Front-End Planning: How to Set Billion-Dollar Projects Up for Success

Episode Overview

Front-end planning is where billion-dollar projects are won or lost, and most people in the industry arrive after those decisions are already made. Orion Matthews hosts Roger Farish, a mechanical engineer who spent 25 years on the contractor side of mega projects at Bechtel, Fluor, Linde Engineering and Kiewit before founding Roman Consulting Group to work exclusively on the front end. Over an hour, Roger gives the FEL-and-stage-gate 101, unpacks the Bechtel-published figure that 98% of mega projects miss at least one objective (cited from memory, marked directional below), tells stories from inside PDRI scope reviews where “complete” deliverables turn out never to have been started, and lays out how to build a career in the phase where influence is highest.

Episode Timeline

The conversation as it unfolded: seven chapters across 62 minutes, following the show’s four-act arc: intro, topics, advice, wrap-up. The strip maps each chapter’s share of the runtime; timestamps are approximate. Click any block to jump to its chapter.

Who carried the mic

The Guest: Roger Farish

Roger Farish

Owner & Principal Consultant, Roman Consulting Group · 25 years across Bechtel, Fluor, Linde Engineering & Kiewit

Sugar Land, TX · Greater Houston

Front-End PlanningCost EstimatingRisk WorkshopsPDRI FacilitationOwner’s Engineer

Roger is a mechanical engineer by trade (bachelor’s and master’s from Tulane, and a Registered Professional Engineer) who deliberately chose the contractor side because that’s where the mega projects were. His 25 years run from six years at a self-contained Bechtel DOE site in Aiken, South Carolina (design engineering, then running construction work packages in the field), through eight and a half years at Fluor in Greenville and Houston (power, chemicals, biotech, mining, a functional assignment overhauling project-management processes, and an MBA from Chicago Booth along the way), to Linde Engineering in Tulsa and then Houston, where as Vice President he oversaw all North American EPC projects (a roughly $1B portfolio spanning four US offices), and finally seven years at Kiewit, leading lump-sum EPC bids, FEED efforts and FEL-1/FEL-2 for large mega projects, latterly directing multi-billion-dollar energy-transition studies and Kiewit’s rollout of Aspen Capital Cost Estimator. In October 2024 he founded Roman Consulting Group in Sugar Land, Texas, where he helps owners, developers, EPCs and OEMs reduce cost and uncertainty on the front end: independent estimating, scope definition and PDRI reviews, risk workshops, governance evaluations, and managing entire front-end phases as owner’s engineer.

“You look back at all the projects you’ve done, and you see where you’ve had the ability to influence the project the most, and it’s always at the front end. The decisions made during that development phase really shape the project, and can sometimes make or break a project.”Roger Farish · this episode, ~8:30

Career journey

Sequence as told on the record; years, titles and locations from Roger’s LinkedIn bio. Timeline bands are approximate to the year.

The Host: Orion Matthews

Orion Matthews

Founder, Queryon · Host, The Major Project Podcast · PMO data & analytics

Eugene, Oregon, United States

Project Controls ReportingPMO Data AnalyticsCapital PortfoliosAI in ProjectsPower BI

Orion is an award-winning programmer, entrepreneur and technology executive who has spent the last decade in the engine room of major capital projects. He founded Queryon in Washington, D.C. in 2015 with a focus on data analytics and software for the construction industry, starting with bellwether client BP, whose project-controls solutions his team deployed globally, from the North Sea to Azerbaijan to Houston. Today he's a recognized expert in PMO data analytics and project-controls reporting, serving billion-dollar portfolios for owners including BP, Santos, Balfour Beatty and Turner & Townsend, a regular speaker at major industry conferences on AI and analytics in capital projects, and the host of this show.

The Road to the Front End

The welcome, and a career told at speed. Roger came out of Tulane with a bachelor’s and master’s in mechanical engineering knowing he wanted big projects, and picked the contractor side on purpose, because the big contractors made their living on the very largest jobs. Twenty-five years later, when it came time to hang out his own shingle, he chose the front end just as deliberately.

“I got drawn more to the contractor side because I think they did a better job at selling these mega projects… The big contractors really make a lot of their profit, on a yearly basis, on the very, very large projects.”Roger Farish · ~1:45

1. Owner side or contractor side? The college-days fork in the road, and why the contractors’ mega-project pitch won.~1:30
2. Bechtel: design engineer in a specialized group, then quickly out to a job site, running and leading multiple construction work packages. “It was just so rewarding.”~2:50
3. Fluor, Greenville then Houston: power, chemicals, biotech, mining, plus a functional assignment feeding lessons learned back into the firm’s project-management processes.~3:30
4. Kiewit: lump-sum EPC bids, a co-led joint-venture pursuit, leading FEED efforts, then FEL-1 and FEL-2 for large mega projects, and rolling out estimating software to sharpen early-phase numbers.~5:20
5. The consulting leap: “helping clients reduce cost and uncertainty on the front end of projects so they can ultimately make better investment decisions.”~6:30
6. Why the front end, out of everything? The influence answer, and the exponential framing that names the episode.~8:30

Front-End Planning 101

Most listeners arrive on a project after sanction, so Orion asks for the 101. Roger maps the vocabulary: a stage-gate, waterfall-style approach where FEL 0/1/2/3, concept–feasibility–select–define, and pre-FEED/FEED are mostly different names for the same disciplined march toward an investment decision.

