Junior Post-Mortem: post-event survey and collecting the numbers

As a junior coordinator the post-mortem isn't yours: it's the senior's. Your job is to collect the raw numbers, distribute the post-event survey within 24-48 hours, fill in the KPI vs target sheet without interpreting it, archive the materials in a structured way. It seems like the least glamorous work of the entire cycle - in reality it's what determines the quality of the senior's post-mortem, and therefore the speed with which the team improves event after event. Three case studies (Eventbrite Organizer Lab, Cvent CONNECT customer ops, Scalapay all-hands T+30) and three first-year anti-patterns.

TL;DR

The post-mortem is the most neglected and most critical skill of the event cycle. Skift Meetings, in its State of Corporate Events 2024 report, states that only 38% of corporate events receive a structured post-mortem within two weeks, and 19% never receive one. For the junior coordinator at L1 level, the post-mortem isn't run: it's fed. Your job is to collect four assets that the senior will use to conduct the full post-mortem: (1) post-event survey distributed within 24-48 hours with a completion rate ≥25% (Cvent benchmark), (2) KPI vs target sheet with three columns (KPI, target from brief, measured actual), (3) structured materials archive (slides, photos, recordings, signed invoices), (4) a 1-2 page L1 report with raw facts without cause-effect interpretation. The three things that separate a junior who grows from one who burns out are: distributing the survey super fast (24-48 hours), reporting numbers without interpreting them (the junior who says the event went great without numbers is in an anti-pattern), and keeping the archive organized from day one (recovering a lost invoice after the fact costs 4-6 hours). You'll also find three case studies (Eventbrite four-section format, Cvent CONNECT playbook iteration, Scalapay T+30 follow-up loop), three anti-patterns (survey distributed late, DIY interpretation of the numbers, disorganized post-event archive) and a Post-Mortem Retrospective template that the senior fills in starting from your raw inputs.

When you need it / Prerequisites

This guide is applicable in the days immediately following the event. The typical case is a single-BU event of 50-150 attendees where you are the junior coordinator and you handled the operational execution; the senior conducts the strategic post-mortem in the days T+5 to T+15 but needs your raw inputs collected between T+1 and T+5. Without your inputs, the senior's post-mortem starts from memory and impressions, not from data - an enormous loss of value.

There are four prerequisites. First: you have the participant distribution list updated post-event, with the actual attendance figure (who was actually in the room, validated by registration or check-in). For a single-BU event of 80 people, typically 75 registered, 62 actually present (presence rate 83%, in line with the Cvent benchmark of 80-85%). Second: you have access to a company survey tool (Typeform, Google Forms, Microsoft Forms, SurveyMonkey, Mailchimp survey). For small events free Google Forms may be enough; for events with an external audience you need a tool with company branding and better deliverability. Third: you have the team's standard survey template. Fourth: you have access to the original event brief where the baseline KPIs you'll have to measure are declared.

What is not a prerequisite at L1 level is the ability to interpret the numbers strategically. Interpretation is L2-L3: the senior who receives the KPI vs target sheet reads not only the numbers but also the underlying causes (why is the presence rate 83% and not 90%? Why is the NPS 7.2 and not 8.5?), and proposes strategic action items for the next event. At L1 level your job is to provide complete and accurate numbers, not to comment on them. The junior who starts interpreting early on risks producing two categories of errors: false causal attributions (e.g. low NPS because the catering was poor when in reality it was the session timing that was badly calibrated) and unwarranted projections (e.g. for the next event we should do X where X is the junior's opinion without supporting data).

If one of the prerequisites is missing, stop. Without an accurate attendance list, the presence rate is made up; without a survey tool, you distribute by hand via email and the completion rate collapses; without a standard template, every event has a different survey and cross-event benchmarks are impossible; without the original brief, you don't know what to measure against. The cost of fixing this now is a few hours; the cost of proceeding without is the total loss of the post-mortem's value.

