Skip to guide

See Distill in action

Request Demo

Tell us a little about yourself and we’ll follow up to schedule time.

Browse guides
Read Markdown

Reporting and results

On this page 11 sections

Distill reports who outreach reached, how people responded, which conversations need attention, and recorded meeting bookings. It also provides tools for investigating where outreach loses momentum.

Reports distinguish accepted connections, replies, positive replies, and recorded bookings. Attendance, sales, and partnerships are later outcomes that need their own evidence.

Available reporting#

Report or viewWhat it helps explain
Project and campaign cardsPeople contacted, connections accepted, replies, recorded meetings, campaign status, and work waiting to happen. Project totals combine activity across campaigns.
Project overview statisticsOutreach activity by week, month, or quarter, with campaign selection and a switch between activity dates and results grouped by outreach period.
ResponsesActual customer responses, positive response examples, and breakdowns by job title, industry, channel, and campaign. Rows link to the conversation.
Campaign dashboardWhere campaign entries sit now, including work that is new, blocked, ready, waiting, in conversation, or closed. Counts can open the corresponding entries.
Weekly assistant reportA selected complete week and the two weeks before it, with project and campaign counts, comparison rates, and explanations for unavailable metrics.
Performance analysis through an assistantResults at each outreach step, reply examples, audience comparisons, and changes to targeting or messaging, to help explain performance.

Each view has its own counting rules, described below.

What the main numbers mean#

People contacted counts each person contacted at least once within the report's project or campaign and date range. Over all time, the same person counts once in the project even if several campaigns reached them. A sent connection request counts, including one that has not been accepted. Recorded direct messages from the sender can also count when they belong to the campaign's contact history. An imported person who has not been contacted does not count.

Outreach started counts a person in the period when they were first contacted within the report's project or campaign. A follow-up can make someone count as People contacted again in a later period, but it does not start their outreach again. Filtering a project's report to a campaign still uses the person's first contact in the project. An individual campaign card uses their first contact in that campaign.

Outreach sent, the top row of the overview statistics table, counts sent connection requests plus first campaign emails from campaigns that do not send connection requests. It is an activity count, not the total number of distinct people or all messages and follow-ups. Two senders sending a connection request to the same person contribute two requests.

People emailed counts distinct people emailed in the group being reported. Multiple emails to one person do not become multiple people.

Email sequences started, in the weekly assistant report, counts people sent their first email within that report's project or campaign during the week. The project summary uses the first email across its campaigns; an individual campaign uses its own first email. A later follow-up does not count as another start.

Replies and New responses usually count people who replied, not every message in an ongoing exchange. The overview activity view counts people whose first reply belongs to the period. The Responses list can also show a later reply or booking that occurred in the selected period. This is why an ongoing conversation may appear in the list without adding another New response.

Positive responses in the customer Responses report means replies classified as positive. Neutral, negative, and unclassified replies remain available in the All view, but do not appear in the positive selection. That classification is an interpretation, so the conversation itself matters when judging interest or intent.

A person can appear in both the LinkedIn and email reply counts. Adding channel counts is not necessarily a count of unique responders across the project.

Activity dates and outreach groups#

The default overview view, Activity by week, places activity in the period when it happened. A connection accepted this week belongs to this week's acceptance count, even if its request was sent earlier. A booking belongs to its booking period.

Group Stats by Cohort changes the question to: what happened to the outreach sent in this period? The table becomes Results by outreach week. A cohort is simply a group defined by when its outreach went out.

For campaigns that send connection requests, the request date determines the group. For campaigns that do not send connection requests, the first campaign email date determines it. Later acceptances, relevant replies, emails, and recorded bookings can add to that original group's results.

For example, a request sent in week one, accepted in week two, followed by a booking in week three contributes to three different weeks in the activity view. In the grouped view, its acceptance and booking belong to week one's outreach results. That older column can improve as outcomes arrive.

Recent groups have had less time to respond. A lower rate for the newest week alone does not establish that targeting or messaging got worse.

How is each rate calculated?#

The customer overview's grouped view and the weekly assistant report use these comparisons:

RateWhat is countedWhat it is divided by
Connection acceptanceRequests from that outreach group that were acceptedConnection requests sent in that group
LinkedIn replyDistinct people behind the group's accepted requests who replied on LinkedInAccepted connection requests in that group
Email replyDistinct people in the group's emailed population who replied by emailDistinct people emailed in that group

Negative replies still count as replies. People emailed as part of a group's outreach belong to that group even if the email happened later. The denominator is the number a rate is divided by. If it is zero, there is no meaningful percentage to calculate.

