Case study: Panda's Greater Toronto network

Greater Toronto is the market we built the platform in. It's directly operated by Panda — same software, same payment rail, same dashboard you'd license. This is the live proof that the model works in a North American market, and where the operating playbook for new operators was forged.

The numbers

Across the Toronto network, since launch in early 2024:

Metric	Value
Charging hours delivered to guests	63,000+
Guest rentals processed	20,000+
Months in continuous operation	27+
Average rental duration	3+ hours
Equivalent of continuous device-charging	7+ years
Settlement currency	CAD

The headline metric we lead with is hours delivered, not station count — because hours-of-charging is the unit that matters to a venue's guest experience and to an operator's bottom line. Every minute of the 3.7+ million we've delivered ran on the same platform and the same payment rail you'd be licensing.

What the network looks like

The network is concentrated across the Greater Toronto Area, with density in the neighborhoods where high-dwell hospitality is densest:

• Hot pot corridors (Markham, North York, Scarborough — multi-block density of hot pot and Chinese hospitality)
• KTV and entertainment venues (Markham, Thornhill, Scarborough) — long bookings, peak weekend nights
• Hotel & business-district lobbies (downtown Toronto, Niagara Falls)
• Premium spa and clinic partners — where the Amenity model fits naturally
• Casual dining and bar / lounge venues

This wasn't deployed all at once. The first stations were trial deployments — single venues, single neighborhoods. We placed, observed, kept the winners, pulled the losers, and built density around the categories that produced.

What we learned in Toronto that's now in the playbook

1. Hot pot is the strongest single category in NA

Our hot pot stations consistently outproduce other categories by a meaningful multiple. The combination of long waits, multi-hour meals, phone-heavy group dining, and the customer base's familiarity with scan-to-rent makes it the cleanest fit.

This finding shaped the operator playbook: when a new operator is selecting their launch batch, we point them at hot pot first if the category exists in their city.

2. The 3+ hour average rental tells you something

A 3-hour average rental duration means the customer is keeping the powerbank with them through their entire visit — not grabbing it for a quick 15-minute top-up. That's the venue-dwell-time thesis being validated in the data: when you place stations in venues where guests actually stay, the product becomes part of the visit rather than a fleeting convenience.

This is the kind of insight that flows back into venue selection. We coach operators to evaluate venue candidates on dwell time first; the per-rental duration in their early data confirms (or contradicts) the placement call.

3. Premium spas justify the Amenity model

Stations at premium spa partners run on our Amenity tier — venue pays a flat fee, charging is free for guests, the screen is fully on-brand. This is a different economic shape than revenue-share, and we proved out the model in Toronto with multi-station rollouts at flagship venues.

The case study we published on the venue site (Go Place) is from this network — a single-station pilot that expanded to seven stations across the property.

4. Density compounds, sparseness doesn't

Stations clustered in the same neighborhood produce more reliably than scattered placements. Customers learn that "there's always a Panda station nearby" — confidence in the network increases rental conversion.

We coach new operators to build density in 2–3 neighborhoods before spreading citywide.

5. Payment localization on day one

We integrated Canadian card rails and mobile wallets, with the customer-facing flow in EN + 中文 (the GTA has a significant Mandarin/Cantonese-speaking customer base). The bilingual UI noticeably increases conversion in the right neighborhoods.

This is the kind of localization a generic factory app doesn't do; it's the kind of thing the Toronto experience taught us to build into the platform itself.

What a new operator inherits from this network

When you license the platform, you don't just get the software. You get:

• Calibrations for what venue categories produce at what level, drawn from the 20,000+ rentals above
• Pricing defaults by category that we've tuned in real operation
• Venue conversation playbook — what works, what doesn't, in real venue outreach
• Hardware spec recommendations by venue type
• Operational rhythm — when to swap powerbanks, how to handle station issues, how to read the dashboard
• All the platform features we built because we hit a real operational problem in Toronto first

This is the value of licensing a platform from someone who runs networks themselves. The factory will sell you a box. Panda hands you a system that's already been load-tested in a real market.

What's different about your city

We're not pretending Toronto and your city are identical. Specifically:

• The category mix will differ (your city might have less hot pot density, more KTV, more banquet halls)
• The price point will differ (markets have different willingness-to-pay)
• The payment rails will differ (your country, your cards, your wallets)

But the structure — venue selection driven by dwell time × phone usage, density-first deployment, dashboard-driven iteration — transfers without modification. That's what an operator inherits.

The honest version

Toronto isn't a closed case study. We continue operating it directly. We use the network as the test environment for new platform features, the calibration source for what new operators should expect, and the proof point that the platform handles real volume.

If a new operator wants to see the dashboard live, see the rental flow live, or talk to us about a specific category's per-month numbers in Toronto for comparison to their own market projections — we can do that on the demo.

Apply to become an operator and we'll walk through the Toronto network's data live.