Conceptual framework: collect raw numbers, don't interpret them

The L1 post-mortem framework distinguishes three types of output and attributes them to different seniority levels. The L1 raw collection level produces the uninterpreted numbers: presence rate, NPS, survey completion, costs vs budget, invoice count. The L2 causal analysis level produces the interpretations: why the numbers are what they are, which underlying causes, which patterns emerge. The L3 action plan level produces the action items for the next event: what to change, who owns it, deadline.

The operational rule for the junior is clear-cut: you produce L1, the senior produces L2-L3. If at L1 level you're asked to produce L2 or L3, you have two options: either your role is already a mid disguised as a junior (a common case in scale-ups under 200 employees where roles are fluid), or there's a delegation problem worth discussing with the senior. The difference is not one of value - both levels are important - but of temporal sequence and developed sensitivity. Correctly interpreting the numbers requires 12-18 months of exposure to multiple similar events to build benchmark intuition; the junior who interprets without that exposure risks misguiding the senior.

The framework anchors to the CMP domain Evaluation Process (5%) and the MBECS domain L. Evaluation. The 5% weight in the CMP exam seems low, but it's deceptive: the evaluation process is the multiplier of everything else. Events that don't measure accurately can't improve; events that measure accurately improve event after event with an exponential curve. CMP International Standards 2025 makes this logic explicit in the rationale of the Evaluation domain: it defines it as the skill that closes the learning loop.

The operational principle is the 24-48 rule: the post-event survey is distributed within 24-48 hours of the event. Beyond 5 days, completion rates drop drastically (the Cvent Blog cites 28% completion at 24 hours vs 9% at 7 days). If the junior skips the survey in the first two days because I need to rest or I need to finish the invoicing first, the post-mortem loses the most important dataset. The operational rule: the survey is the FIRST post-event task, not the last. It's sent while the public's memory is still warm.

The final principle for the junior is the discipline of the structured archive: every piece of material (speaker slides, event photos, recordings if there are any, signed invoices, inbound feedback emails) ends up in an organized folder on the same day it arrives. Not I'll sort it out at the end of the week. The folder has 5-6 canonical sub-folders: speaker_materials, event_photos, recordings, invoices, feedback_emails, final_report. A scheme replicable at the next event, a scheme queryable 6 months later for internal audits or for the lessons learned of a subsequent junior.

Operational steps: six steps to close the event cycle

The six steps cover from days T+1 to T+10. Times indicated for an event of 50-150 attendees.

Step 1 - Distribute the post-event survey (T+1 to T+2)

The first post-event task is to distribute the survey. The operational rule: it must go out within 48 hours of the close of the event, ideally in the first 24 hours. The completion rate has a very steep descending curve: 28% at 24 hours, 18% at 72 hours, 9% at 7 days (Cvent Blog data).

The team's standard survey typically includes 8-12 questions in three categories: (a) overall satisfaction (NPS-like, 0-10 scale), (b) quality of specific operational dimensions (catering, A/V, location, duration, agenda, speakers), (c) three open questions (what you liked most, what you liked least, suggestions for the next event). Total length: completable in 3-5 minutes. Above 15 questions the completion rate collapses.

The junior doesn't write the survey: they receive it ready-made from the team (company template) or derive it from the standard industry template. Cvent Blog, MPI and PCMA publish free, adaptable survey templates. The junior's rule: don't add or remove questions from the standard template without the senior's OK. Adding questions stretches the length and lowers completion; removing questions breaks cross-event comparability (if we measure catering 4 times a year with 5 standard questions, we have a trend; if every event has 5 different questions, we have nothing).

On distribution, two practical rules. First: send from the official channel (company Mailchimp or Outlook with an institutional signature) to ensure deliverability. Second: in addition to email, post the survey link in the dedicated Slack/Teams channel too (see L1-C4 internal communications): the public who don't read the email find it in chat and complete it anyway. The cross-channel pattern doubles the completion rate from an average of 25% to 35-50% in scale-up corporate events where the Slack channel is active.

