← Reading listSEPARATE WHAT ACCUMULATED FROM WHAT HAPPENED THIS PERIOD
Text

Practical framework / 5 MIN READ

Separate what accumulated from what happened this period

Use running totals for progress and period values for pace.

Published by SNS Media Tech

AI-assisted research and analysis for the SNS Creative Fieldbook

Sources checked 2026-11-06

A cumulative total can keep rising while current performance weakens. The right view depends on whether the decision asks how much has been achieved so far or how quickly new results are arriving now.

What you will leave with: Create a paired progress-and-pace report with a documented reset, window and reconciliation.

Use this when

  • A growing lifetime total is used to claim accelerating performance.
  • A campaign’s daily pace is hidden inside a cumulative chart.

This lesson does not interpret cumulative unique users as additive daily users or provide a forecasting model.

Choose the clock and reset

Define whether the running total starts at campaign launch, the calendar year or another meaningful boundary. State the unit and whether values can be revised downward through refunds or corrections. A cumulative net measure may fall, unlike a simple accumulation of nonnegative counts.

Choose the period view needed for action, such as daily orders or weekly qualified enquiries. Use comparable complete periods and explain partial ones. A campaign-to-date total is useful for progress toward a goal, while the period series helps decide whether delivery or operations need attention.

Verify the running calculation

PostgreSQL documents ordered window calculations and the importance of their frame. In any reporting tool, check the ordering, partition and reset rules behind a running total. A valid query can accumulate across the wrong campaigns or restart at an unintended filter boundary.

Reconcile a short fixture by hand. For an additive measure with a consistent definition, differences between successive cumulative values should recover period increments, allowing for documented revisions. Check duplicate dates and missing periods before interpreting the slope. Do not assume a missing day means zero activity.

Sources for this sectionPostgreSQL: Window functions →

Treat unique counts and rates carefully

Daily unique customers can overlap across days, so adding them does not produce unique campaign customers. A cumulative unique measure requires deduplication over its whole window. Differences in cumulative unique counts indicate newly added identities under that measurement method, not necessarily all customers active in the period.

Rates also need their own cumulative numerator and denominator. Summing daily conversion rates is meaningless, and an unweighted average can misrepresent the overall rate. Keep the underlying counts available and calculate the measure that matches the defined window.

Pair progress with current pace

Show the cumulative view for progress and a period view for recent flow. If a target line is included, explain whether it assumes an even pace or a planned changing schedule. A gap from a straight line can be intentional when activity is concentrated around an event.

Write the takeaway using both views. The total increased, but weekly additions slowed, is often more useful than continued growth. Avoid projecting a future total from a brief favorable slope without a suitable forecast and uncertainty. Preserve report revisions so a change in accumulation is not mistaken for new activity.

Make the reasoning concrete.

A hypothetical campaign adds 40, 30 and 20 qualified enquiries over three weeks. Its cumulative totals are 40, 70 and 90. The total rises every week, while the weekly additions decline by ten each time.

Show only accumulation

Opening: Call the rising line improving performance.

Hypothesis: The campaign has produced more total enquiries over time.

Risk: The current pace is slowing.

Show progress and flow

Opening: Display 90 to date alongside weekly additions of 40, 30 and 20.

Hypothesis: The team sees both achievement and the recent slowdown.

Risk: The pattern alone does not explain why pace changed.

Investigate the slowdown using relevant delivery and operating evidence. Keep the cumulative total for goal progress, but do not use its upward direction as proof of acceleration.

Take the questions into your next brief.

Use these prompts in your own notes. Nothing is submitted from this page.

  1. Reset

    Where does the cumulative window begin?

  2. Measure

    Is it additive, unique or a ratio?

  3. Calculation

    How are ordering, partitions and missing periods handled?

  4. Pace

    Which complete-period view supports action?

  5. Interpretation

    What does each view establish and leave unexplained?

Watch for the wrong conclusion.

  • Calling any rising running total acceleration.
  • Adding daily unique people into a campaign unique total.
  • Summing or casually averaging rates.
  • Comparing a partial current period with complete past periods.

Your next action

Add a period-flow companion to one cumulative campaign chart and reconcile both against a short hand-calculated fixture.

Sources & scope

Frameworks, worksheets and hypothetical examples are original editorial synthesis. Linked primary sources support only the stated concepts. Source checking is not an independent audit of effectiveness. No named human author or reviewer is claimed.

  • PostgreSQL: Window functions →Primary technical documentation · Supports: Documents ordered window calculations and frame behavior. The cumulative-versus-period marketing interpretation and hypothetical arithmetic are original editorial analysis.