The Scenario
You are a retention analyst. Every Thursday at 2 PM you present the weekly leadership review. The standing ask: show the January and February signup cohorts on the same retention curve and explain what changed between them.
The retention data is in Amplitude. The chart is in a Google Sheet that you update every week by hand.
The slow version every Thursday:
- Open Amplitude, navigate to Retention Analysis, configure the January cohort, set weekly intervals, export the CSV
- Go back, configure the February cohort, export a second CSV
- Open Google Sheets, paste the January curve in one column, paste February in another, manually align the week-number rows so week 1 is next to week 1
- Build the overlay line chart
- You walk into the 2 PM with a chart that took an hour to assemble.
The fast version is one prompt.
The Easy Way: One Prompt in SheetXAI
SheetXAI queries the Amplitude Retention API directly and writes both curves into the sheet in the right structure — no two-export, two-paste sequence.
Open the SheetXAI sidebar and type:
Fetch retention curves for my January 2025 and February 2025 signup cohorts from Amplitude. Use weekly intervals. Write week numbers in column A, January retention percentages in column B, and February retention percentages in column C. Add a header row.
SheetXAI pulls both cohort curves, aligns them by week number, and writes the three-column table. Your chart just needs a click to refresh.
What You Get
A clean three-column retention table, ready for charting:
- Column A — week numbers (Week 0 through Week N)
- Column B — January 2025 cohort weekly retention percentages
- Column C — February 2025 cohort weekly retention percentages
- Row 1 — headers
Both curves are aligned by week number, not by calendar date, so Week 4 for January sits next to Week 4 for February in the same row. No manual alignment required.
The line chart you built last quarter still works — the data shape is identical, only the numbers change.
What If the Data Is Not Quite Ready
Retention analysis rarely starts with two clean, perfectly comparable cohorts. SheetXAI handles the edge cases in the same prompt.
When you want 8-week and 12-week views side by side
Leadership asked for both the short and the full retention horizon this week.
Fetch 8-week and 12-week retention data for my February 2025 signup cohort from Amplitude. Write week numbers in column A, 8-week retention in column B, and 12-week retention in column C. Add headers. Mark any week beyond Week 8 in column C as "outside 8-week window" in column D.
When you want to compare users who completed onboarding versus those who skipped it
The product team is debating whether onboarding completion actually drives retention.
Pull 8-week retention data from Amplitude for users who completed onboarding versus users who skipped it. Write week numbers in column A, 'Completed Onboarding' retention in column B, 'Skipped Onboarding' in column C. Add a fourth column D calculating the difference (B minus C) for each week.
When a product change mid-month means the cohorts are not cleanly comparable
The new onboarding flow launched January 18th. Users who signed up before and after that date had different experiences.
Fetch retention curves from Amplitude for users who signed up January 1–17 and users who signed up January 18–31 separately. Write both curves side by side with week numbers in column A, pre-launch cohort in column B, post-launch cohort in column C. Add a note in cell D1 explaining the January 18 cutoff.
When leadership wants the retention curves, the week-over-week delta, and a written takeaway all in one output
The Thursday review requires a chart plus a one-paragraph summary of the most significant retention difference between cohorts.
Pull weekly retention for January and February 2025 signup cohorts from Amplitude. Write week numbers in column A, January retention in column B, February retention in column C. Calculate the difference (February minus January) in column D and flag any week where the difference exceeds 5 percentage points with "NOTABLE" in column E. Then write a two-sentence summary in cell G1 identifying which week shows the largest difference and what it implies about the February cohort.
The pattern: instead of pulling two exports, reconciling them, and drafting a summary separately, you ask for all three in one prompt. The summary references the numbers SheetXAI just wrote.
Try It
Get the 7-day free trial of SheetXAI and open any Google Sheet, then ask it to pull retention curves for any two Amplitude cohorts with whatever interval and date range the review requires. The Amplitude integration is included in every plan. For related workflows, see how to pull funnel conversion data for quarterly reviews or the Amplitude in Google Sheets overview.