Value Packs / Retail / Conversion & Checkout Funnel Radar / Wisdom
Which parts of the site are converting worst against what we expect of them today, and how much traffic is behind each one?
NWhich surfaces are off their own normal
IN A NUTSHELL
Three combinations are outside the band we expect of them today. Mobile app on paid search is furthest — 1.92% against a 2.71% floor, a 0.79-point gap — on 192K sessions today. Desktop on email is 0.34 points low on 30K, and mobile web on paid social 0.27 points low on 127K. Two surfaces are running above theirs. The app on paid search alone is about 1,500 orders short of what that many sessions usually turn into.
Conversion Rate today against the band we expect of each surface
Every surface is read against its own floor for today, not against the others — each floor is built from the same weekday in the three weeks before, which is the only place a week enters this notebook.
Ask a follow-up
How has Checkout Rate on the mobile app moved day by day against what we expect of it?
NIt broke overnight, and it is still broken
IN A NUTSHELL
Checkout Rate on the mobile app sat between 66.8% and 69.1% every day from 1 May to 19 May, inside what we expect of it. Yesterday it fell to 61.4% and today it is 61.0% — two days below the range, and today is lower than yesterday. It dropped 6.8 points overnight, nearly four times the biggest move any ordinary day made in the three weeks before.
Checkout Rate by day against what we expect · mobile app
Three weeks of ordinary days, then two below the range — the range dips each weekend because it is built from the same weekday in the three weeks before, which is the only place a week enters this notebook.
The biggest move on any ordinary day in three weeks was 1.8 points. The night before yesterday was 6.8. Yesterday closed at 61.4% and today at 61.0%, against a range that expected between 67.0% and 70.0% on a Wednesday.
Wisdom noticed
Ask a follow-up
Is the drop in one step of the checkout or all of them?
NOne step, and only one
IN A NUTSHELL
One step. Against a usual day, sessions to bag is unchanged at 8.9% versus 8.8%, and bag to checkout is unchanged at 46.2% versus 45.9%. Checkout to order fell from 68.1% to 61.0% — 7.1 points at the last step while the two above it held. Multiply the three together and 2.46% of app sessions became orders today against a usual 2.80%, which is 12% fewer orders on the same traffic.
Each step of the checkout, today against a usual day
| Step | A usual day | Today | Change | Share of the loss |
|---|---|---|---|---|
| Sessions → bag | 8.9% | 8.8% | −0.1 pp | No change |
| Bag → checkout | 46.2% | 45.9% | −0.3 pp | No change |
| Checkout → order | 68.1% | 61.0% | −7.1 pp | The whole of it |
Two rows barely move and the third does all of it — a usual day here is the nineteen days before this started, and the steps above checkout change by less than half a point between them.
Traffic went up, not down — sessions 4% and add-to-bag 3% above a usual day — so more people reached checkout today than usually do, and fewer got through it. The two steps above it moved by under half a point.
Ask a follow-up
Put the checkout report together.
NThe report lists every step of the checkout that is running behind, with the two steps before it so you can see where the problem starts.
checkout_steps_21_may.xlsx
One row per step, grouped by the team that can change it. Mobile app, today against the nineteen days before.
22 KB · 7 rows · 2 new, 4 persisting, 1 cleared
Ask a follow-up