The activity view shows counts rather than these percentages. Dividing this week's replies by this week's sends mixes people at different stages. The deeper performance-analysis funnel uses different denominators, stated in its results, including replies after a first LinkedIn message. Its reply rate is not interchangeable with the overview's LinkedIn reply rate.

Dates, filters, and totals#

The overview supports weekly, monthly, and quarterly periods and a campaign filter. Reporting weeks run Monday through Sunday in UTC. When comparing reports, check which campaigns and date ranges each one includes, especially near period boundaries.

Total in the customer statistics table covers all time for the selected campaigns. It is not a sum of just the visible columns. Person-based metrics count each qualifying person once across the selected history, while activity metrics keep their activity-counting rule. A total can include older history and still be lower than a sum that counts the same people in several periods.

Loading or unavailable totals are distinct from a successfully loaded zero.

Selecting a period in the statistics table filters the Responses activity below to those dates. It does not filter those responses to the clicked metric, selected campaigns, or people in that outreach group. The Responses headline cards still show all-time project values. A selected week can therefore show responses from different people than the corresponding grouped statistics column.

Without a selected statistics interval, response sections have their own All time, Past week, Past month, and Past quarter controls.

Why campaign and project counts can differ#

A person reached in two campaigns can count once in each campaign card, but once in the project total. Adding campaign cards can therefore overcount distinct people.

Historical project engagement totals keep relevant recorded activity when a candidate is later dismissed. A candidate is a person's entry in a particular campaign. The campaign dashboard counts these entries, excluding dismissed ones. It can contain more than one entry for a person, for example through different introduction paths, so its entry count is not necessarily a count of unique people.

Reporting counts also differ from billing usage. Agent campaign billing counts a target once per organization within six months, while a report applies its own project, campaign, and date scope. Use the account's billing view for paid usage and balances, rather than treating a performance total as an invoice. Plans and usage.

These counting rules can explain why totals differ. A discrepancy within the same report, scope, period, and counting rule still deserves investigation.

Responses, evidence, and export#

Responses includes the person's role and company, the channel and campaign, their response's positive, neutral, or negative classification when available, and a link to the conversation. Positive response examples make it easier to review what people actually said behind a summary count.

The response list currently loads up to 200 candidate matches. Its examples and breakdowns are based on those loaded rows. The Positive responses headline also uses those rows, so it is not a complete lifetime positive-response total when the list reaches that limit. Most other headline metrics are calculated separately using all records that qualify for each metric.

This also means the response breakdowns describe who appears among the responses. A large share of responses from an industry is not by itself that industry's reply rate; a rate also needs the number of people reached in that industry. The performance-analysis segment tools supply reached and replied counts for that comparison.

With the required project access, the handed-to-client CSV export lists people whose conversations were handed to the customer. It contains one row per unique person, with name, role, company, handoff date, meeting date when recorded, LinkedIn URL, a conversation link, and email when available. Missing information leaves an empty cell. A handoff is not proof that a meeting occurred or that the customer replied.

For exporting contact and company lists, see Lists. The report export described here covers handed-off conversations.

Meetings and calendar coverage#

A calendar link sent, a link visited, a detected booking, and an attended meeting are separate events. The Meetings booked metric reflects recorded booking information. It does not count deals won or verify that the meeting took place.

In the project overview statistics, meeting values depend on usable Google Calendar connections for the senders in the current reporting scope. If coverage is incomplete, the meeting row is hidden and the page explains what is missing. Filtering to campaigns whose senders are connected can make their meeting counts available even if another campaign's sender is not connected.

Loading, failed coverage checks, and missing calendar connections do not mean zero meetings. A zero shown in another summary also does not prove no meeting occurred unless the calendar coverage is known.

Coverage is only one part of recording a meeting. The booking must also be recognized and associated with the person or campaign. A meeting arranged elsewhere may be absent from Distill's reporting. Corrected or cancelled booking records can change historical meeting counts.

The current overview shows bookings in their booking period in the activity view and against the outreach group in the grouped view. A positive reply alone does not create a meeting count. Campaigns and personalization explains booking and reply handling; Integrations explains the connected services.

Sends, opens, bounces, and matching results to campaigns#

A draft, scheduled message, recorded send, and detected reply are different stages. Reporting a send does not prove that an email landed in the inbox rather than spam, or that a person read it. Failed attempts and bounces need their own evidence.

When tracking is enabled and the initial email qualifies for it, Emails opened counts that email if its signature image loads. Repeated image loads do not add more opens for the same tracked send. Blocked images can hide real reading, while a provider loading images automatically can record an open without a person reading. The count does not verify that someone read the email and is not a reliable measure of interest. The Responses report can show this count; the overview statistics table does not show an email-open-rate row.

