I
Portfolio governance, stood up
One register of what is funded, who owns it, and what it is worth to the group. Steering committees that decide rather than receive updates. Thresholds and stage gates that hold when a sponsor pushes.
Leave no doubt
Visibility and control over the initiatives that matter.
Doubless gives leadership a single honest picture of every project in the group — what it costs, who owns it, and whether it will land. Decisions get made on evidence rather than on the last confident update.
Book a discovery call
or write to hello@thedoubless.com
The problem
Everything is reported green until it isn't.
Groups rarely fail for want of effort. They fail because forty initiatives are running at once, each with its own version of the truth, and no one can say which five actually matter this quarter.
The board receives a pack it cannot interrogate. Sponsors sign off on work they never see again. Delivery teams escalate late, because escalating early has never been rewarded.
Which projects are genuinely at risk? What did we approve, and what did we get? If we stopped this tomorrow, who would notice?
What we build
Governance your people will actually use.
Not a framework binder. A working cadence — one register, clear decision rights, gates with teeth, and reporting the board can read in a single sitting. Designed with the people who have to live inside it, then handed over.
Practice
Four engagements
Each one leaves the group with something it can run without us.
II
Independent delivery oversight
For programmes already in flight — core banking, ERP, operating model change. Honest status, early warnings while they are still cheap to act on, and a costed route back to plan.
III
Process and AI-enabled improvement
We map how the work actually moves, then apply automation and AI where it pays. Every opportunity is named, sized and tied to the revenue line or the cost base before anyone builds anything.
IV
Governance-in-a-Box
A packaged PMO for groups starting from nothing: charter, registers, RAID, stage gates, reporting pack and meeting cadence. Configured to your business in weeks rather than quarters.
Where we work
Built for the way groups run here.
Family conglomerates, holding companies and regulated institutions across the GCC — where ownership is concentrated, decisions carry personal weight, and the governance still has to satisfy a regulator. We have spent eighteen years inside that.
Our purpose
Certainty, in a place that rewards it.
Doubless exists so that leadership never has to decide on a guess. Spoken aloud, the name is doubtless — and that is the whole product: a single defensible picture of the portfolio, current enough to act on and honest enough to trust.
One version of the truth
Every initiative in one register, on one set of definitions.
Decisions on the record
Who approved what, on what evidence, and what happened next.
Audit-ready by default
Built for regulated environments, where governance has to survive scrutiny.
Handover, not dependence
Your team owns the cadence when we leave. That is the point.
Approach
How an engagement runs
Short, sequenced, and built to hand over. You should not need us permanently.
Diagnose
Interviews with sponsors and delivery leads, a read of every active project, and a scored view of where governance is actually failing.Two to three weeks
Design
The operating model for the portfolio: decision rights, thresholds, gates, forums and the reporting that feeds them. Agreed with the people who will run it.Three to four weeks
Stand up
We run it ourselves for a full cycle — chairing the forums, producing the pack, working the exceptions — so it is proven against your real portfolio.One to two quarters
Hand over
Your PMO takes the cadence, with the artefacts, training and a written playbook. We stay available for board reviews, and nothing else.Four weeks
Diagnostic
Project Portfolio Healthcheck
A structured review of every active project in the group, scored against the ten dimensions that predict whether work lands. It produces one page the board can act on, and a ranked list of what to stop, fix or fund.
- Sponsorship and decision rights
- Benefit definition and ownership
- Funding against stage gates
- Scope and change control
- Delivery capability and capacity
- Dependency and sequencing risk
- Reporting integrity
- Risk and issue discipline
- Vendor and contract exposure
- Regulatory and audit readiness
Most groups find a portion of the portfolio consuming capacity without a named owner or a defined benefit. The healthcheck usually pays for itself in what it lets you stop.
Case notes
Retail · ERP transformation
Healthcare · Artificial intelligence
Technology · Algorithmic pricing
Three programmes that were stopped, and what a gate would have asked.
None of these organisations lacked money, talent or intent. Each ran a large multi-year programme that was eventually written off. In every case the decision to stop was available years before it was taken. These are public failures, documented in audits and filings, chosen because the governance lesson in each is different.
A seven-year ERP programme, abandoned at €500 million
€500m
A European grocery group began replacing its in-house merchandise management system in 2011. The rollout reached several countries and the vendor publicly recognised the group as a leading customer. In 2018 the board stopped the programme and returned to the legacy system. Reporting traced the core problem to a mismatch between how the group valued inventory and how the standard software did — a difference the group chose to solve by adapting the software rather than changing its own operating model. Costs rose from an original estimate near €200 million to roughly €500 million.
