Chris M. / 11 August 2026 / Reviewed 11 August 2026

Your Business, Your Construction Platform: One Product or Something Bespoke?

A managing director, three tools deep into a Procore renewal conversation, asks the sales team a direct question: does this replace the accounting system too, or do we still need Xero running alongside it? The answer, on Procore's own marketing page, is telling. The platform describes itself as "a true platform," not "cobbled together point solutions," in the same breath as advertising "500+ integrations and apps," including named categories for accounting, estimating and CRM.¹ The pitch and the fine print contradict each other, and the contradiction is the actual shape of this decision.

No platform is actually one platform

This is not unique to Procore. Autodesk's construction product, since renamed Forma, claims to be "your single source of truth for every project" on the same page that advertises "over 400 pre-built integrations for ERPs, CRMs, document management, analytics tools, and more."² Oracle Aconex, the platform in this group that comes closest to genuinely covering the inspection and test plan workflow end to end, still positions itself as one part of a wider Oracle suite alongside Primavera and Unifier, not as a self-contained replacement for everything a contractor runs.³

The one real exception among the major players is RIB, and it is instructive precisely because of how it achieves the coverage. RIB does sell accounting, estimating and BIM takeoff under its own name, but not as one product. It sells thirteen separately branded systems, RIB 4.0, RIB BuildSmart, RIB Candy, RIB CostX and nine others, stitched into a portfolio rather than unified into a platform.⁴ Even the vendor that gets closest to all-in-one does it by selling you several platforms with a shared name.

The practical conclusion is that "should I buy one platform" is not really the decision in front of you. Accounting, estimating and CRM sit outside every product examined here except through integration, so the real choice is how much you configure a commodity platform, how much you connect several specialised ones, and how much you build to keep for yourself, and each of those has a genuinely different, checkable cost profile.

The Green Book test, applied to a construction platform decision

HM Treasury's Green Book gives the closest thing to a rigorous, publicly available decision framework for exactly this choice, and it is worth borrowing even though it was written for public spending. Its Options Framework Filter separates the decision into independent categories, scope, solution, delivery, implementation and funding.⁵ Ahead of that filter, the guidance insists one option always be carried forward regardless of how attractive the alternatives look: business as usual, which is not the same thing as doing nothing, but running your existing arrangement forward honestly rather than deliberately underselling it. The reasoning is explicit: "Overestimating the costs and risks of BAU will make government action appear unduly attractive."⁵ Any comparison that quietly makes staying put sound worse than it is has already decided the answer before the analysis starts, and the same discipline should apply to a construction firm comparing its current systems against a shiny new integrated platform, or against the promise of something built specifically for it.

The framework's costing window matters too. Software decisions are costed over ten years, covering "the whole lifetime of the proposal… from construction or development, through to operation or delivery, and then winding-down or decommissioning."⁶ Most vendor demonstrations show you month one. The real comparison is what each option costs to run, maintain, and eventually leave, across a decade, not a quarter.

Why the bespoke option is riskier than most people budget for

If the commodity platform genuinely will not do what you need, custom development is the fallback, and it deserves an honest look at what that has actually cost other organisations. The single most useful figure here is HM Treasury's own optimism bias guidance, still the recommended fallback under the current Green Book where a firm has no measured history of its own forecast accuracy.⁷ For a standard building project, the recommended upper-bound capital expenditure uplift is 24%. For an equipment or software development project, the same table recommends starting at 200%, a figure the guidance insists you should not reduce "without clear and tangible evidence of the mitigation of contributory factors" that has itself been independently verified.⁸ An eight-fold gap between building a warehouse and building software, in the government's own planning assumptions, is not a small margin of error. It is the government telling you, in writing, that bespoke software is a fundamentally different risk category from the physical construction most contractors understand intuitively.

