A proposal for Beck Building Company — prepared for Kevin O'Donnell and the principals.
We want to make sure we understood you correctly — before we say anything else.
Beck Building has been doing this for 55 years. You have 66 people, four principals with a combined 130-plus years at the company, and a reputation built entirely on word of mouth. No advertising. Clients come to you because you've earned it.
The business model is cost-plus — but you run it almost like a hard bid. Owners trust you because you treat their money like your own. Nothing moves without owner permission. You're a fiduciary first, a builder second. That reputation is the brand.
Nothing you build ever looks the same. Luxury custom homes in Vail, Aspen, and Denver. Ranch work in western Eagle County. Small projects that grew from honeydew lists into a serious division. Every project is unique — which means your estimators are constantly starting from scratch.
"Nothing we build ever looks the same."
— Kevin O'DonnellYou have two full-time estimators. One of them is retiring in a couple of years. They're fantastic — fast and accurate. They work through three phases: early design drawings (SD), developing design (DD), and then construction documents (CD) where you send plans out for full sub bids.
At the early stages, the estimating tool is a giant Excel file — multiple tabs, built up over decades. It's a feel as much as a formula. Your estimator looks at a house, compares it to houses they've built before, and works from there. When they retire, that file becomes a static artifact. The feel — the judgment — goes with them.
"This house feels like that house."
— How the unit cost file actually worksAt CD stage, you go out to subs — usually three to five bids per trade. Leveling those bids is real work. Every sub formats differently. Multi-page documents. You're hand-plugging numbers from each one to compare them. If the system could just read those bids and level them, it would save hours every project.
You work with a limited sub pool. Two or three per trade per county. You know them. You're not trying to discard the ones that struggle — you'd rather bring them up. But you need to know which ones are slipping before it becomes a problem on a live job.
You're running Sage for cost accounting, Crystal Reports for pulling data out of Sage, and Excel for everything else. Data flows: Sage → Crystal → Excel. You enter things in all three. You described it as triple entry, and that's exactly what it is.
Sage actually produces estimate vs. actual by line item automatically. That's more than most GCs have. But getting into that data, comparing it across projects, and feeding it back into future estimates — that's still an effort. The loop isn't closed.
"Every time we've gone out and looked to integrate, there's nothing. There's no one company that's strong."
— Kevin O'DonnellEvery new architect means starting over. You've built 350-plus projects, but when a new design team comes in, there's no quick way to say: "Here's what we've learned building in this terrain, at this altitude, with these materials." You start from scratch. The institutional knowledge lives in a few people's heads.
Your IT team wants your data handled securely. Your partners need to approve before anything is shared. That's exactly the right instinct, and it's something we address directly.
We see this as a walk-alongside relationship. No rush. No pressure. Start with five completed projects, see what surfaces, and decide from there whether it's worth going further. The pace is entirely yours.
This isn't a technology problem — it's a structural one that the industry has never addressed.
You described it clearly yourself: every time you've looked for a system that connects your estimating to your actual costs, it doesn't exist. Plenty of tools handle one side or the other. Nobody closes the loop.
The companies that build project management software are good at tracking what happens during a project. The companies that build accounting software are good at recording what got spent. But the step between — comparing what you estimated against what you actually paid, then feeding those corrections forward into the next estimate — that's the gap.
Estimate vs. actual data exists in your system — at every level. What you estimated for concrete forming versus what you actually paid. What a specific tile subcategory cost versus what the architect assumed. Nobody has built the layer that turns those reconciliations into future accuracy, trade by trade, line item by line item.
Decades of judgment live in a few people. When they retire, the data stays but the wisdom goes. No one has built a system to preserve the judgment, not just the numbers.
You know your subs. But tracking which ones consistently come in over or under, which specific scope items trigger their change orders, which ones generate RFIs that delay your schedule by weeks — that intelligence stays in people's heads. No one is connecting the bid line item to the change order to the warranty callback to show you the full picture per vendor, per trade, per scope.
Vail labor costs differently than Denver. Altitude affects materials. Mountain access changes schedules. None of that is captured in a way that gets automatically applied to the next mountain project.
These aren't gaps because no one was smart enough to solve them. They're gaps because you need a lot of completed projects from one company — all structured the same way — before the patterns become visible. Beck, with 350-plus projects across three markets, is unusually well-positioned to close them.
