Author: will_tygart

Restoration contractors are paying for Encircle. And PSA. And DASH. And a CRM. And a project management tool. And a call tracking service. And a reputation management platform. And an estimating integration. By the time you add it all up, a mid-size restoration company might be running eight separate software subscriptions, each with its own login, its own invoice, its own support line, and its own way of storing data that doesn’t talk to anything else.

I’ve been watching this stack accumulate for years. And I’ve been thinking about a question I haven’t seen anyone ask out loud:

Who owns the data when the job is done?

What Software Companies Are Actually Selling

Encircle is a genuinely good product. So is PSA. So is DASH. I’m not writing this to trash them. They solved real problems — structured photo documentation that insurance carriers accept, drying logs that meet IICRC standards, scope writing that integrates with Xactimate. These things are hard to build from scratch and they matter in a claims-dependent business.

But here’s what all of them are also selling, whether they say it or not: a structured way to store your business’s data. Customer records. Job histories. Equipment logs. Photo sets. Communication trails. Every one of those platforms is capturing the operational intelligence of your company and holding it in their database, in their format, accessible through their interface.

The subscription isn’t just for the software. It’s for continued access to your own data.

That arrangement made sense when there was no alternative. You needed the structure, and the only way to get the structure was to accept the terms. The software vendor provided the architecture. You provided the data. The architecture stayed with them.

That’s the deal. It’s been the deal for twenty years. And it’s changing.

What’s Actually Different Now

The thing that changed isn’t AI, exactly. It’s the integration layer.

For most of the software era, building custom business tools required engineering teams, expensive infrastructure, and months of development time. That’s why SaaS won — you couldn’t build it yourself, so you rented it from someone who could. The subscription model was the price of access to capability that was otherwise out of reach.

What’s different now: a single developer — or an operator who knows how to use modern AI tools — can assemble custom business infrastructure in days that would have taken a team months in 2019. A Google Cloud VM costs $60/month. A CRM custom-built on WordPress with webhooks firing into CTM, Slack, and a Firestore job log costs fractions of what PSA charges. An AI intake agent that handles emergency calls, qualifies the job, creates the customer record, and pings the on-call crew — built on Twilio and Claude on Vertex AI — costs less per month than most restoration companies spend on coffee.

The capability gap that justified the subscription is closing. Not for every business — not yet — but for businesses that have someone close enough to understand what they need and how to build it. And critically: when you build it, you own it. The data lives on infrastructure you control. It doesn’t leave when you cancel a subscription because there’s no subscription to cancel.

What Encircle Still Does That Matters

I said I wasn’t writing this to trash these companies and I meant it. So let me be specific about what they do that’s genuinely hard to replicate.

The compliance layer. Insurance carriers have specific documentation requirements. IICRC has drying log standards. Xactimate has a particular way of handling scope line items. Encircle has spent years building integrations with those systems, getting their formats accepted by carriers, making their documentation hold up in adjuster reviews and litigation. That institutional trust is not a feature you can code in a weekend. It’s accumulated credibility that took years to build and is worth real money to contractors whose revenue depends on claims getting approved.

The field mobile experience. Technicians in the field need something fast, offline-capable, and purpose-built for how they actually work — photos, moisture readings, equipment logs, job updates — all from a phone in a flooded basement. Generic platforms aren’t optimized for that workflow. Encircle is.

So no — the Company OS doesn’t make Encircle irrelevant for everything. What it makes irrelevant is the parts of Encircle — and PSA, and DASH, and the CRM, and the project management tool — that are really just coordination and data structure. The scheduling, the customer records, the communication trails, the job status tracking, the lead attribution, the revenue reporting. All of that can live in a system you own, wired together through APIs, with your data staying on your infrastructure.

You keep Encircle for what Encircle is uniquely good at. You stop paying for the eight other subscriptions that are just doing coordination work you could own.

