Case study · Bristol · Live since April 2026

Social Climbing's online revenue ran at 2.5× baseline for six weeks. Then they removed the widget — and lost £80 a day.

Six weeks after Social Climbing went live with DeskRelay, online revenue is running at 2.5× the previous year's baseline — £85/day pre-deploy, £220/day after. About 55% of online bookings in the window are attributable to widget conversations. The widget has fielded around 3,600 customer questions a month at an average reply time under 3 seconds, every hour of the day — including the Sunday-evening enquiries the desk used to lose to Monday morning.

Pre-widget baseline: £85/day
With widget: £220/day
Without widget (after relaunch): £140/day

The strongest piece of evidence

The £29,000-a-year hole the widget had been filling.

In May 2026, Social Climbing redesigned their website. The new site is otherwise improved — better design, better SEO — but launched without the widget. In the first nine days, online revenue dropped to £140/day, an £80/day decrease versus the widget period.

Annualised, that represents approximately £29,000 in foregone online revenue specifically attributable to the widget's absence. This drop, isolated against an otherwise improved site, confirms the original attribution analysis and demonstrates a clean causal relationship between the widget and online bookings.

Methodology, in full

Where the 55% comes from, and what it isn't.

Honest attribution beats a tidier-looking number. Here's exactly how the figure was built, what's still uncertain, and what would tighten it.

How attribution was estimated

Social Climbing's booking system does not currently expose UTM parameters or referral data to the operator. Instead, attribution was built by matching widget conversations (timestamped, identifiable by visitor session) against booking events from the same browser session within a 30-minute window. Where a conversation immediately preceded a booking and the conversation referenced the session that was booked, that booking is counted as widget-attributable.

Across the post-launch window, approximately 55% of online bookings followed this pattern. That figure is the basis of the 2.5× headline.

What this method can over-count

Some customers would have booked the same session even without the widget. The widget may have shortened the path to booking (better, faster) without being the deciding factor. The 55% includes those cases — meaning the figure overstates incremental bookings.

What this method can under-count

Some customers have a widget conversation, leave the site, return days later from a different device, and book without re-engaging the widget. The 55% does not capture those bookings. It also doesn't capture customers who learn from the widget that the gym isn't a fit and recommend it to a friend instead.

Why the May control experiment matters

The control experiment cuts through both biases. By removing the widget while keeping everything else equal (or better), it isolates the incremental contribution. The £80/day drop is the answer to "what would happen without the widget" — measured, not modelled.

What would tighten the figure further

Full UTM access on the booking system would let us trace every booking back to its first-touch source. We've requested that change from Social Climbing's booking provider. A second control window — for example, A/B'ing widget-on vs widget-off across two-week blocks — would build a confidence interval around the £80/day figure. Both are on the roadmap.

Supporting metrics, six-week window

The conversation pattern underneath the revenue line.

~3,600

Conversations per month

To be verified against Social Climbing telemetry; figure quoted from operator estimate.

Under 3s

Average reply time

Across all hours; performance does not degrade during peak.

16%

Site visitor engagement rate

Visitors who initiate or accept a conversation, against site sessions.

~33%

Booking-link click-through (estimated)

Of engaged conversations. Estimate pending UTM tracking on booking system.

The bottom line

The monthly fee was back in the account inside a week.

Standard DeskRelay pricing of $600/month (roughly £480) is recovered in 3–4 days of widget operation at Social Climbing's attribution levels. The bespoke build & delivery is custom-scoped per gym and typically a four-to-six-week engineering engagement — recovered against widget-attributable revenue within two months. Year-one return at SC: 5–9× depending on methodology and build scope.

A Widget-attributable revenue at SC: ~£135/day (≈ $170/day) £220/day with widget − £85/day baseline
B Monthly subscription: $600 (≈ £480) Recovered in 480 ÷ 135 ≈ 3.5 days of widget operation
C Bespoke build & delivery: custom-scoped per gym A 4–6 week engineering engagement; at SC's attribution levels, recovered against widget-attributable revenue within two months of widget operation
D Year-one net at SC: £35K–£49K (≈ $44K–$62K) Range reflects attribution methodology; control-experiment basis sits in the upper half

About Social Climbing

Independent climbing gym based in Bristol. Boulder-focused, with intro sessions, route classes, and weekly community evenings. Active across walk-in, members, and group bookings.

About this case study

All figures provided by Social Climbing or measured directly from their site analytics and the DeskRelay conversation log. Numbers tracked here will be updated as the booking-system integration matures. Last updated 18 May 2026.

Twenty minutes is enough to know if your numbers would look similar.

On the call we'll look at your enquiry volume, how many of those questions a trained widget could answer, and what an honest projection looks like for your business — not Social Climbing's.

Book a 20-minute call