Five completed projects. Your data. Structured and returned to you as patterns.
We're not bringing industry averages. We're not using national benchmarks. Everything we surface comes from Beck's own completed work — priced in Beck's markets, built by Beck's subs, estimated by Beck's team.
"All the things you're talking about — these are things we're trying to make — it just takes a lot of time."
— Kevin O'DonnellThree phases. You keep building. We run alongside.
The core commitment: No software change. No migration. No disruption to how Beck operates. Your team keeps doing exactly what they're doing — we work from the data that already exists.
Using the 14-item checklist we're sending along with this proposal — original estimate, final cost report, sub bids, change orders, closeout accounting, and a few others — you send us five completed projects. These are projects that are finished. No active job information. No live pricing. Just history.
How you deliver those files is entirely up to you — whatever method your team is most comfortable with. A shared Dropbox or Google Drive folder, email, a USB drive, or Garry can pick them up in person at the office. There is no required format or upload portal. We work with however Beck already handles files. And your projects are complex — that is actually ideal. The more variety in the five projects, the more patterns the system can identify. A mix of custom homes, renovations, and ranch work gives us the broadest baseline to work from. You do not need to clean anything up before sending it. We handle the structuring on our side.
We structure that data, analyze it, and return our findings to you within four weeks. You see exactly what we found before we build anything further. You own every finding. If you want to stop there, you stop there.
No software change No fee 4 weeksFrom five projects, we can show you: where your early estimates typically drift by the time you reach closeout, which trades have the widest variance, how your Vail numbers compare to your Denver numbers, and where your subs are most and least predictable.
This is the first time Beck's project history will have been looked at as a connected dataset instead of fifty separate files. Most GCs have never seen their own patterns laid out this way.
Findings returned to you Your data, your patternsAs more projects feed in — new completed jobs, additional historical data — the software application gets more accurate for Beck specifically. Not for GCs in general. For the way Beck estimates, the subs Beck uses, and the markets Beck builds in.
Your retiring estimator's judgment gets encoded into the system before they walk out the door. New estimators onboard faster. Every project makes the next estimate tighter. The institutional knowledge stops living exclusively in a few people's heads and starts living in a system Beck owns.
No double entry Runs alongside your current workflow Beck owns its project dataWhat Beck gets at each phase — and what compounds over time.
Beck keeps building exactly as it does today. You send five completed projects. We structure the data and come back with concrete findings.
Deliverable
"Here's What We Found" report — estimate-to-actual variance by trade category and line item, sub performance chain (bid → CO → RFI → warranty), cost-per-SF baselines per market, and the specific patterns in your five projects that are not visible in Sage or Excel today.
The software application now knows Beck's cost history well enough to flag problems on new bids before they go out the door.
Deliverable
Live estimating dashboard — sub scorecards with bid-to-closeout chain, estimate accuracy by division and line item, missing scope alerts on active bids, RFI tracking with schedule impact and invisible general conditions costs, and the unit cost intelligence your estimators have been building for decades — now structured and searchable.
Every completed project automatically recalibrates the system. The more Beck builds, the more accurate the next estimate becomes — without anyone manually updating a file.
Deliverable
Self-improving estimating software — every closed project recalibrates accuracy, line-item cost baselines update automatically per market, prediction tracking shows the system getting tighter over time. Owned by Beck, built for Beck, closed to the outside.
What this is not
✕ Not a new software platform you have to learn
✕ Not a replacement for Sage, Excel, or Crystal Reports
✕ Not a migration or data conversion project
✕ Not a generic off-the-shelf product shipped to every GC
✕ Not a disruption to how your estimators or PMs work
✕ Not a commitment beyond Phase 1 until you have seen results
Everything above runs on Beck's data alone. What follows is what happens when the network grows.
Beck's private intelligence — your estimate vs. actual patterns, your sub performance history, your institutional knowledge — gets better with every project you complete. That alone changes how you estimate. But there is a second layer that no single company can build on its own, no matter how many projects they complete.
Before you price Division 09 on a mountain renovation, the system already knows: across 50 or more closed projects from 10 or more contractors in this region, finishes average $8.40 per square foot, with a normal range of $6.80 to $9.20. Your number is in range — or it is not, and you know before you submit. These are common divisions that every contractor touches regardless of project type — concrete, framing, electrical, mechanical, finishes. The benchmarks come from work that every GC does, not from what makes any one company unique.
