Churn deep dive #3

1Reasons · by size

Smaller merchants churn silently; larger ones over features and surcharge

Multi-select; shown as % of each size group's churns. Bands by expected annual TPV at signing.

• Silent churn is most prominent among smaller merchants — potentially business closures.
• Overpromised / missing features is highest among higher-TPV merchants — as expected, larger merchants have higher demands. This is mostly Pay by Link / MOTO.
• Surcharge is the biggest issue among higher-TPV merchants — though it is rarely the sole reason.
• Technical issues are higher among smaller merchants — mostly battery issues. One hypothesis (Ale): larger merchants keep set hours while smaller ones trade longer and wear the hardware down. Not verified.

Deep dive · overpromised / missing feature, £150k+ merchants

2Reasons · by industry

Hospitality churns to competitors (mostly Dojo), surcharge & payout frequency; Retail to surcharge & missing features

Limited to the three largest industries by churn volume. Multi-select; shown as % of each industry's churns.

Every percentage here is the share of that industry’s own churns — not of total churn. Rows are multi-select, so they sum to more than 100%.

What stands out by industry

• Hospitality is the most competitive segment — competitor is its top reason at 23%, as expected. Dojo is the named competitor in 8 of the 15 competitor churns.
• Payout frequency is joint-highest in Hospitality at 20% — the most payout-driven churn of any industry.
• Contract length is also elevated in Hospitality, at 12%.
• Services skews to silent churn (21%, the highest of the three) and technical issues (15%).

3Lifespan

Almost half of FS - Canvas churns between 8–30 days, without a common churn reason

On the master's own lifespan (activation → last transaction). All three channels are deep-infancy leavers, with medians of 20–25 days.

The 8–30 day FS - Canvas cluster

4Value at risk · expected TPV

Facebook churns fewer merchants but larger TPV merchants

MCC - Facebook, while churning fewer merchants, churns a similar amount of TPV — at a higher average per merchant. That is mostly driven by 5 merchants worth £8.8M in expected TPV between them, with no common theme in their churn reasons.

In FS - Canvas, the £75–150k band churns at 4.7% (13 merchants) — with no common theme in the reasons: a mix of technical, fraud, silent, payout, surcharge, contract and held-payout.

5Ramp speed

FS - Canvas: churned merchants ramped fastest, while being the smallest on TPV

How quickly churned merchants reached £2,500 in processed volume, and the expected TPV they represented.

6Value at risk · realised TPV

Merchants churn before they ramp — Facebook realises the least

Actual volume processed by churned merchants before leaving, and how little of their expected annual TPV they realised.

MCC - Facebook has the lowest expected-TPV realisation of the three channels, at 2.8%. Its few largest merchants — which carry almost all of the channel’s expected TPV — barely processed anything before churning, consistent with the earlier read that Facebook churns merchants at a higher average expected TPV. One possibility is that these larger merchants purchase ahead of opening, so take longer to start processing.

33% of Facebook churns processed 5–10% of their expected TPV — and 3 of those 9 were fraud, in line with Facebook being the main driver of fraud across the three channels.

FS - Canvas, while having the lowest average TPV churned, had the shortest number of days to activation among its churned merchants. This is in line with its highest TPV to date among the three largest channels. Potential hypothesis: fewer FS - Canvas churned merchants had other card machines — so although their average TPV was smaller, they were putting more TPV through our terminals. Not verified.