The Model That Makes This Work

The reason most restoration contractors won’t build this themselves isn’t that they can’t afford it. It’s that they don’t have the time or expertise to architect it — and even if they did, they’d have to manage it forever. That’s not a restoration contractor’s job. Their job is running jobs.

The Company OS model I’ve been developing solves this by flipping the arrangement entirely. Instead of the contractor buying software subscriptions and managing a fragmented stack, I build and host the entire infrastructure — VM, CRM, call tracking, AI intake, content engine, ad management — and take a percentage of revenue I can prove I drove through the system. The contractor pays nothing upfront and nothing ongoing for the infrastructure. They pay on verified results.

The difference from the SaaS model: the data architecture belongs to the system I built, which is operated in the contractor’s interest and accessible to them. The attribution data, the customer history, the job records, the communication logs — all of it lives in a structure we both can see, verified by Call Track Metrics, not locked behind a vendor’s dashboard.

That’s not a software product. That’s an infrastructure partnership. And it produces a fundamentally different answer to the question of who owns the data when the job is done.

The Question Worth Sitting With

I want to be careful here about the scope of what I’m claiming. The vertical software companies — Encircle, Xactimate, PSA — aren’t going away. The contractors who need carrier-compliant documentation and field mobile tools will keep paying for them. The compliance layer is real and the field experience is real and those are genuinely hard problems.

What I think is ending — or at least what I think deserves to end — is the part of the software subscription economy built on the coordination tax. The $200/month CRM that stores your customer records in someone else’s database. The project management tool that knows your job pipeline better than you do. The reporting dashboard that shows you your own business through someone else’s lens. That category of software exists because the integration layer didn’t. Now it does.

So here’s the question I’d ask any restoration contractor right now: for every subscription you’re paying, do you own the data when you stop paying? Do you know exactly where your customer records live, who controls the schema, what happens if the vendor raises prices or shuts down?

Most contractors have never asked this because they’ve never had to. The subscription was the only option.

---

It isn’t anymore.

The question isn’t whether your software does the job. The question is who owns the data when the job is done.

Nobody sits down and says “I’m going to build an operating system for an entire industry.” That’s not how it starts. It starts with one client who needs a website. Then another who needs their Google Ads cleaned up. Then someone asks if you can help them figure out why their phone isn’t ringing.

You solve problems. You move on to the next one. You don’t zoom out.

I zoomed out recently — for the first time in a long time — and what I saw surprised me. I hadn’t been building a marketing consultancy. I’d been building a vertical operating system for the restoration industry, one problem at a time, without ever calling it that.

How It Actually Started

The first piece was SEO. A restoration contractor needed to show up when someone searched “water damage restoration” in their city. Straightforward enough. I built the content, optimized the site, tracked the rankings. It worked. They referred someone else. That someone else had a slightly different problem — their ads were running but the calls weren’t converting. So I looked at that.

Call Track Metrics came in because I kept running into the same argument: the client thought the calls were coming from one place, I thought they were coming from another, and neither of us could prove it. CTM solved that. Now every call is tagged to the source — the keyword, the page, the campaign, the full journey. Attribution stopped being a debate and became math.

Then I noticed that the calls were coming in but jobs weren’t closing at the rate they should. That’s not an SEO problem. That’s an operations problem. So I started looking at intake — how calls were answered, how follow-up happened, how estimates were scheduled. An AI intake agent started to make sense. Not because I was trying to build AI products, but because the gap was right there and I could see it.

The Restoration Golf League came from a completely different direction. Restoration contractors need referral relationships with insurance adjusters and property managers. That’s the commercial side of the business. A golf league is one of the best relationship-building structures that exists in professional services — relaxed, repeated contact, shared experience. It wasn’t a marketing idea. It was a relationship infrastructure idea that happened to use golf as the mechanism.