Your electrical bid comes in 24% below the rolling average of the last 50 comparable projects in your region. The system flags it before you submit: verify scope coverage. Maybe the sub missed transition strips, underlayment, or specialty items from the architect spec. That one alert could be the difference between a profitable project and a six-figure overrun. This is the cross-check that lives in a veteran estimator's head — except it is backed by data from every contractor in your market, not just one person's memory.
Across dozens of closed projects in your region, mechanical trades average 11% over initial bid. Concrete averages 3% under. These are not one company's numbers — they are the combined closeout data of every contractor on the platform, anonymized into percentages and ranges. No project names, no company names, no dollar amounts. Just the pattern: this trade, in this region, for this project type, tends to land here. Your next estimate starts with that knowledge built in.
Mountain construction costs differently than flatland. Altitude, access, labor markets, seasonal timing — these variables shift every number. The platform tracks cost per square foot by division, by project type, by region, updated with every closed project. Your Vail numbers look different from your Denver numbers. The system knows that, and it adjusts. Your next estimate starts with a data-backed baseline, not a guess.
What stays out of the benchmark layer: If only two or three luxury custom home builders in your region use a specialized division — custom millwork, specialty stone, high-end A/V integration — those divisions do not generate benchmarks. The contributor count is too low. The system only publishes benchmarks where enough contractors contribute to make the data truly anonymous. The divisions that make Beck unique stay private. The divisions that every contractor shares become intelligence that helps everyone estimate more accurately.
You told us that constructability is where your real efficiency lives — not estimating. You said you start over with every architect. Just about the time you figure each other out, the job is done. A new architect walks in, and you retrain from scratch.
The network intelligence layer changes that dynamic. Here is how.
Right now, an architect designs something, hands you plans, and you tell them it is over budget. Redesign. More time. More cost. More frustration on both sides. With Beck's own historical cost data structured by division and line item, your estimators can give an architect a cost confidence range before plans are finalized — based on what Beck has actually paid on comparable projects, not an industry guess. The architect adjusts during design, not after. You stop getting plans that are 40% over budget on day one. This is Beck's data informing Beck's architect conversations — no other company's project details are involved.
When a new architect comes on, your estimators already have Beck's full cost history organized by division, subcategory, and line item across every market you build in. The new architect does not need three projects with Beck to understand your cost reality — your team can show them what similar scopes have actually cost on past Beck projects. The ramp-up time that costs you weeks on every project gets compressed because Beck's cost knowledge is in the system, not just in people's heads. No architect-specific design details or client information leave Beck's private environment.
Every completed Beck project reveals where specific material specs cost more than expected. Stone veneer that always runs over. Mechanical specs that never align with actual install costs. Material callouts that look reasonable on paper but land 20% high in the field. As Beck completes more projects, those patterns become visible in Beck's own data — which specs reliably underestimate, which line items carry hidden cost, and where the gap between design intent and construction reality is widest. Your estimators already know some of this from experience. The system gives them Beck's own numbers to back it up. No project-specific details, architect names, or client information are shared outside Beck's private environment.
You described it yourself: in the middle of a foundation, the owner decides they want a 15-foot water feature. Everything stops. Your team has to estimate something they have never priced at that site, get the owner a sense of cost, get the architect started on detailing, and keep the project moving — all while subs are sitting idle and your building window is shrinking. Today, that answer takes days or weeks of back-and-forth between your estimators and the architect. With Beck's historical cost data structured by line item and market, the system can pull comparable scope from past projects within minutes — what similar features have actually cost Beck on similar builds, what trades and materials were involved, and what the typical change order impact looks like. Your estimator gives the owner a data-backed range the same day. The architect gets a realistic cost target to design around instead of guessing. The project keeps moving. Every day a project sits idle waiting for a cost answer is general conditions money — and that cost is invisible until closeout.
This is not a future concept. It is a direct extension of the same reconciliation intelligence that starts with Beck's five projects. Every completed project teaches the system what construction actually costs versus what the design assumed. Multiply that across dozens of contractors and hundreds of projects, and the architect conversation changes permanently — for everyone on the platform.
