Harbourstone Construction Group Ltd.

Technology decisions at Harbourstone are made in the truck cab, project by project, usually when a general contractor requires the firm to use Procore on a specific job. There is no digital roadmap and no one owns systems as a mandate. A Buildertrend trial was started last year and abandoned when the busy season hit, which is the common pattern: a tool is bought on enthusiasm, rolled out to one champion, and never institutionalized. The owner carries the operating picture in his head, which works at the current size but does not scale and does not survive a key person being off. Software is treated as a project expense rather than as infrastructure that compounds.

Reactive buying means the firm pays for tools it does not fully use and repeats onboarding costs on every new client platform. Without a roadmap, each digital effort starts from zero, stalls, and leaves the underlying margin problem untouched.

A one-page digital roadmap with a named systems owner, sequencing estimating-to-accounting integration first, so every tool decision serves one plan rather than one job.

• Low Name a systems owner and write a one-page 12-month digital roadmap tied to the two biggest margin leaks 0 to 1 month
• Low Set a standing quarterly review of the roadmap against actual project margins to keep it honest Ongoing, quarterly

Job-costing lives in QuickBooks and a set of spreadsheets that are updated after the fact. There is no live work-in-progress view, so the firm cannot see cost-to-date against budget on an open job without a manual reconciliation. Committed costs from subtrade purchase orders are not booked until the invoice arrives, which can be weeks after the commitment was made, so the budget always understates true exposure. The rich estimating history sitting in past takeoffs is never mined to sharpen future bids, which means the firm re-guesses unit costs it has already paid for many times over. Reporting to the owner is a monthly look backward rather than a forward-looking instrument.

Margin erosion goes undetected until it is locked in. A job trending 8 points under bid looks fine on the books until the final subtrade invoices land, by which point the loss is real and unrecoverable rather than a variance someone could have managed.

A live WIP and committed-cost view per job, refreshed weekly, with estimate-versus-actual reconciled so margin drift is visible while there is still time to act on it.

• Medium Stand up a weekly WIP report reconciling estimate to actual and committed costs on every open job 1 to 3 months
• Medium Start capturing subtrade purchase orders as committed costs at the point of commitment, not at invoice 1 to 2 months

The stack is capable but siloed. Takeoffs are done in PlanSwift and Bluebeam Revu, the books run in Sage and QuickBooks, and daily logs come off the field on paper or as texted photos. None of these are integrated, so the estimate never flows into the budget and the field never flows into the ledger without someone re-keying it. The firm runs an aging on-premise server for Sage with a shared drive that is the single point of failure for every project file, and backup discipline is informal. Superintendents capture site information on whatever is at hand rather than in a structured field app. The tools to close these gaps exist and are partly owned already, but they sit as islands.

Integration gaps force manual re-entry that costs administrative hours and injects errors into costing and billing. The aging server and informal backups put every active project file one hardware failure away from a serious disruption.

Estimating output flowing into the job-cost budget, a structured field app replacing paper logs, and project files on backed-up cloud storage rather than a single office server.

• Medium Integrate estimating takeoff output into the job-costing budget so bids populate budgets without re-keying 2 to 4 months
• Medium Migrate project files off the local server to backed-up cloud storage with enforced backup 1 to 3 months

The workforce is strong on the tools that matter for building. The gap is that digital systems have no owner and no rhythm. The project manager is the de facto administrator for scheduling, costing, and client updates, and is overloaded to the point that anything requiring a new tool gets deferred. Younger field staff would readily use a mobile daily-log or timesheet app and are frustrated by paper, while some senior hands are wary of change, so adoption stalls in the middle. There is no training plan when a tool is introduced, which is why past rollouts died. The talent to run better systems is on the payroll; the structure to support them is not.

A single overloaded PM as the systems bottleneck caps how many jobs the firm can run well and creates key-person risk. Failed rollouts also breed cynicism that makes the next adoption harder.

A named systems owner distinct from the lead PM, plus a short structured onboarding whenever a new tool goes live, so adoption sticks past the busy season.

• Low Assign systems ownership off the PM's plate and give that person time to run adoption 0 to 1 month
• Low Pair every new tool with a half-day field and office onboarding rather than a link and a hope Per rollout

The operational core is manual. Change orders are written on paper or in email and take weeks to reach costing, so the budget lags reality. RFIs move by email with no register, so items get lost and rework follows. Subtrade coordination is done by phone, which means the schedule lives in the caller's head and in an Excel sheet or a whiteboard rather than in a shared plan everyone can see. When a field condition changes and the information does not reach the office in time, the result is rework, which is the most expensive category of waste on any site. There is no standard closeout process, so lessons from one job rarely improve the next.

Slow change-order flow means work gets done before it is priced and approved, which erodes margin directly. Lost RFIs and stale schedules drive avoidable rework, and rework on a single significant item can erase the profit on a job.

A tracked change-order and RFI register with clear status, and a shared schedule visible to office, field, and subtrades so coordination stops depending on phone tag.

• Medium Move change orders and RFIs into a tracked register with status and dollar value so nothing gets built unpriced 1 to 3 months
• Medium Publish a shared project schedule that subtrades and field staff can see and update 1 to 2 months

This is Harbourstone's strongest axis and its buffer. The firm holds durable relationships with regional general contractors and public owners, delivers quality, and earns repeat work on reputation. The exposure sits in governance. COR safety documentation, WHMIS records, and toolbox-talk sign-offs are kept on paper in binders, which are slow to assemble and risky at audit time. Client reporting is ad hoc and built by hand for each request, which consumes time and varies in quality depending on who produces it. Because so much compliance evidence lives in physical form and in individuals' routines, a lost binder or a missed sign-off becomes a real problem rather than a minor one, and it is the kind of failure that can jeopardize a prequalification.

Manual safety records put COR certification and prequalified status at risk, and losing either can close the firm out of the public and larger private work that anchors its book. Hand-built client reporting also quietly consumes senior time every week.

Digitized, on-demand safety and compliance records tied to each job and worker, plus a standard client reporting format, so audits and updates are a lookup rather than a scramble.

• Medium Digitize COR, WHMIS, and toolbox-talk records into a searchable system tied to job and worker 1 to 3 months
• Low Standardize a client progress report template so updates are consistent and fast to produce 0 to 1 month