How to Read the statpx Weekly Retention Cohort Table

Total visitor counts can climb every month while the site quietly gets worse at keeping anyone around — new acquisition simply papers over the leak. The statpx Retention page removes that blind spot with a color-coded weekly cohort table: each row is a group of visitors who first arrived in the same week, and each column shows what share of that group came back in the weeks that followed. This guide covers how the grid is built, how to read the color coding, and what to do once you spot a pattern.

What the Retention Page Shows

Open retention.php from the sidebar of any site and you'll see a table with cohort weeks running down the left side (most recent at the top) and week numbers — Week 0 through Week 7 — running across the top. Week 0 is always 100%: it's the full cohort on its first day. Each cell after that shows the percentage of that same cohort's original visitors who returned during that specific week, based on distinct visitor IDs matched back to their first-seen week.

Because the grid is limited to the last eight cohort weeks and caps at Week 7, you always see a manageable, current window rather than an ever-growing table. Recent cohorts naturally have empty cells in the later columns — they simply haven't had time to reach those weeks yet.

Reading the Color-Coded Cells

Each cell is shaded by its percentage value, so a whole week's health is visible without reading a single number: darker or more saturated cells mean a higher share of that cohort returned, while pale cells mean most of that group never came back. Scanning down any single column tells you whether that week-N return rate is improving or declining across cohorts — a column getting visibly darker as you move toward the newest cohorts means something you changed recently is working.

• A strong Week 1 column across every cohort means your onboarding or first-week experience reliably earns a second visit.
• Fading color past Week 2–3 for every cohort means you hook people briefly but don't give them a reason to stick around long-term.
• One cohort row noticeably darker than its neighbors usually lines up with a specific acquisition week — check what campaign, content push, or traffic source was active that week, since it's clearly sending higher-quality visitors.

Turning Acquisition Weeks Into Loyal Visitors

The real value of the grid is spotting which specific acquisition weeks produce visitors who keep returning, then asking why. If a cohort week correlates with a particular landing page, referral source, or content release, that's a repeatable playbook — double down on whatever attracted that cohort. Conversely, if a cohort week produced a burst of visitors who all vanish by Week 1, that traffic source or campaign is inflating your totals without building an audience, and its cost isn't paying off in the way the visit count alone suggests. Pair the weekly grid here with a full cohort analysis when you need day-level or behavioral cohort breakdowns rather than the standard weekly acquisition view.

The Bottom Line

The Retention page turns raw visitor totals into the one question that actually matters: are the people who come back growing, or is growth just a bigger front door with the same leak in the back? Check it weekly, read down each column for the trend, and use one strong or weak cohort row as a starting point for investigating what happened that week. Combine it with returning vs new visitor data for a complete picture of whether your audience is compounding or just churning through new arrivals.