The question a gate would have asked
Once the decision was taken to bend the software to the operating model, the programme stopped being an implementation and became a re-engineering effort with entirely different economics. Nobody re-baselined it as one. A stage gate exists precisely to force that re-approval: this is no longer the thing you funded — do you still want it, at the new price?
Where we add value. We have run business-side vendor management on a four-market Microsoft Dynamics programme, where the same fit-versus-adapt decision arises in every module. Our design phase forces that call into the open at gate one, with the cost of each path priced before anyone commits, and writes a stop rule into the charter so that "we have already spent too much to stop" is not an argument the forum will accept.
Sources: Consultancy.uk, reporting Handelsblatt · RetailDetail
An AI decision-support build that never reached a patient
$62m
A leading US cancer centre contracted a technology vendor in 2012 to build an AI-based clinical decision tool. The original scope was six months and $2.4 million. By the time the work ended, the contract had been extended twelve times and roughly $62 million had been spent across the technology vendor and a consulting firm. A University of Texas System audit found the tool had been piloted against the old records system and never integrated with the hospital's current one, alongside procurement exceptions, delays and cost overruns. The audit was careful to note it took no view on the science. The system never entered clinical use.
The question a gate would have asked
Integration with the live records system was the condition on which the entire benefit depended, and it was never made an exit criterion for any stage. Twelve extensions is not a procurement failure so much as a governance vacuum: no forum was required to re-approve the spend, because the work had been classified in a way that placed it outside standard project governance.
Where we add value. Our background is in regulated financial services and risk consulting, where an engagement that cannot survive an internal audit is not finished. We write decision rights and procurement thresholds into the charter at the start, define what "done" means for each gate in terms the sponsor can test, and name the one dependency that, if unresolved, makes the benefit unreachable. On an AED 100 million infrastructure portfolio that discipline is what separated the initiatives that closed from the ones that drifted.
Sources: Journal of the National Cancer Institute · Forbes, on the UT System audit
An algorithm scaled faster than it could be proven
$881m
A US property platform built an algorithmic home-buying business on the strength of its own valuation model. It scaled aggressively, holding thousands of homes. In its November 2021 filing the company recorded a $304 million inventory write-down, having bought at prices above its own later estimates of resale value, and disclosed further expected charges. The programme was closed, about a quarter of the workforce was cut, and the segment ended 2021 with a loss of roughly $881 million. The chief executive attributed the decision to the unpredictability of forecasting prices exceeding what had been anticipated.
The question a gate would have asked
The model was not obviously wrong; it was deployed at a scale its error margin could not support, and the balance sheet absorbed the difference. The governance question is not "does the model work" but "at what volume does being wrong stop being survivable, and who is watching that number?" Scaling should have been gated on model performance in live conditions, not on commercial ambition.
Where we add value. We size the downside before the upside. A claims analysis we ran surfaced SAR 50 million in recoverable cost precisely because the numbers were interrogated rather than accepted, and reporting a USD 100 million book at CEO level teaches you quickly which figure the board will act on. For AI and analytics initiatives we set the exposure limit and the stop trigger at the design stage, and put a named owner on the metric that would tell you first.
Sources: SEC Form 8-K · GeekWire · Stanford GSB
The common thread is not technology. In each case a decision that should have been forced by a forum was instead made by default — by extension, by momentum, or by nobody at all. That is what the healthcheck looks for, and what the governance we build is designed to prevent.
About
Eighteen years of watching where programmes break.
The person who sells the work is the person who does it. There is no bench, and nothing is handed to a junior team once the engagement starts.
Shatakshi
Eighteen years inside banking, insurance and development finance across the GCC and Asia-Pacific — private and corporate banks, a sovereign development fund, an insurer, global broking. Layered structures, eight markets, several regulators. Programmes rarely fail on the technology; they fail at the seams between entities. Every portfolio here is read for those seams first.
Certified across the disciplines the work demands — PMP, Certified ScrumMaster, ITIL — and, more recently, MIT Professional Education's Applied Generative AI for Digital Transformation. The next decade of abandoned programmes will be AI programmes, and someone has to be able to govern them.
Based in Dubai. Works across the GCC.
How we work
Three commitments
Ethics
You get the status as it is. If a programme is failing, you hear it from us first, with a recovery plan attached — not a softened version that protects the relationship.
Empathy
Governance fails when it is imposed. We design with the sponsors, PMs and finance leads who have to run it, so the cadence survives after we go.
Efficacy
Every engagement is measured on what changed — decisions made faster, benefits actually claimed, work stopped that should never have started.
Start here
Leave no doubt.
Governance-in-a-Box is set out in a short document: what is included, how long it takes, and what your team owns at the end. Leave an email and it comes straight back to you.