Once Upon a Time,
There Was an Industry.

Restoration Operators run their businesses in an industry where the work itself is born of crisis. The hours are unreasonable. The responsibility is to put a stranger’s home back together with care, while the people who are supposed to pay for it push back on almost every line.

The job takes everything you have. Running the business that runs the jobs takes whatever is left. The day starts before anyone else’s and ends after. Operators show up either way, whether they want to or not.

It is tiring. It is rewarding. And most of it stays invisible to anyone outside the work.

Operators are the reason this industry exists. They are the people this story is about.

Start with the claims process itself. Every estimate, photo, moisture reading, and line item an operator submits is data. Carriers and the TPAs that manage their claims analyze all of it. They compare one operator against the next, track cycle times and severity, and decide where the next assignment goes. The operator submits the same file and does nothing with it. The ownership question follows from that.

There is a dominant pricing tool in this industry. Operators are required to use it. The data they enter belongs to the company that owns the tool. That company aggregates it into analytics and contractor-performance benchmarks that the carriers paying the operator’s invoices rely on.

The asymmetry compounds in a quieter way. Industry reporting notes that the monthly price list is trued up against the estimates that get marked complete and sent back, and that carriers complete theirs while most contractors do not. So the prices an operator is held to are shaped by carrier-completed files, and the operator’s own pricing is under-represented in the very list used to settle their claims.

Earlier this year that company announced an AI integration. Carriers get cleaner underwriting insights. Contractors get a chatbot that sits beside the pricing tool, in a separate product most contractors do not use. The workflow does not change. Both run on the same analytics, built on decades of operator submissions. Operators now pay an additional subscription to ask questions about data they helped build, on terms the vendor decides, at a price the vendor sets. They get to use it only after it has been packaged and sold back to them.

Another major property-data service is owned by private equity. Its primary customers are insurance carriers and financial institutions. It also sells software directly to restoration operators. Where operator and carrier interests come into conflict, the structural alignment sits with the largest checks.

The industry is consolidating. Private-equity rollups and franchise systems aggregate operators faster than ever before. The data layer that serves them has not changed shape with them. Restoration is dissolved into the broader “claims” or “insurance” industry, and with that it loses the right to its own data.

The two most established industry benchmark reports are the Restoration Industry Association’s Cost of Doing Business report and C&R Magazine’s State of the Industry report. Both are produced with KnowHow as the software-vendor sponsor. They have done pioneering survey work in this category for years. That work deserves credit. An industry needs benchmarks to think about itself, and these institutions have been doing that work.

C&R Magazine in particular has been a singular voice in keeping the Restoration industry aligned year after year. The State of the Industry report has given operators category-level reference points that simply did not exist before. Beyond the annual survey, its reporting on methodology, certification, safety, and the day-to-day realities of running a Restoration business is what most Operators read to stay informed about what others in the trade are doing. That editorial work deserves recognition on its own terms.

Co-authoring and sponsoring a benchmark as a category-building exercise is neither wrong nor unusual. The insights have real value, and the reports give Operators category-level reference points they would not otherwise have. But the harder question is the methodology. Both reports are surveys. Operators answered questions about their own performance and their own opinions of vendors. Nothing was sourced from actual business documents or transactions. Any sponsored survey naturally reflects the questions the sponsor chose to ask. That is the structural shape of the format.

An annual survey and a continuous benchmark serve different purposes. One gives the industry a shared yearly snapshot to reflect on. The other gives an operator a live read on specific aspects of their own business as it changes. The next generation builds on what RIA and C&R have established: vendor-independent benchmarks, sourced from real operational data, with a published methodology and a structure operators can audit.

AI is everywhere. In every restoration tool. Spatial and photo recognition for damage capture. Estimate analysis. Process advice. Different surfaces, the same underlying pattern.

Most of these tools call the same foundation models through an API. The vendor adds a sticker and a vertical wrapper. Without proprietary data behind the model, AI tends to be submissive. It tells you what it can without context. The kind of recommendation an operator actually needs to make a decision rarely shows up.

