Back to all articlesReports and dashboards

Choose Between Event-Date and Entry-Date Reports

Alactic Editorial10 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

In a hypothetical reporting example, six completed visits appear in today's report, while the previous two days show only one each. Before treating that as a surge in activity, check which date the report groups by. Some of today's entries may describe work completed earlier.

Use event date to answer when the work happened. Use entry date to answer when its record entered the reporting system. When late recording could change a decision, pair the activity view with a recording-delay summary rather than asking one date field to explain both.

The distinction matters even when every record is accurate. A report can count the right records and still tell the wrong operational story because its date basis does not match the question.

1. Start with the operational question

Event date, also called occurrence date, represents when the activity being measured happened. For completed visits, that would normally be the completion date. Entry date represents when the record was first recorded in the relevant system. Neither should be confused with the date a record was last edited.

Choose the event before choosing its date. A scheduled visit, a completed visit, and the recording of that completion are different events. A report titled “Completed visits” should not silently group records by scheduled date because that field is convenient.

  • “How many visits were completed each day?” Group qualifying completed visits by completion date.
  • “How many completed-visit records were entered each day?” Group those records by entry date.
  • “How much earlier work was recorded today?” Select records entered today, then compare their occurrence and entry dates.

The third question needs both dates. An entry-date total alone cannot distinguish same-day activity from older work newly recorded.

Also define what “entry” means. If records move between systems, a destination's creation timestamp may describe arrival there, not the original recording time. Label the timestamp you actually have.

2. Work through the same records both ways

Hypothetical worked example: A fictional team records eight completed visits across 6–8 April 2026. The reporting cutoff is the end of 8 April in one agreed reporting timezone. Each row represents one completed visit, with no duplicates. “Delay” means calendar days between occurrence and entry, not elapsed hours.

The following displays contain exactly the same fictional records. Only their grouping changes.

View A: grouped by occurrence date

Record  Visit          Occurred   Entered    Delay
V-101   Cedar site     6 Apr      6 Apr      0 days
V-102   Maple site     6 Apr      8 Apr      2 days
V-103   Birch site     6 Apr      8 Apr      2 days
        6 Apr total: 3 completed visits

V-104   Elm site       7 Apr      7 Apr      0 days
V-105   Willow site    7 Apr      8 Apr      1 day
V-106   Ash site       7 Apr      8 Apr      1 day
        7 Apr total: 3 completed visits

V-107   Oak site       8 Apr      8 Apr      0 days
V-108   Pine site      8 Apr      8 Apr      0 days
        8 Apr total: 2 completed visits

Occurrence totals: 6 Apr = 3 | 7 Apr = 3 | 8 Apr = 2
Total: 8 completed visits

This view answers the activity-timing question: three visits happened on each of the first two days, and two happened on the third. It does not, by itself, reveal when that activity became visible to the reporting team.

View B: grouped by entry date

Record  Visit          Occurred   Entered    Delay
V-101   Cedar site     6 Apr      6 Apr      0 days
        6 Apr total: 1 record entered

V-104   Elm site       7 Apr      7 Apr      0 days
        7 Apr total: 1 record entered

V-102   Maple site     6 Apr      8 Apr      2 days
V-103   Birch site     6 Apr      8 Apr      2 days
V-105   Willow site    7 Apr      8 Apr      1 day
V-106   Ash site       7 Apr      8 Apr      1 day
V-107   Oak site       8 Apr      8 Apr      0 days
V-108   Pine site      8 Apr      8 Apr      0 days
        8 Apr total: 6 records entered

Entry totals: 6 Apr = 1 | 7 Apr = 1 | 8 Apr = 6
Total: 8 records entered

The entry-date view correctly shows a recording surge on 8 April. Four of its six entries describe earlier work; two describe work completed that day. Calling all six “visits completed today” would turn a recording pattern into an unsupported activity claim.

Both views total eight because they contain the same fixed record set. Their daily distributions differ because they answer different questions.

3. Choose the primary view with a decision table

Use this decision table to select the basis for the main total. The supporting date remains useful even when it is not the primary grouping.

Operational questionPrimary viewSupporting detail
When did visits happen?Occurrence dateEntry cutoff and recording-delay summary
When were records entered?Entry dateOccurrence dates of those entries
Was 8 April an activity surge?Occurrence dateCompare three, three, and two visits in the hypothetical example
Why did 8 April show six entries?Entry dateFour earlier visits and two same-day visits
How much delayed recording is visible?Both datesCount and age of entries recorded after occurrence
What was visible at an earlier cutoff?Historical as-of viewSaved snapshot or suitable record history

The last question is different from ordinary date grouping. A current occurrence-date report distributes known records to when their events happened. It does not necessarily reproduce what a manager could see at an earlier cutoff.

In the fictional set, only V-101 was recorded by the end of 6 April. In a live dataset, later edits or deletions can also affect historical reconstruction. Without suitable history, do not present today's data as yesterday's knowledge.

4. Separate date basis, period, and cutoff

A clear report states which date groups the records, which period is selected, and how current the data is. A date range alone is not enough.

