Donastar

Donastar runs a field service business on two tools that were never designed to talk to each other.

Zoho Desk is a customer support tool. Striven is a generic ERP. Neither one was built for a company that manages a network of third-party service providers, and the gap between them was being filled by hand.

Copy the customer name here. Paste the PO number there. Email the completion form to everybody and hope it lands on the right ticket.

Here is the thing: that's not a configuration problem. It's an architectural mismatch. And you cannot configure your way out of an architectural mismatch.

Donastar is a ten-person shop servicing commercial equipment. They're an intermediary. Clients send them work, and they dispatch it to a network of independent service providers. That one detail turns out to be the whole story, so hold onto it.

Every job started with someone reading a freeform email or a voicemail and manually turning it into an estimate. Completion forms came back from providers as a blast email to the whole team, unattached to any ticket, and somebody had to figure out where each one belonged. Aggregator platforms like ServiceChannel and ServiceNow spun up a fresh duplicate ticket on every single status update. Filter assets expiring, an entire replacement business line, were tracked by hand. Bills showed up by email, by text, by whatever channel the provider felt like using that day.

Add it up and the territory managers were losing one to two hours a day to copy-paste, re-entry, and routing. Not to the work. To moving the work between two systems.

The obvious move is to go buy field service software. So that's where we started, not where we ended.

We evaluated three paths in the open: connect the systems they already had, migrate the whole company into Zoho One, or move to a purpose-built field service platform. ServiceTitan, Jobber, FieldEdge, Housecall Pro, Workiz, ServiceMax. All of it, scored against Donastar's own must-have list.

And that's where the real insight surfaced, the one nobody selling software is going to volunteer.

Every SMB field service tool is built for the company that is the service provider. Jobber assumes you're the plumber. FieldEdge assumes you're the HVAC company. Housecall Pro assumes you're the cleaning crew. They're all built for a business dispatching its own technicians.

Donastar is none of those. Donastar is the intermediary managing a network of them.

The platforms that do serve that model natively (ServiceMax, Salesforce Field Service) start at $500 to $1,000+ a month and are built for organizations five to ten times Donastar's size. The tools priced right for a ten-person team weren't designed for the business at all.

Even the one candidate that looked perfect on paper fell apart under inspection. ServiceTitan has a confirmed ServiceChannel integration, which seemed like the answer to the aggregator problem. But that integration was built for the other side of the transaction: for providers receiving work orders from ServiceChannel. Donastar creates and manages those work orders. The integration didn't touch their pain point at all.

So the honest conclusion was that there was nothing to buy. Not because we hadn't looked hard enough, but because the gap is structural. It isn't closing.

That is what justified a custom build. Not a preference for writing code.

We kept Zoho and Striven exactly where they were, and built the connective tissue between them.

One design detail worth naming, because it shows the shape of the problem: neither Zoho nor Striven supports real-time webhooks. So TAP syncs on a schedule. The dashboard exists partly for the moments when someone needs a record updated right now, and one button does it. We designed around the real constraint instead of pretending it wasn't there.

Donastar had been burned by a developer before. So transparency wasn't a nice-to-have on this project. It was the contract.

Fixed price, defined scope, phased delivery. No open meter.

Sandbox first, every time. Nothing touched production without testing and their sign-off. 337 automated tests, all green at handoff, because "seems to be working" is not the same as working.

Kickoff on April 16. Production cutover on May 28. A real handoff with a user guide, a team walkthrough, and a person who picks up the phone.

Thirteen days after cutover, pulled straight from the production database:

The copy-paste tax is gone. The hours that used to disappear into moving data between two systems now happen in the background, on a schedule, without anyone thinking about it. The team sees every job in one place for the first time. And when something goes wrong, it surfaces on the board today instead of in a phone call thirty days later.

The dashboard was never just a dashboard.

It's a data layer Donastar owns, sitting between the two systems, built to be extended without starting over. That's the difference between buying software and building it. When the tool fits the business instead of the other way around, everything after it gets cheaper.

A portal where service providers bid on jobs and upload their own bills. Profitability reporting that answers what each job actually made. AI agents that read an incoming request and draft the estimate before anyone opens the email. None of that is a rebuild. It's the next floor on a foundation that's already poured, and it's what Phase 2 is being built on right now.

No platform was built for a business like Donastar's. That used to be a problem you had to live with.

With AI changing the economics of custom software, it's starting to look more like an unfair advantage.