Some vendors do train their own models on user data. That underlines a deeper problem. The operator’s data is distributed across many vendors. No single vendor has the whole picture. The operator certainly does not. Vendors do not collect data for the benefit of the operator. They collect it to better monetize their solutions.

What is left for the operator is the hope that the product improves at all. The price of the subscription will rise either way. And there is no easy way out.

Tech has significantly improved the quality of work and profitability of many restoration businesses. But every tool in the stack also generates more data. An overwhelming amount of data. Almost no tool generates decisions.

The estimating tool gives numbers and probabilities. The job-management tool tells you what is open. The accounting tool tells you what came in last month. None of them tell you that your field-documentation contract is priced 17% above what twenty-seven peers pay, that the renewal window opens in seven weeks, or that two of those peers switched last quarter, and to what.

Operators do the translation themselves. Open six tabs, hold the picture in their head, decide between meetings or on the drive home. The decisions get made anyway, just later than they could be, and on less of the information than the operator was paying all these tools for in the first place.

Verinode delivers decisions, not dashboards.

What if you had a Chief Data Officer who knew your entire business? Every job, every vendor, every email. Who could analyze every operational process in a split second. Who knew every benchmark in your category. Who was always on your side. Who was available whenever you needed to make an informed decision.

The natural objection: operators could already drop everything into ChatGPT or Claude and ask. That works for some questions. It does not work for the questions that decide a business.

Foundation models reason by predicting what is most likely based on the patterns in their training data. They were trained on the open internet, not on the books and ledgers of working restoration operators. So an LLM will give a plausible-sounding answer to almost any question. Plausible is not the same as verified. “Your dehu rentals look high” and “your dehu rentals are 23% above the peer median for jobs of this size in your region” are different statements. Only one is built on data that actually exists.

But this Chief Data Officer can only operate on data provided by a data trust that is secure, partisan, and independent. Any data, from any tool, any interaction, any document, at the operator’s terms. That is what Verinode builds.

Verinode is the data trust. It sits on top of your tech stack, not next to it. The estimating tool, the job-management tool, the field documentation, the property-data services. None of them go away. Verinode does not replace them. It does not judge their use. It helps you use them better, and pay less for them. What it adds is a secure place for all data an operator already produces: a place only the operator controls, where the data works for them and them alone.

Now this part is important: The data stays the operator’s. What gets shared with the network is anonymized aggregate, contributed safely. In return, every operator gets analysis they could not produce alone, and the industry finally gets something it has never had: a data counterweight to the large tech companies, TPAs, and carriers that have controlled the picture until now.

Bloomberg pooled every stock trader’s behavior. Today every trader reads it. Verinode is what that looks like for restoration.

Operators join Verinode as members. They want to take control of the data their operations produce, and to make better decisions with it. Here is how that works in practice.

Nothing has to be wired up to begin. The email and calendar can connect directly now, and so can the books through QuickBooks, if an operator wants that. Everything else just flows in from the exports and the inbox they already have. Upload, photo, paste, voice, call and meeting notes, email, bulk import. Anything an operator already has can be stored and ingested.

Invoices. Receipts. Supplements. Ledgers. Emails from adjusters. Field documentation. Standard operating procedures. Whatever surface produces it, the data flows in.

A canonical chart of accounts normalizes margin and cost data across operators with completely different bookkeeping conventions, so cross-operator benchmarks finally mean something.

Operators own the raw data. The Vault Key, issued at sign-up, column-encrypts the identifying parts of every operator’s data (customer details, claims, team identities, full street addresses, free-text records, vehicle and equipment identifiers) with a 30-character key scoped to the operator. No other operator or carrier can read them, and no Verinode employee can browse or export them through any tool we operate. The operator retains the raw record; Verinode publishes the cohort.

That data lives in a separate PII database, region-deployable and operator-scoped, with row-level security on every table. Analytical dimensions (carrier and vendor names, equipment models, dollar amounts, dates, categories) stay in clear so peer benchmarks can exist at all, protected by service-role-only access and an immutable audit log.