This intelligence is getting built.
Every general contractor who joins the platform contributes to it and benefits from it. The benchmarks get more accurate with every closed project. The overrun patterns get sharper. The cost ranges get tighter. The architect feedback loop gets smarter. The early participants shape how it works — what gets tracked, how it is structured, what matters most in practice.
Beck would be one of the first companies in. That means direct input on how the intelligence layer is designed. That means your priorities — constructability, sub performance, architect coordination — reflected in the system architecture from the ground up. And that means access to every improvement from day one, rather than catching up later to a system that was built around someone else's workflow and someone else's pain points.
None of this requires exposing Beck's private data. The only inputs to the network layer are anonymized percentages, cost-per-square-foot averages, and overrun rates — stripped of every project name, company name, sub name, and dollar amount. How that protection works, precisely, is what we cover next.
Written for your IT team and your partners — plain language first, technical detail below.
You told us your IT team wants your data handled securely and your partners need to approve before any data is shared. That's exactly the right instinct — and here is how we address it.
Every project Beck shares is stored in a private, isolated account. Your estimates, your sub bids, your cost reports, your closeout data. No other company can access it. The database architecture enforces this at every level.
Another company using the same platform cannot access Beck's data. Not your project details, not your sub pricing, not your estimates. The isolation is structural, not just a permission setting.
Beck's project files, estimates, bids, and cost reports belong to Beck. You can request a full export at any time in a standard format your team can open and use. We are stewards of your data, not owners of it.
Nothing moves until both parties have signed a mutual non-disclosure agreement. We have a standard NDA ready, and we welcome your counsel's review or amendments.
Only the specific people named in the NDA can access Beck's data. During the initial five-project build, that is four people. After the build phase, it reduces to three. Access is gated by name and role — not open to a broader team.
You get a complete copy of every project file you shared. All Beck-specific records are deleted from the Supabase database with written confirmation.
Your raw project data sits in your private account. But the value of this platform is what it turns that data into. Here is what Beck gets from its own project history, visible only to Beck:
Which divisions consistently come in over or under budget? At the trade category level, you see the big picture. At the line item level, you see exactly which materials and scopes are driving those variances — ceramic tile running 12% over, engineered lumber consistently under. Two levels of intelligence from your own completed work, visible only to your team.
Which subs are most predictable? Which ones struggle on certain project types? More specifically — which scope items within a sub's work are causing problems? Which line items generate the most change orders, the most RFIs, the most warranty callbacks? Your sub relationships and their specific rates stay completely private. The performance patterns — traced from the original bid line item through every downstream event — help you make better decisions on the next project.
The judgment your estimators carry — how this project feels like that project — gets structured and retained. New team members inherit decades of Beck-specific experience from day one.
How Beck's Vail numbers compare to Beck's Denver numbers. Altitude, access, labor market differences — applied automatically based on project location and your own history.
Everything above comes from Beck's data and is visible only to Beck — at both the trade category level and the specific line item level. No other company contributes to it or benefits from it. This is your private intelligence layer.
In addition to your private intelligence, the platform also maintains a regional benchmark layer. This is a separate system that answers a different question: what does mountain construction typically cost across the broader market?
This raises a fair question: where does that benchmark data come from, and does Beck's data feed into it? The answer is yes — but only in a specific, controlled way. Here is how it works, precisely.
Yours alone. Everything Beck shares stays here.
Belongs to Beck. No one else can access it. Think of it as a private vault — only Beck's team and the people Beck authorizes can open it.
Completely separate. Contains only statistical averages at two levels of detail.
No project names. No company names. No raw dollar amounts. No addresses. No identifying information of any kind. Every published number is a statistical average with built-in variance protection. A specific project cannot be reconstructed from these outputs.
Beck's private database gets smarter as your own project history grows — without depending on any external data source.
Regional benchmarks calibrate what "normal" looks like for mountain construction — so when Beck's numbers diverge from the regional pattern, the system flags it as an early warning, not a surprise at closeout.
Beck's data makes the benchmarks more accurate — but is never exposed through them. No competitor, no partner, no outside party can see a trace of Beck's specific numbers in the regional layer.