A useful practice is the T+5 survey reminder: for those who haven't yet completed it 4-5 days after sending, a second polite email with the line 5 minutes to help us improve. The reminder typically recovers an additional 10-12% of completion. Beyond T+10 you don't insist - the public's fatigue outweighs the value of the additional data.

Step 2 - Fill in the KPI vs target sheet (T+2 to T+4)

While the survey collects responses (typical cadence: 60% of responses arrive in the first 72 hours, 90% within 7 days), the junior fills in the KPI vs target sheet. It's an Excel document with three columns for each KPI in the brief: KPI name, target from brief, measured actual. For a single-BU event of 80 people, typically 3-5 primary KPIs (e.g. presence rate, NPS, cost per attendee, coverage of the target audience).

The KPI vs target sheet is a document of facts, not opinions. Correct example: Presence rate: target 85%, actual 83% (62/75 registered). NPS: target 7.5, actual 7.2. Cost per attendee: target €120, actual €128 (+6.7%). Target audience coverage: target 80% of the Engineering team, actual 87% (62 present, team target 71). Wrong example: Presence rate good; NPS disappointing; costs a bit over budget but acceptable. The difference is the level of interpretation, and the wrong level is L1 (the judgments good/disappointing/acceptable are already interpretation).

A useful sub-practice is the percentage variance: instead of writing only target vs actual, also calculate the % variance (positive if under-target, negative if over-target). Example: cost per attendee variance +6.7% (over budget); NPS variance -4% (under target). The % variance makes the magnitude of the deviation immediately readable and lets the senior prioritize the deep-dives during the post-mortem (variances above 10% are high priority, below 5% are low impact).

An important point: some KPIs aren't measurable at T+5 because they require a delay (e.g. increase in the team's sense of belonging by 0.5 points measured 60 days after the event via Officevibe survey). For those KPIs, the sheet states measurement at T+60 and the junior schedules a calendar task at 60 days to collect the data. Without this planning, the data is lost and the KPI remains unverifiable.

Step 3 - Post-event materials archive (T+1 to T+5)

Structured archiving is the most routine task and the one where the junior's discipline makes the difference between everything in order in 6 months and nothing recoverable in 6 months. The post-event folder typically has 5-6 sub-folders: speaker_materials (slides, any scripts), event_photos (organized by moment - registration, keynote, networking, break, group photos), recordings (audio or video of sessions if there are any), invoices (receipts from suppliers, organized by supplier with a file name that includes date + invoice number), feedback_emails (export of the important inbound emails collected during preparation), final_report (the 1-2 page document of the junior's operational report, see step 5).

On speaker materials: the operational rule is to ask for the slides within 48 hours via email with the subject Slides [Speaker name] | Engineering All-Hands | 06/06. Typically 70-80% of speakers send by the deadline; for the remaining 20-30% a second reminder is needed at T+5. A useful practice: asking for the slides BEFORE the event (e.g. at T-2 for A/V back-up) already gives you the copy. Cvent Blog cites this practice as the pre-collect rule: better to ask before you need it than to look foolish having to chase it down afterwards.

On photos: if you've hired a professional photographer, you receive the edited photos within 5-7 days of the event. You organize them into 4-5 thematic sub-folders and select 15-20 highlight photos (the best ones) for internal communication or marketing use. The highlight photos end up in the follow-up (see L1-C4) and in the marketing team's cross-event archive.

On invoices: the junior validates each invoice against the original quote, identifies variances, and forwards to administration within 5-7 days of the event. If the variance exceeds 10%, escalate to the senior before forwarding. The practice works well when the junior already has an Excel tracking sheet with three columns (supplier, quote, final invoice) filled in during preparation. At the end of the event you just compare the two columns.

Step 4 - Raw analysis of the survey results (T+7 to T+10)

By 7-10 days after the event the survey has collected the bulk of the responses (typically 85-90% of the final total). The junior does the raw analysis of the results and produces a numeric sub-report for the senior. The analysis is basic and technical: mean, median, distribution on numeric scales (NPS, satisfaction); top-3 mentioned in the open responses (what was liked most, what least, suggestions).