The Inventory I Didn’t Know I Had

When I actually sat down and listed everything that exists right now across the work I’ve been doing, here’s what came out:

A content intelligence platform — a BigQuery knowledge base that logs every session, surfaces patterns, and drives automated publishing. A lead tracking infrastructure built on Call Track Metrics, wired to every traffic source. A referral network of restoration contractors meeting through a structured golf league across multiple cities. A commercial compliance strategy using fire extinguisher inspections as a loss leader to get in the door with property managers. An AI receptionist product purpose-built for restoration intake — Twilio, Claude on Vertex AI, Cloud Run, Firestore. A Company OS model — a fully hosted GCP environment where I run a contractor’s entire revenue infrastructure and take a commission on verified results. A WordPress CRM being built and dogfooded on my own site before being offered to clients. A knowledge cluster of five interconnected websites building topical authority in the restoration and risk intelligence space.

None of those were planned in sequence. Each one was the answer to a specific question that kept coming up. But together they cover almost every layer of how a restoration business actually operates — lead generation, lead tracking, intake, conversion, referral relationships, commercial acquisition, operations tools, and content authority.

That’s not a service menu. That’s a stack.

Why Accidental Might Be Better Than Planned

I’ve thought about whether it would have been better to plan this from the start. Design the full system upfront, build it in sequence, launch it as a coherent product.

I don’t think so. And here’s why.

Every piece of this was validated before the next one got built. The CTM infrastructure exists because attribution disputes are real and expensive. The AI intake agent exists because I watched calls get dropped after I’d already driven them. The golf league exists because I saw contractors lose commercial accounts to competitors who had better adjuster relationships, not better work. Each problem was visible because I was close enough to the industry to see it — not designing from a distance.

The version of this that gets designed upfront has a different failure mode: it’s theoretically complete but practically wrong. The problems you think exist from the outside are never quite the same as the ones that actually exist on the inside. Building problem by problem, staying inside the industry, means every piece of the stack is load-bearing because it was built under load.

There’s also something that happens when you’re not trying to build a system. You’re more honest about what’s actually needed. You don’t add things because they complete the picture — you add them because the gap is genuinely painful. The result is a leaner, more accurate stack than anything I could have designed in a planning session.

The Question I’m Sitting With

The thing I keep coming back to: is this replicable in other verticals, or is it only possible because of the depth of time I’ve spent inside restoration specifically?

I genuinely don’t know. The honest answer is probably both. The approach — stay close, solve real problems, let the system emerge — is transferable. But the specific inventory I ended up with is deeply shaped by restoration’s particular quirks: the insurance dependency, the emergency-driven intake, the adjuster relationship dynamics, the commercial vs. residential split, the franchise structures, the IICRC certification culture.

A different vertical would produce a different stack. HVAC has different intake patterns. Personal injury law has a completely different referral economy. Healthcare has different compliance requirements and trust dynamics. The method of paying attention and building toward what you see would be the same. The pieces that emerge would be different.

What I’m more confident about: you can’t fake the depth. The reason the stack works is because I know what it’s like to be a restoration contractor well enough to feel the pain of each layer. That knowledge isn’t transferable quickly. It’s accumulated. Someone who decided tomorrow to “build a vertical OS for HVAC” would be designing from the outside. They’d get some things right and miss the things that matter most, because those only become visible from inside.

What This Changes

Naming a thing changes how you relate to it. Before this realization, I was a marketing consultant who did a lot of different things for restoration companies. That description is accurate but it undersells the coherence of what’s actually there.

Now I think of it differently: I’m a vertical infrastructure builder who happened to start in restoration and went deep enough that the full stack became visible. The individual services aren’t the product. The system is the product. Any one piece of it — just the SEO, just the CTM setup, just the AI intake — is less valuable than the whole because the whole is integrated in ways that individual pieces can’t be.

That changes what I build next, how I talk about what I do, and who I build it for. It also changes what “being done” means — because a vertical OS is never really done. Industries evolve, problems shift, new gaps appear. The work is staying close enough to keep seeing them.

---

I didn’t plan any of this. I just kept solving the next problem.

Turns out that’s a strategy.