The line between encrypted and plaintext is drawn by the law. Carrier contracts, employee privacy laws, and state privacy laws say certain categories of operator data cannot be aggregated across operators. The vault enforces that line; the rest of the platform sits on top of it.

The agent that reads operator-private data is the operator’s agent. It cannot send the data to other operators or to the intelligence layer. The architecture itself enforces this, not a marketing promise.

What flows out is one-way-transformed aggregate of the analytical dimensions: hashed identifiers, bucketed numbers, cohort-floored. No individual operator can be reconstructed from it.

Operators contribute toward the collective. The collective never exposes the individual.

The data trust is advised by independent restoration operators. Working operators will sit on the Verinode Operator Advisory Council. They are not on Verinode’s payroll. They run real businesses.

The Council reviews material changes to the data-use policy. It is consulted on uses of network data that could affect operator interests. Its review is public, so any move against operator interests would be visible to the network.

Carriers cannot acquire this seat. Vendors cannot fund it. The seat belongs to operators by design.

Based on the data its members contribute, Verinode publishes research the way universities do. Every benchmark, every report, every comparison has a methodology page operators can read.

What questions were asked. What data was used. How the sample was constructed. How variables were normalized. How outliers were treated. What confidence intervals apply.

The methodology comes before the conclusion. The conclusions follow from the methodology. Anyone who wants to challenge a number can read exactly how it was produced and challenge the part they disagree with.

An illustrative example is already published: the Job Management Quadrant.

This is what becomes possible when research is independent of vendor sponsorship and sourced from real operational data. A different instrument than an annual survey, built for a different job.

Peer benchmarks the carriers will never see. Labor margin, vendor pricing, carrier payment cycles, scorecard performance, equipment utilization. Compliance, safety, recruiting, reputation. Your line next to the peer line.

Regular benchmark and industry performance reports. Built from real operator data, not survey responses. Vendor comparisons based on what operators actually pay. Performance metrics based on what operators actually do.

Insights operators could not produce alone. Peer comparisons with people doing similar work. Goals worth aspiring to, drawn from what others have actually done in the same conditions.

Verinode IQ is the operator-side product. Inside it lives a personal Chief Data Officer that is onboarded by the Operator, and only works for the Operator. IQ is not another software platform with seats and licenses to learn.

IQ changes how operators interact with their own data. No longer through a dashboard with rows and filters, but through a Chief Data Officer they can talk to, email, or ask anything. IQ is the personification of the operator’s business: every job, every invoice, every supplement, every email, given a voice that answers only to its owner.

The operator works with their Chief Data Officer the way they work with a chief of staff. Forward an email and the Chief Data Officer takes it from there. Reply to a thread and IQ updates the plan. No platform to log into, no new habit to learn. Open the app when you want the full picture, and it is built for the two minutes between jobs: a feed to scroll, decisions to swipe. Same agent, same memory, wherever you reach it.

Beneath IQ runs a network of dozens of specialized agents, one for each operational area. Jobs, vendors, clients, equipment, safety, certifications, compliance, team, recruiting, P&L. Each agent has its own data domain, its own retrieval layer, its own outcome capture. They are not chat wrappers around a foundation model.

IQ watches for patterns across them, the way a real Chief Data Officer would.

Every set of meaningful data signals terminates in one decision card. The gain is stated. The cost of inaction is stated. The operator presses Act, or Park, or asks Why. Once a decision is made by the operator, an action plan is drafted, executed, and tracked.

Picture a four-truck operation in Atlanta. On a Tuesday morning, IQ surfaces a card: your dehu rental cost per water-mitigation job is running 23% above peer median this quarter, and the operators paying less use two specific vendors. The renegotiation email is drafted, with the peer benchmarks cited and alternative vendors loaded in. The operator presses Act. Three weeks later the contract is signed, and IQ records what landed.

Outcomes feed back into the model. Plans for similar signals across the network learn from what landed.

Beneath the decision card and the Chief Data Officer is a discipline called process mining. The practice of comparing how a business actually works to how it is supposed to work.