For the hypothetical example, an occurrence-view heading could read: “Completed visits by occurrence date, 6–8 April 2026; records available through end of 8 April.” State the reporting timezone alongside it. The entry-view heading should instead say “Completed-visit records by entry date.”

Keep the population consistent when diagnosing differences. An occurrence report selecting work completed during a period and an entry report selecting records entered during that period may legitimately have different totals. Work completed before the period or recorded after it can cross those boundaries.

Start reconciliation with a fixed set of record identifiers, as the example does. Once those totals agree, restore the operational filters and explain any boundary differences.

  • Timezone: derive reporting dates consistently from timestamps. An event near midnight can fall on different calendar dates in different zones.
  • Field meaning: verify that the entry field is not a last-modified timestamp. A correction should not silently become a new recording event.
  • Counting unit: count visits consistently. A report of record changes is not necessarily a report of distinct completed visits.

5. Show recording delay beside activity

If the main question is when work happened, keep occurrence date as the main chart's basis. Add a compact delay summary to explain how late recording affects interpretation.

For the hypothetical record set, the calculation is:

  • Same-day recording: V-101, V-104, V-107, and V-108; four of eight records, or 50%.
  • Later-date recording: V-102, V-103, V-105, and V-106; four of eight records, or 50%.
  • Delay distribution: four records at zero days, two at one day, and two at two days.
  • Earlier work among 8 April entries: four of the six entries.

These measures describe available records. They cannot count completed work that has not yet been recorded. A high same-day share among visible records is not proof that the activity report is complete.

Calendar-day delay is also coarse. Recording shortly after midnight may count as a one-day delay despite little elapsed time. If that distinction matters, use timestamps to calculate elapsed time and label the unit explicitly.

Describe the result neutrally as “entered after occurrence date.” The timing difference alone does not establish its cause or demonstrate poor performance.

6. Explain revisions and uncertain dates

An occurrence-date report can revise yesterday's total when today's entries describe yesterday's work. In the hypothetical example, the known occurrence total for 6 April rises from one at that day's cutoff to three by the end of 8 April. The work did not move; the available record set expanded.

For a current activity view, allow those totals to update and show the cutoff. For recurring management comparisons, retain dated snapshots where the reporting workflow supports them. If recent periods remain open to additional records, make that qualification visible.

Do not invent a universal waiting period after which activity is complete. Review local recording delays before choosing a review point, and describe that point as an operating convention rather than proof that no more records will arrive.

Missing occurrence dates need a separate treatment. Keep such records in an “occurrence date unknown” group rather than silently substituting entry date. Otherwise, uncertainty becomes apparent same-day activity. Show the unknown count beside the dated total.

If a corrected occurrence date changes an earlier total, explain that revision using whatever reliable history is available. An overwritten field alone is not a full audit trail.

7. Correct a misleading activity report

If an entry-date chart has already been described as an activity chart, correct both the label and the interpretation. Relabeling alone will not address a decision made from the earlier story.

  1. Identify the original record set and reporting cutoff.
  2. Regroup those records by occurrence date where available.
  3. Separate same-day entries from entries describing earlier activity.
  4. Check whether the apparent activity surge remains.
  5. Publish the corrected interpretation with the revised figures.

For the fictional example, the correction would be: “Six records were entered on 8 April, but only two visits occurred that day. Four entries related to 6–7 April. The available occurrence counts are three, three, and two.”

If the fixed-set totals do not reconcile, investigate filters, missing dates, duplicate rows, and counting units before interpreting the trend. A date-basis explanation should account for redistribution, not conceal missing or extra records.

8. Reuse this filled report-definition card

The following filled artifact captures the choices behind the hypothetical report. Copy its field labels for another metric, then replace the values. It is a reporting specification, not a claim that any particular product implements these controls.

REPORT DEFINITION — HYPOTHETICAL EXAMPLE

Question: When were completed visits carried out?
Counting unit: One distinct completed visit
Primary date: Visit occurrence date
Secondary date: First record entry date
Period: 6–8 April 2026, inclusive
Reporting timezone: Etc/UTC
Data cutoff: End of 8 April 2026 in Etc/UTC
Fixed comparison set: V-101 through V-108

Primary result:
6 Apr: 3 visits | 7 Apr: 3 visits | 8 Apr: 2 visits

Entry-date comparison:
6 Apr: 1 entry | 7 Apr: 1 entry | 8 Apr: 6 entries

Delay definition: Entry date minus occurrence date,
in calendar days in the reporting timezone
Delay result: 4 same-day; 2 one-day; 2 two-day
Unknown occurrence dates: 0 in this example
Reconciliation: 8 distinct visits in each view

Interpretation:
8 April was a recording surge, not an activity surge
within this fixed set. Four entries describe earlier work.

Revision rule:
Update occurrence totals as records arrive; show cutoff.
Keep dated snapshots for historical comparisons.

Review check:
Confirm field meanings, counting unit, and filters
before distributing the report.

Test one existing report with this card before changing the wider dashboard. The observable result should be a reconciled record set, explainable daily differences, and a heading that makes the date basis and cutoff clear. That lets readers distinguish work happening from records arriving.

Reports and dashboardsData qualityReporting practices

Put the next conversation to work.

Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.

Explore Communications