Charles Thomas Real Estate.
A spreadsheet empire and a rented Salesforce login, replaced with one Zoho CRM. Two engines, owner prospecting and buyer/seller enquiries, share a single property data model.
Visit live site →A spreadsheet empire, and a system he didn't own.
Charles Thomas sells real estate in Dubai's established villa communities. His edge is knowing which owners might sell before anyone else does, and matching serious buyers to the right property fast. Each community uses its own internal "Type" codes for layouts. Most buyers have never heard of them, but they decide whether a property actually fits the brief.
Before this project, that knowledge lived in two disconnected places. A personal spreadsheet tracked hundreds of units, grown so tangled that Charles had already abandoned an earlier version of it. The other was his employer's Salesforce, which he didn't own and would lose the moment he left. On top of that, he was paying someone a flat rate plus commission just to manually call and WhatsApp a list of owners, asking if they'd consider selling.
A data model that matches how the communities actually work.
Rather than starting with software, I started by mapping the business as it actually runs day to day. That included working out exactly why the old spreadsheet had failed. Property Types turned out to be community-specific. The same letter code means a completely different layout from one community to the next. So Community and Type are combined into a single record. Real listing detail, square footage, price, condition, is held separately at the unit level, since no two "identical" units are actually identical once renovations enter the picture.
The public brand site carries that same clarity through to buyers and sellers. A background section covers Charles's real story, from a welding apprenticeship in Glasgow to building a Dubai property business from nothing. An FAQ section is written in plain, direct language for a person, or an AI assistant searching on their behalf. A genuine client testimonial and a No. 1 Bayut ranking for Mira and Mira Oasis sit on the page as proof, not just a claim.
Two engines, one shared source of truth.
The system runs as two connected engines on one data structure. The first is an owner-prospecting engine, automating the outreach Charles used to pay someone else to do by hand. The second is a buyer and seller enquiry engine, turning website visitors into pre-qualified leads matched to the right property type and community before Charles ever opens the record.
Rather than building two separate systems for buyer and seller enquiries, the standard Leads module does the work of both. A single "Enquiry Type" field controls which questions appear. Buyers see budget and move-in timing questions; sellers see unit condition and tenancy questions. Both share Zoho's native duplicate-contact detection. That solves a real edge case: an owner selling a second property through Charles ends up as one contact with two linked properties, not two separate records.
Urgency matched to actual risk, not blanket automation.
Seller leads convert to full contact records automatically the moment their property type is confirmed. Buyer leads stay on manual conversion, so Charles keeps a deliberate checkpoint over which enquiries are worth formally tracking. Owners who go quiet get a repeating quarterly check-in over WhatsApp Business API. It's paired with real market data, so it reads as useful rather than a generic "just checking in."
Every automated message checks a Do Not Contact flag before it sends, and the only way off that list is the contact reaching back out themselves. That boundary was a non-negotiable design rule from day one, not something bolted on after the fact.
A working system, proven before he committed to it.
The proof of concept was built and demonstrated within a 15-day trial, populated with real community and unit data. That let Charles see and use it before committing to anything. Here's the headline number. The fully licensed CRM costs roughly a fifteenth of what he was previously paying a person to do a fraction of the job, with no commission cut on top. The extra reach from public launch and search visibility is a welcome bonus arriving over time, not the main pitch.
Rollout is happening in three deliberate phases. First, prove the core CRM structure. Then populate it with Charles's own contacts and live listings, to refine it on real data. Only then does the site open to public traffic, with full outreach automation switched on. Each phase adds capability without outrunning what's actually been tested.