Your subcontractor relationships are protected at every level. The regional benchmark layer never reveals which sub bid what, or the specific rates your trade partners gave you. Vendor performance data is only visible to GCs who have used that same vendor, and only when five or more contractors contribute. What enters the anonymous benchmark is statistical averages at the trade category and line item level — stripped of all vendor names, project names, dollar amounts, and identifying detail. The relationships you've built over decades stay yours.
What you get back: A complete export of all project data within 30 days. Every project file, every estimate, every cost report you shared. All Beck-specific records are deleted from the Supabase database with written confirmation.
What stays with the platform: The regional benchmark layer — which contains no identifiable project, company, or individual information — is retained as a platform asset. This is consistent with how all enterprise software platforms operate. The statistical averages that remain carry no trace of who contributed them and cannot be reverse-engineered to reconstruct any specific project.
This distinction is documented in the data processing agreement and mutual NDA. Both parties know exactly what belongs to whom before any data is exchanged.
The team that would work with Beck.
Forty years in the construction industry across plumbing, electrical, general contracting, and development. Started and grew a company to 140 employees before selling it to a public utility. Since then, construction management consulting for municipal and government projects.
Skip brings the construction credibility to this work. He knows what an estimator is actually doing at the SD stage, what a sub bid is really telling you, and where the pain lives in the process — because he's lived every part of it. He's the reason this tool is built around how construction actually works, not how software companies think it works.
His background in government construction includes projects requiring federal security clearance. That background shapes how seriously he takes data handling and confidentiality.
Data systems and software architecture. Built the software application underlying this tool — the system that takes Beck's completed project history and turns it into structured, actionable patterns. Eagle County based — across the street from Beck's Eagle office. Not a remote team pitching from a coast.
The combination that matters here: deep construction domain knowledge from Skip, technical execution from Garry. Neither of us is trying to understand the other's world from a distance. We're translating between them every day.
Prior work includes building data systems and software architecture for clients at scale. The focus here is narrow: one niche, mountain GC cost intelligence, built right before expanding.
CEO of Vaadin, an enterprise software platform serving over 1,000 paying customers including Apple. Four successful exits across software, banking technology, and enterprise platforms. Started his career building software for contractors — his first company, ACS, was purpose-built for the construction industry. Later built banking software that scaled to 2,000 community banks nationwide.
Steve advises on enterprise scaling, go-to-market strategy, and operational architecture. He has built and sold companies at the exact intersection we operate in — niche vertical software that becomes critical infrastructure for the businesses it serves. His role is making sure we build this the right way from day one, not just for Beck, but for the long term.
Director of the Project Management Office at Domino Technologies, an IT staffing and consulting firm serving PennDOT and Pennsylvania state government. Previous PMO Director roles at Ciber and Applied Educational Systems. U.S. Air Force veteran, Reports and Analysis. BA from Goddard College, Master Certificate from Villanova.
Garry Sr. brings enterprise project governance to this team. His background is standing up PMOs from scratch, building delivery frameworks, and making sure nothing ships without proper review. He oversees quality assurance, visual review of all deliverables, and ensures the process between our team and Beck stays structured and accountable.
Executive Director of Acclaim Systems, an IT services and software firm serving government clients. SBA 8(a) certified, GSA Schedule holder, Microsoft Gold Partner. BBA in Finance from Drexel University's LeBow College of Business. Background spans ERP solutions, business intelligence, data normalization, and enterprise operations.
Trish is responsible for validating how Beck's data gets structured inside the platform. When an estimator's spreadsheet comes in with line items labeled differently than the next one, Trish makes sure the normalization is accurate — that "Window Restoration" maps to the right AIA division and the numbers land in the right columns. Her ERP and BI background is an exact fit for making sure the data layer is clean from day one.
Who sees Beck's data — by name: During the initial five-project analysis, four people will have direct access to Beck's project data: Garry Dubbs, Garry Dubbs Sr., Patricia Dubbs, and Skip Cox. Skip's involvement is specifically for the initial structuring phase — his 40 years of estimating experience is what ensures the data gets mapped correctly into the system. Once the initial analysis is complete and the system is running, ongoing access is limited to Garry Dubbs, Garry Dubbs Sr., and Patricia Dubbs. Skip does not retain access after the build phase. Steve Grandchamp advises on strategy and operations without data access at any phase. All named access is documented in the mutual NDA.