On the NPS calculation, a note: the standard NPS is the difference between the percentage of Promoters (9-10) and the percentage of Detractors (0-6); the Passives (7-8) don't count. Example: 100 responses, 55 Promoters, 30 Passives, 15 Detractors → NPS = 55 - 15 = 40. The value runs on a -100 to +100 scale. For internal corporate events, the Cvent benchmark: NPS 30-50 = good, 50-70 = excellent, above 70 = outstanding. Below 30 is a signal of operational problems. The junior reports the number, the senior interprets the meaning.

On the open responses (the three qualitative questions), the L1 technique is the mention count: you read all the responses, identify the three most frequently mentioned themes for each question, report the numbers. Example: What was liked most: 1) speaker Marco (mentioned 18 times), 2) catering (12 times), 3) breakout format (9 times). What was liked least: 1) duration (15 times), 2) guest wifi (10 times), 3) last-minute venue change (7 times). Suggestions: 1) invite external speakers (12 times), 2) reduce duration by 30 minutes (10 times), 3) improve wifi (8 times). The junior doesn't interpret why the public mentioned duration 15 times - they report it. The senior interprets.

A useful sub-practice is the weak signals note: if a single but relevant mention emerges in the open responses (e.g. the catering had very few vegan options mentioned once), the junior reports it separately as weak signal, to be verified at the next event. These weak signals are often more predictive than the high ones - whoever mentioned the vegan option is the person who probably won't return to the next event if it isn't resolved.

Step 5 - 1-2 page L1 report to the senior (T+8 to T+10)

The junior's final deliverable is a 1-2 page report delivered to the senior 8-10 days after the event. The report contains: (1) numeric summary (KPI vs target with % variance), (2) survey summary (completion rate, NPS, top-3 mentions for each open question), (3) three relevant operational facts that emerged during the cycle (e.g. the A/V supplier had 2 technical incidents, recovered within 4 minutes, presence rate of the Engineering team 87%, target was 80%, budget variance +6.7% due to an unplanned time extension), (4) any weak signals or anomalies deserving the senior's attention.

The report does NOT contain: causal interpretation (the A/V supplier was imprecise because we chose the lowest quote), overall quality judgments (the event went great / badly), proposed action items for the next event (next time we should change A/V). That content is the senior's L2-L3 in the official post-mortem; the L1 report is the factual base to start from.

A sub-rule of writing: the L1 report is in a neutral voice, data-based, written like a status report not like a narrative. The typical sentences of the L1 report start with measurement verbs (the NPS is 7.2, the budget variance is +6.7%, 83% of registrants attended); they don't start with evaluative verbs (we managed to, we failed on, a positive surprise was). The rule: nouns and numbers before adjectives and adverbs.

The L1 report delivered to the senior 8-10 days after the event allows the senior to conduct the official post-mortem between T+12 and T+15. The cadence is important: the post-mortem should happen while the team's memory is still vivid (within T+15) but after the data is complete (post T+8 when the survey has collected >85%). This 7-day window is the golden zone of the post-mortem.

Step 6 - T+30 follow-up (iterative culture)

A less common but highly valuable practice is the T+30 follow-up: 30 days after the event, the junior verifies the status of the action items generated by the senior's post-mortem. Have they been implemented? Do they have an active owner? Are they about to slip to the next event without being closed? The scheme is simple: an Excel sheet with the action items, owner, deadline, status (todo / in-progress / done / dropped). The junior updates the sheet with the senior in 30 minutes.

This practice is explicitly codified in the GitLab Handbook and in the practices published by Scalapay (see the case study below). The value is in closing the learning loop: without the T+30 follow-up, the post-mortem's action items end up in a drawer and the next event repeats the same mistakes. With the T+30 follow-up, the implementation rate of action items goes from an average of 14% to 71% (PMI Library data on 1500+ post-mortems).