Replies and bookings also depend on what the connected channels recorded and how Distill matched that activity to a campaign. A missing record does not conclusively prove that no response occurred. If an outcome appears under the wrong campaign, the conversation, sender, recipient, channel, and timestamps help distinguish a reporting issue from a difference in date range or selected campaigns.

Activity and screening history#

A project has separate Activity History and Screening History views. Both require the relevant project access; the current reads require permission to edit the project.

  • Activity History records supported targeting, campaign, and membership changes, with who or what made the change and when. Filters help narrow the timeline. A campaign-template update can show that a change happened without reproducing every wording change in that row.
  • Screening History records screening runs, rule changes, and supported per-person decisions. A run can show counts and a breakdown of affected people, with reasons and available evidence. This helps explain why a person was screened, released, or held.

These views support investigation, but they are not a complete reconstruction of every past setting or event. An Updated timestamp can change without a corresponding detailed history entry. Missing history does not prove that nothing changed. A run marked unfinished may have processed only part of its audience, so its counts should not be read as a completed review.

Project history also differs from a campaign's message history and conversation activity. Each answers a different question: what changed in the project, what changed in the message, or what happened with a particular person. The available links and returned records establish which details can be inspected.

What an assistant can read#

The assistant's tools provide several views of performance. Which ones are available depends on its current Distill connection and the signed-in person's access.

CapabilityInformation available
Project performance summaryAll-time and recent 7-day or 30-day people contacted, replies, connections, recorded bookings, responses waiting on the customer, campaign states, and targeting totals.
Weekly performanceA selected complete UTC week and the two before it, project and campaign activity counts, grouped rates with the counts used to calculate them, and explanations when channel or calendar data is unavailable.
Outreach funnelResults at each step: connection requests and acceptance, people who accepted but were not messaged and what held them back, first messages and emails, bounces, replies, and positive replies.
Audience comparisonsPeople reached, replies, and positive replies by seniority, job function, industry, company size, time in role, and source list, with uncertainty ranges.
Outreach examplesA sample of people, sent connection notes and first messages, reply text, profile context, and current hold information.
Change historyRecorded targeting changes, campaign changes, screening changes, published message versions, and weekly outreach alongside them.

The weekly tool defaults to the most recent completed week and refuses an unfinished week. A company-wide report includes active projects the person can edit, up to 20 active projects. It reports the number left out when that limit applies. Quiet campaigns and projects may also be omitted, as may a metric that is zero in all three weeks. The result separately identifies metrics unavailable because of channel or calendar coverage.

Weekly activity counts and the counts used for grouped rates can differ because they answer different questions. Each rate uses the numerator and denominator returned with it. Dividing nearby activity counts may produce a different and misleading rate.

The performance-analysis tools generally look at people first reached within the selected date range, then follow their outcomes. They can compare rates with percentiles across other campaigns, showing where a rate falls within that comparison group. This provides context for performance, rather than identifying its cause.

Small groups, recent outreach, incomplete historical change records, and limited samples are meaningful limits. Outreach sampling returns up to 40 examples per request; it is not a full campaign export. Earlier edits and list imports may be absent from the change timeline, so no recorded edit does not prove nothing changed.

Using an assistant explains the available actions when analysis leads to a proposed campaign change.

Common questions#

Does a low reply rate mean the message is bad?#

Not by itself. The outreach funnel separates people who were never successfully reached, accepted connections that did not receive a message, email bounces, and sent messages that did not get replies. Audience fit, timing, the sender's profile, and the message can all matter. Reading the replies and comparing groups with enough people helps explain the result.

Can I compare this week's results with last week's?#

Yes, with the same scope and definition. Activity counts describe what happened in each week. Grouped rates describe outcomes from each week's outreach and can keep changing. The newer group has had less time to produce responses.

Why do I see responses when New responses is zero?#

The Responses list can include a later message in an existing conversation or a booking. New responses counts first responses, not every subsequent message.

Do positive replies mean qualified leads?#

No. A positive sentiment classification is useful evidence of a response's tone, not a completed qualification process. The person's fit, what they actually said, and the next agreed step still matter.

Does zero meetings mean nobody booked, even if my calendar is not connected?#

No. Without the required calendar connections, meeting reporting is incomplete or unavailable. A meaningful zero requires knowing which senders' calendars are connected and which campaigns and dates the report covers.

Can the Responses breakdown tell me which audience responds best?#

It shows the mix of loaded responses. A fair reply-rate comparison also needs how many people in each audience were reached. The assistant's segment analysis supplies those counts and uncertainty, but small samples still limit the conclusion.

All guidesBack to top
Distill product guidesAll pages as text