Data Studio's comparison feature looks straightforward until you run a year-over-year report and a stakeholder asks why January traffic is up 8% when nothing changed. The culprit is almost always the weekday shift: Previous year compares 1–31 Jan 2026 to 1–31 Jan 2025, but those weeks start on different days of the week. For traffic-heavy sites, weekend vs. weekday composition alone can swing figures by 10–15%, creating calendar noise that reads like a real trend. Picking the right comparison mode — or using the manual workaround — eliminates that problem before it starts.
The three built-in comparison modes
Data Studio exposes three comparison modes through the date range control. Here is what each one actually calculates:
Previous period — shifts your selected window back by the same number of days. If your report shows 1–31 Jan 2026 (31 days), the comparison window becomes 1–31 Dec 2025 (the preceding 31 days). This is accurate for month-over-month and week-over-week work because the two windows are the same length and sit immediately adjacent. Weekday composition is not aligned, but the difference is small enough over a full month that it rarely distorts the figures materially.
Previous year — maps your selected dates back exactly twelve months. 1–31 Jan 2026 compares to 1–31 Jan 2025. The window length is the same, but weekday composition is not guaranteed to match, because calendar dates fall on different weekdays year to year. This is the mode most people reach for YoY comparisons — and it works fine for monthly and quarterly rollups. The problem appears when you're analysing weekly patterns or when your business has a strong day-of-week signal (e-commerce, news, entertainment).
Advanced — you specify the comparison start and end dates manually. Data Studio does not calculate anything for you; you enter the exact dates you want to compare against. This is the escape hatch for weekday-aligned comparisons, irregular periods, or any situation where the automatic modes produce misleading numbers.
Adding a comparison date range control
The comparison date range is set on the date range control, not on individual charts. Add one from the toolbar:
- In edit mode, click Add a control in the top toolbar and select Date range control.
- Place the control on the canvas and select it.
- In the right-hand Setup panel, find Comparison date range and open the dropdown.
- Choose Previous period, Previous year, or Advanced (for a custom range).
- Right-click the control and select Make report-level so the comparison applies to every page at once — see the date control setup guide for full details on scoping.
Once the comparison is active, time series charts show a second line and tables show delta values with up/down indicators automatically — no per-chart configuration needed.
When to use which mode
| Goal | Recommended mode |
|---|---|
| Week-over-week trend | Previous period |
| Month-over-month | Previous period |
| Quarter-over-quarter | Previous period |
| YoY (calendar aligned) | Previous year |
| YoY (weekday aligned) | Advanced |
| Ad-hoc or irregular comparison | Advanced |
The key distinction: use Previous period when continuity of days matters more than calendar alignment, and use Previous year when you want the same calendar dates from twelve months ago. Switch to Advanced the moment either automatic mode is giving you numbers that don't reflect what actually happened.
The weekday alignment problem
ISO 8601 defines weeks as starting on Monday. ISO W1 of a given year is the week containing the year's first Thursday — which means W1 can start in late December of the previous calendar year.
Concrete example:
- W1 2026: Mon 29 Dec 2025 – Sun 4 Jan 2026
- W1 2025: Mon 30 Dec 2024 – Sun 5 Jan 2025
If you set your date range to W1 2026 (29 Dec 2025 – 4 Jan 2026) and use Previous year, Data Studio compares it to 29 Dec 2024 – 4 Jan 2025 — a completely different week in the ISO calendar. That window starts on 29 Dec 2024, which is a Sunday — the last day of W52 2024 — before continuing into W1 2025 (30 Dec 2024 – 4 Jan 2025): a mixed-week comparison that misaligns the two periods by one day, not the clean W1-vs-W1 comparison you intended.
To compare W1 2026 against W1 2025 correctly, use Advanced mode:
- Set your primary date range to 29 Dec 2025 – 4 Jan 2026 (W1 2026).
- Open the Comparison date range dropdown and select Advanced.
- Enter the comparison start date as 30 Dec 2024 and the end date as 5 Jan 2025 (W1 2025).
Data Studio will now query both windows independently and display the delta based on equivalent weekday periods. The same logic applies to any rolling weekly report: always verify which ISO week the automatic modes are actually comparing before sharing results with stakeholders.
If your use case goes further — for example, comparing two data sources that each need their own independently controlled date window in the same chart — that requires data blending rather than a comparison date range. See the guide to blending data sources in Data Studio for how joins and date windows interact in that scenario.
Quick recap
- Data Studio's built-in comparison modes are Previous period, Previous year, and Advanced — the third lets you specify the exact comparison window manually.
- Previous period is reliable for MoM and WoW because the two windows are the same length and sit back-to-back.
- Previous year compares identical calendar dates twelve months apart — correct for monthly and quarterly rollups, but problematic for weekday-sensitive analysis.
- The weekday shift is the most common source of misleading YoY figures: W1 2026 (29 Dec 2025 – 4 Jan 2026) does not align calendar-date-for-calendar-date with W1 2025 (30 Dec 2024 – 5 Jan 2025).
- Use Advanced mode and enter the comparison dates manually whenever weekday alignment matters.
- Set the comparison on a report-level date range control so it applies across all pages in one step.
- Getting all of this right means your comparison numbers reflect what actually happened in the business — not an artefact of how the calendar falls differently year to year.