your messages went out before the buyer's first reply. |
| Buyer response time | Median time from your message to the buyer's next reply. |
| Your response time | Median time from the buyer's message to your next reply. |
| Waiting on | You when the last message is theirs, them when it is yours, with the time since. |
| Participants | Other people outside your company who joined or were copied into the thread. |
| Deals | Deals linked to the campaign directly or created from one of its conversations, with open value and won value. |
| Unsubscribed | Recipients who opted out or reported spam after this campaign. |
Opens and clicks follow the rules in Email tracking: people only, repeats merged, scanners left out.
Small samples
A percentage from fewer than 20 conversations is shown in grey with its count beside it, like 50% (2/4). Two replies from four emails reads as 50%, but it tells you very little yet.
Every percentage on the page carries its count, so you can always see how many conversations are behind it. When the page picks a best or lowest campaign, it only compares campaigns with at least 20 delivered conversations.
Filters
Every filter lives in the page address, so a filtered view can be bookmarked or sent to a colleague.
- Period. Last 30 days, last 90 days, this month, last month, this quarter, or a custom range. Conversations are counted in the period they started.
- Mailbox. Only conversations started from one sending mailbox.
- Sent by. Only conversations whose first email that person sent.
- Offer category. Only conversations about offers in one category.
- Buyer type. Only conversations with contacts of one buyer type. Shown when your contacts have a buyer type set.
Reading each tab
Overview
The funnel for the period, and how it compares with the previous period of the same length. Rate changes are in percentage points, so a reply rate going from 10% to 15% shows as +5.0 pts.
Below it: Waiting on you, the conversations where the buyer wrote last, oldest first, and Hot right now, buyers who opened several times, came back to an old email or clicked an offer page in the last day. Start your day with these two lists.
Campaigns
One row per campaign, with its funnel. Click a column to sort. Rate columns always put real samples above small ones, so a 100% reply rate off one email never tops the list.
Offers and categories
Which lots pull buyers. Interested buyers clicked the offer page or replied. Offers are sorted by positive replies, then interest. Conversations with no offering linked can't count toward an offer, so the page tells you how many there are. Link an offering on the campaign to include them.
The categories table sums the same numbers by offering category. It helps when deciding what to buy next.
Buyers and accounts
Everyone at a company, summed. In B2B, several people at one account engaging at the same time is the strongest buying signal there is. People without a company record are grouped by their email domain.
Accounts are ranked by a simple score: a reply is worth 3, a positive reply 2 more, a click on an offer page 2, and an open 1. It is a way to sort the list, not a metric to report.
Gone quiet lists accounts that engaged before and then had no opens or replies for 30 days. Those are worth a call.
Follow-up effectiveness
Reply rate by touch shows, for each message (the first email, follow-up 1, 2, 3 or later), how many still-unanswered conversations it went out to and how many replied right after it. If follow-up 3 almost never gets a reply, stop at two.
Days before the first follow-up compares reply rates by how long you waited before the first follow-up. Only conversations still unanswered at that point count.
Timing
A weekday by hour grid of when buyers open and reply, on your workspace's clock. Replies count three times as much as opens in the shading. Use it to pick send times.
Senders and deliverability
Only workspace owners see this tab. Per mailbox: what went out, the bounce rate, spam complaints, opt-outs and the reply rate, with weekly sends. Weeks with a bounce are marked.
It is a deliverability alarm, not a leaderboard. A sudden rise in bounces or complaints on one mailbox usually means a sending problem with that mailbox or its domain.
Links
The offer and deal pages buyers clicked, by clicks from people and by how many buyers clicked. Login and other utility links are listed separately because they say less about interest in a lot.
Asking Claude
With the AgentlyLeads connector, Claude reads the same numbers:
- campaign_analytics: one campaign's funnel and every conversation with a reason line, or every campaign in a period.
- offer_interest: buyers ranked by interest in one offer.
- account_engagement: everyone at one company, or the most engaged accounts and the ones that went quiet.
Try "Which golf lot had the most real interest this week, and who should I call first?"