The most rigorous recent academic study of this question goes further and finds that averages are the wrong tool entirely. A 2022 peer-reviewed analysis of 5,392 real IT projects, spanning 66 countries and 56.5 billion US dollars of spend, including 2,653 projects obtained directly from 104 US government agencies through Freedom of Information Act requests, concluded that "the average cost overrun for IT projects does not exist (i.e., cannot be calculated)."⁹ Cost overruns on software follow what statisticians call a fat-tailed power law distribution: most projects land close to budget, and a smaller number blow through it by an extreme, unpredictable margin, in a pattern where averaging conceals more than it reveals. The commonly quoted claim that most software projects run over budget by a fixed percentage should be treated with real caution for the same reason; the most-cited version of that statistic, the Standish Group's CHAOS report, was directly challenged in a peer-reviewed paper that tested it against 5,457 real forecasts and found the underlying method "misleading, one-sided" and productive of "meaningless figures."¹⁰

None of this means bespoke software always fails, and buying is not free of risk either. The same Green Book optimism bias table applies a 41% upper-bound uplift to outsourcing, which is a real reminder that handing a problem to a vendor does not remove risk, it relocates it.¹¹ The point is narrower and more useful than "never build": do not plan a bespoke construction system against a confident average cost, because the evidence says that average does not meaningfully exist, and a fat tail of extreme overruns is a real, well-documented possibility rather than a worst case someone is exaggerating to sell you a subscription instead.

What actually justifies building your own

The strongest, most defensible reason to build rather than buy is a contractual or regulatory requirement that is genuinely yours and that no commodity platform can be expected to anticipate. UK construction gives a sharp, concrete illustration of this. Four different UK infrastructure regimes specify inspection and test plans in four genuinely different ways: Network Rail mandates its own three-part proprietary forms by document number, National Highways supplies only a fill-in-the-blanks sample appendix rather than a template, and a shared industry body, CQIC, simply publishes contributed Word and Excel examples because no single template fits every contractor's arrangement.¹² There is no universal ITP standard to buy off the shelf, because the clients themselves have not standardised on one.

ISO 19650 reinforces the same lesson from the standards side rather than the client side. Its implementation guidance is explicit that exchange information requirements are "appointment specific," meaning each new contract can legitimately demand something a generic platform was never built to produce, and that the standard "does not mandate a template approach."¹³ Where a client has put a specific format in writing, and no platform's default output matches it, that is a real, testable justification for building something of your own, whether that is a bespoke report generator or a genuinely custom system. Where the justification is instead "our current spreadsheet process works differently and we like it," that is not the same kind of requirement, and the Green Book test above should be applied honestly before choosing to build around it.

The decision this guide actually recommends

Run three checks before either signing a platform contract or commissioning custom development. First, list which of your processes are contractually mandated to be exactly as they are, in the way Network Rail's own quality terms state plainly that ISO 9001 certification "does not demonstrate that the specified QMS requirements under this contract have been achieved."¹⁴ Only those processes justify the cost of configuration or a bespoke build. Second, cost every option, including doing nothing, over a genuine ten-year horizon that includes what it costs to eventually leave the platform or retire the bespoke system, not just what it costs to switch on. Third, if you do choose to build, budget from the Green Book's 200% upper bound rather than from a vendor's confident quote or your own optimistic estimate, and treat any claimed average industry overrun percentage as unreliable, because the best available evidence says that average cannot be meaningfully calculated at all.

Related reading

Two firms who chose the bespoke route have written up what it actually involved: we built our own CRM, what we learned and what building our own tools taught us. If the configuration limits on a named platform are what is pushing you toward building, can I tailor-make Procore or Asite covers exactly where those limits sit.

Where to check this yourself

• "How should I appraise a major spending decision properly?" HM Treasury's Green Book, the framework this guide borrows from, is free to read in full at gov.uk.
• "What does UK government say about avoiding vendor lock-in?" Free guidance on managing technical lock-in is published at gov.uk.
• "What does ICE say about digital transformation in civil engineering?" The Institution of Civil Engineers publishes ongoing coverage of digital adoption and BIM at ice.org.uk.
• "What does CIOB say about technology decisions in construction?" CIOB's industry policy and guidance hub is at ciob.org.

What could not be established for this guide

No independent, methodologically sound figure exists for how many separate systems a typical UK construction business runs day to day; every large number in general circulation traces back to a vendor with an integrated platform to sell. No UK-specific comparative study was found measuring outcomes for construction firms specifically choosing between commodity platforms, integration layers and bespoke builds, so the evidence in this guide is drawn from general public-sector IT project data and from construction-specific contractual evidence, not from a dataset of construction software decisions themselves. And pricing for every enterprise integration platform examined for this guide, Boomi, Workato and Celigo among them, is quote-only, so no genuine cost comparison between "buy one platform" and "buy several and connect them" could be built from published prices alone.