A claim about Quenchly

In year three the company raises prices, and you can watch it land.

Three new plans rows at the increase date; annual share of activations spiking the month before; requested cancels running at least 1.2 times the trailing quarter's count for the quarter that follows.

holds on the current window

as of moments ago

Versioned price book

SELECT id, name, seat_price_usd_month, seat_price_usd_year
FROM plans
ORDER BY name, seat_price_usd_month;

id	name	seat_price_usd_month	seat_price_usd_year
business	business	199	1990
business-v2	business	219	2190
pro	pro	99	990
pro-v2	pro	109	1090
starter	starter	49	490
starter-v2	starter	54	540

holds on the current window

Activations, cancellations, and migrations

SELECT strftime('%Y-%m', e.at)                                           AS month,
       SUM(e.kind = 'activate')                                          AS activations,
       ROUND(100.0 * SUM(e.kind = 'activate' AND s.billing = 'annual')
                   / NULLIF(SUM(e.kind = 'activate'), 0))                AS annual_pct,
       SUM(e.kind = 'cancel' AND e.reason = 'requested')                 AS requested_cancels,
       SUM(e.kind = 'plan_change' AND e.plan_id LIKE '%-v2')             AS migrations_to_new_price
FROM subscription_events e
JOIN subscriptions s ON s.id = e.subscription_id
GROUP BY 1
ORDER BY 1;

View data table

month	activations	annual_pct	requested_cancels	migrations_to_new_price
2023-01	40	40	0	0
2023-02	36	36	1	0
2023-03	32	47	2	0
2023-04	35	49	4	0
2023-05	53	42	8	0
2023-06	59	47	3	0
2023-07	69	46	3	0
2023-08	42	55	5	0
2023-09	48	40	9	0
2023-10	57	23	7	0
2023-11	55	53	12	0
2023-12	44	36	11	0
2024-01	35	43	13	0
2024-02	44	48	8	0
2024-03	42	48	9	0
2024-04	43	40	13	0
2024-05	48	42	13	0
2024-06	39	44	12	0
2024-07	30	50	15	0
2024-08	14	43	13	0
2024-09	22	36	12	0
2024-10	38	37	4	0
2024-11	46	46	13	0
2024-12	32	78	9	0
2025-01	16	56	15	355
2025-02	29	41	17	33
2025-03	58	48	16	36
2025-04	47	47	15	36
2025-05	58	41	10	41
2025-06	51	39	12	42
2025-07	53	40	14	43
2025-08	40	53	13	29
2025-09	78	41	10	28
2025-10	116	42	14	30
2025-11	94	47	18	49
2025-12	60	40	21	39
2026-01	57	51	25	10
2026-02	52	38	16	7
2026-03	70	46	18	6
2026-04	65	35	26	15
2026-05	75	48	14	7
2026-06	80	56	16	7
2026-07	78	45	30	10
2026-08	36	53	22	7
2026-09	58	50	22	9
2026-10	48	44	31	13
2026-11	9	67	2	1

holds on the current window

Demo data usually describes a business where nothing ever happened. Quenchly’s business made a decision: two years in, it raised list prices by about ten percent. The price book gained new rows with the same plan names, the way a billing provider creates new price objects instead of editing old ones. Monthly subscribers moved to the new price at their next renewal; annual subscribers kept the old price until their anniversary. Customers who saw the announcement locked in annual plans the month before. Some left in the quarter after.

What you should see

The plans table has six rows: the three originals, plus the same three names again with a -v2 suffix and prices about ten percent higher. In the monthly table, annual_pct jumps in one month and returns to its usual mid-forties the month after. The next month migrations_to_new_price spikes as monthly renewals move over, then tails off across the following year as annual renewals come due. requested_cancels runs above its trailing level for about a quarter after the increase and settles back.

Run it yourself

SQLite one file, ready to query CSV tar.gz, one file per table