For the junior just starting out, the T+30 follow-up is also an exercise in personal professional growth: tracking what translates into action and what doesn't trains the sensitivity to tell apart the realistic action items (with owner, deadline, clear scope) from the ceremonial ones (vague, without owner, indistinguishable from aspirations). Developing this sensitivity is crucial for the leap from L1 to L2 along the Post-Mortem axis.

A useful sub-practice of the T+30 is the recurrence tracker: if the same action item appears in two or three consecutive post-mortems without being implemented, it's a signal of a systemic problem - the action isn't done not out of forgetfulness but because there's a structural blocker (e.g. budget not available, owner not really accountable, unresolved external dependency). The recurrence tracker is kept on a cross-event Excel sheet with the repeated action items. When an action item appears for the third time, structured escalation to the senior with a proposal: this action item has appeared in the post-mortems of three consecutive events without being closed. Is it worth discussing whether we really want it or whether it should be reformulated?. This practice transforms the junior from a passive data collector into an active contributor to the hygiene of the team's backlog.

A cultural note specific to the Italian context: in the IT/EU scale-up market many companies have not yet codified the T+30 post-mortem practice. It's almost always a practice that the individual team adopts autonomously, and that the junior coordinator can propose if it doesn't exist. The typical conversation with the senior: I'd like to propose a T+30 practice: 30 days after the event, we check the status of the post-mortem's action items. 30 minutes, a shared Excel sheet. It works in scale-ups like Scalapay (case study above). Can we try it at the next event?. Nine seniors out of ten say yes, even just out of curiosity. After two iterations the practice becomes standard, and the junior who proposed it sees the authorship positively attributed to them in their career.

Real case: Eventbrite Organizer Lab, Cvent CONNECT iteration, Scalapay T+30

Three case studies that cover three complementary dimensions of the junior post-mortem: the editorial structure of the Eventbrite case studies, the playbook iteration of Cvent CONNECT as an institutionalized methodology, the T+30 follow-up of Scalapay as a loop-closing practice.

Eventbrite covers the dimension of the editorial format. Cvent CONNECT covers the dimension of institutionalized iteration.

For the Italian scale-up context, the useful reference is a fintech that has publicized its T+30 practice.

Three additional sources for those who want to dig deeper: the PMI Learning Library has articles on post-mortems applied to different operational contexts; the Google SRE Book - Postmortem Culture codifies the blameless culture of the post-mortem in IT operations (applicable to events); Harvard Business Review - Managing Yourself has articles on the personal learning loop.

Anti-patterns: the three most frequent errors of the junior post-mortem

Templates, checklist and next steps

The most useful template for the junior in the post-mortem phase is the Post-Mortem Retrospective: a standard structure of 5 sections that the senior fills in starting from your raw inputs.

The personal growth path along the Post-Mortem axis (C5) goes from the raw-number-collection of the L1 level to the interpret-numbers-and-propose-action-items of the L2 level. The critical transition is the moment when you start to see cross-event patterns (not just numbers from a single event) and to propose improvements based on the patterns. See the next L2-C5 step mid-level blameless post-mortem. On the other L1 skills, the post-mortem ties to: L1-C1 single-BU ownership, L1-C2 direct-team stakeholders, L1-C3 operational checklists, L1-C4 internal communications. The cornerstone contains the extended framework applied to all levels.

Three readings for the post-mortem skill. The Field Guide to Understanding Human Error by Sidney Dekker (Ashgate 2014) is the definitive book on the blameless culture applied to operational contexts. Thinking Fast and Slow by Daniel Kahneman (Farrar 2011) is the book on cognitive heuristics that explains why we misinterpret data and how to correct it. The Google SRE Book - Postmortem Culture is free online and contains real examples of post-mortems in tech contexts (also applicable to events). For the Italian context, the LinkedIn posts of the people ops at Scalapay, Bending Spoons and Musixmatch often publish the backstage of their own post-mortems.

What I learned writing this