Ilya Drozdov from Dwelly on AI in Lettings and Building High-Margin Businesses

Ex-Uber manager aims to triple letting agency margins with AI

Dwelly is betting that UK letting agencies can triple their margins if AI handles the communication-heavy work and humans stay in the loop. Agency work is labour intensive and fragmented, so the company buys or partners with existing agencies instead of trying to disrupt them. Ilya Drozdov brings marketplace scaling lessons from running Uber in Russia.

Takeaways

• Dwelly targets tripling letting agency margins. Traditional small agents earn about 10 to 15% EBITDA, usually near 12%. Dwelly targets 35 to 40%. Letting is labour intensive, and most agent work is communication that language models can digitize.
• Growth comes from acquisitions and agency partnerships. Dwelly recently announced a first acquisition and also partners with agencies that keep their businesses and share upside. It targets agencies with 500 to a few thousand properties, with Dexter's a possible future partner.
• AI works with a human in the loop. Landlords still want a person to talk to, so Dwelly sells work rather than software. It resembles outsourcing but adds technology for landlords, tenants and maintenance providers. Ilya favours evolution over disrupting UK property.
• Letting fits marketplaces but transacts rarely. Supply and demand are fragmented with no dominant middleman, which attracted the cofounders after Uber. Tenancies last a couple of years, though, so transaction frequency is low and limits a pure marketplace.
• Uber paid for scale against Yandex. Ilya calls the Yandex price war among the fiercest outside China and defends discounts as the fastest route to efficient scale, if not so deep that riders choose public transport. The merger gave Uber 40%, versus under 20% in China.
• Uber's general manager role trained founders. The job was entrepreneurial with a 360 view of a market. Teams tracked driver acquisition costs, trips and market share rather than headcount, and managers had five to 10 direct reports across Russia's 17 or 18 cities.

Founder lessons

• Company culture continues the founder's personality, so unresolved co-founder tensions cascade through the organization.
• Timing matters as much as effort, much like catching waves when surfing.
• Scaling before product-market fit removes agility and can end in bankruptcy. Hiring junior staff for senior roles was another early mistake.
• Simplicity depends on the customer's culture, and extra revenue streams add features that erode it. Use data when it is reliable, but avoid optimizing local extremes.
• Founders should recognize the company is not themselves. Failing to detach became a crisis for Ilya several times.