A traditional consultant does this with interviews and clipboards. Process mining does it with the data the business already produces. Every job, every supplement, every invoice, every email is a step in a real process. Map enough of those steps and the bottlenecks reveal themselves. Where work piles up. Where decisions stall. Where the same mistake recurs across different jobs.

Most industries have process mining. Restoration does not. Without operator-aligned data, there is no substrate. Carrier-side data captures only what carriers care about. Vendor-side tools capture only what helps them sell. Neither sees the operator’s full process.

Verinode does. The data is the input. The peer comparison is the reference. The decision is the output.

Imagine an industry where operators are not the cheapest input in someone else’s profit model.

Imagine benchmarks built independently of any software vendor. Real comparisons that mean something.

Imagine vendor pricing so transparent among members that every operator can negotiate their own rate from the same real numbers a national account already has.

Imagine carriers unable to quietly unwind a year of negotiation because every operator can push back on their own file with data.

Imagine the Restoration Industry Association walking into a room with statistics it owns, sourced from operators it represents.

Imagine a franchise system or a private-equity rollup buying into a network and getting actual cross-portfolio intelligence on day one. Verinode HQ is built for that.

This is the industry Verinode is trying to build. None of it without the operators who make the trust real by joining it. Every operator who joins makes the next decision sharper for everyone.

Verinode is partisan. The first independent data trust for Restoration, built for Operators, only works if the Operators who join early stay close to the project.

The first hundred Premier members carry the Founder cohort. Early access to new features before they reach the rest of the platform. A direct feedback channel into the team building Verinode. And when Operator seats on the Operator Advisory Council open, the Founders are the natural pool we recruit from.

The people Verinode is built for are also the people Verinode is built with. The Founder cohort is how we make that real.

I spent the last several years as COO of magicplan, a SaaS tool that quietly runs underneath a lot of restoration documentation. As early as 2018, we were training our own machine learning models on real customer data to recognize floor plans from a phone camera, long before “AI” was a marketing word.

Across the United States, Canada, and Europe, I noticed the same pattern across vendors. Operators do the work. Other people own their data. Field documentation, estimating, job management, ERP. Every category, the same template. More AI features, more attempts to own a bigger share of the operator’s job lifecycle, more value extracted on the surface. Beneath all of it, the same thing makes every one of those tools work: structured data. But almost none of them treat the data foundation itself as the problem worth solving. They treat it as the asset to monetize, often through backroom conversations with carriers about access.

That is where the idea for Verinode came from. The next company in restoration tech should not be another process tool, another submissive AI, or another dashboard. It should improve the data foundation each operator runs on, turn that data into decisions the operator can act on every day, and keep control with the operator as the data moves to carriers, TPAs, or vendors.

AI alone is not enough. It can take fragmented data and produce confident-sounding answers. It cannot, on its own, tell whether the answer is right. What makes the answer right is the data foundation underneath it. Real, normalized, consented, attributed. That is infrastructure. The data pipeline, the canonical schema, the consent framework, the agent architecture, the outcome capture, the learning loop that closes back into the model. It takes years to build. It is invisible from the outside. It is what separates a tool that gets better as the network grows from one that gets a new sticker every quarter. Most of what is sold today as “AI for restoration” is the latter.

This trend has already played out in tech. It is happening to vertical industries now, one trade at a time. Restoration will not be exempt. The question is who builds the infrastructure that powers AI in this industry: carriers, vendors, or operators. Once it is entrenched, it stays in place for a long time. That is why the timing of Verinode matters.

The research and benchmarks share the dynamic. Restoration gets folded into the wider “Insurance” universe so quietly that insights and reports on what the industry should do are written or sponsored by companies affiliated with the carrier side. There is little research that truly belongs to the people working in this industry.

My belief is this: in the age of AI, your operational data is your business’s greatest asset. The operators who own it and build on it are the ones who will pull ahead. Real insights, working in their favor. Better margins. Better business performance.

If you are reading this, you are someone I respect so much that I want to get your feedback before anyone else’s. Tell me what you think. Tell me what you see today. Tell me how you imagine the future. Verinode should be built around operators. I’m excited to hear your perspective.

Stefan