“The whole idea is that organizations are taking a disciplined and structured approach to making those decisions, so that they’re making the best possible decision on how to invest their capital.”Roger Farish · ~12:10

1. The terminology decoder: FEL 0–3 ≈ initiate/assess ≈ concept–feasibility–select–define ≈ pre-FEED and FEED. Different owners, different labels, “very similar meaning.”~10:30
2. Orion plays it back (steps intersected by gate events, hold/advance decisions) and Roger confirms: interchangeable-ish, custom per organization.~12:40
3. The executive scenario: “we just finished this project, it’s way over budget… we’re gonna get the best front-end planning process ever.” Roger’s first question back: one-off project, or portfolio? The governance answer differs completely.~15:40
4. Objectives differ too (schedule-driven speed-to-market versus measured, sustaining growth), and the process must be fit not just for the organization, but for the organization at that moment in its maturity.~17:10
5. The culture catch: a new gate process is an organizational-change project. “You need buy-in from executives, from different layers of management… you can’t just go from one side of the coin to the other.”~17:50

Too Big to Start

Orion floats “too big to fail, or too big to start?” and Roger takes the second half seriously: as projects grow, interfaces multiply, and every interface adds cost and risk. This chapter holds the episode’s hardest numbers, its best de-risking story, and the bias that keeps doomed projects rolling.

“As projects get bigger, the risks become exponential… Bechtel published that 98% of mega projects, a billion dollars and above, fail to meet at least one of their objectives, whether it’s cost, schedule, or another very important project objective.”Roger Farish · ~19:40

1. Why big equals risky: every added interface carries added cost and added risk: “these monsters of projects that have all these interfaces really are the riskiest of projects.”~19:10
2. Execution bias, named: big projects are exciting economic events, so organizations lean toward “go.” Roger’s counter is an objective outside view armed with the client’s own history.~21:30
3. The method: separate systemic risk from event-driven project risk in the owner’s historical data, so there’s a point of reference for what to expect going forward.~23:50
4. Data reality: some owners have lots of data, some little, and the rich datasets are often messy: captured one way on one project, another way on the next. Quantitative analysis plus interviews, never a single data point.~25:20
5. The payoff: “No owner data that we have reviewed to date has told us there’s nothing to learn here… In the end, that’s the golden nugget.”~26:10

Inside the PDRI Room

Orion sets a scene (notebook in hand, walking into the C-suite at a stage-gate moment) and Roger immediately corrects the frame: he supports gate reviews rather than sitting the gate itself, because the real conversations happen through the process. What follows is the episode’s richest stretch: how PDRI scope reviews actually work, and how experienced facilitation catches the gaps a gate checklist never would.

“You have no idea how many times the project manager claims that the deliverable is complete, issued, final, but sometimes the deliverable has not even been started.”Roger Farish · ~30:40

1. The gamer’s gambit: PMs push for lower scores so the project clears the gate threshold, sometimes knowingly, sometimes genuinely unaware the deliverable isn’t what they think it is.~29:30
2. Why the bias is human: “the project manager loves his project, of course, and wants it to keep going forward, and he’ll do anything and everything to keep moving that project forward.”~31:10
3. Two facilitation formats: risk workshops pair a group session with confidential one-on-one interviews; PDRI sessions don’t get the private channel, so the live facilitation has to carry all the truth-finding.~32:30
4. Where gaps hide: people outside the task force. On tight task-force projects Roger maps who’s in and who’s out; the outsiders give the answers the core team doesn’t expect.~34:00
5. The output isn’t a gotcha. It’s alignment: each surfaced gap becomes a PM action to reconnect with the disengaged discipline before it becomes a scope hole.~35:40
6. A cultural aside: communication norms differ across South America, Asia and Europe (even Munich vs Dresden inside one company), and a facilitator has to account for it.~37:30

Data, AI & the Hype Filter

Back to data: this time the industry’s. Roger walks the landscape of subscription cost datasets, the IPA wall between owners and consultants, and then the AI question every conference is shouting about. His answer is unusually honest for 2025: mixed results, one genuinely great use case, and a sharp eye for rebranding.

“I hear a lot of talk, but to me the proof is in the pudding… I think some people are calling a lot of rules-based processes AI, and I think it’s a stretch.”Roger Farish · ~45:10

1. The data landscape: really good US cost data, thinner internationally, which partly tracks risk itself, since labor drives North American cost in a way it doesn’t in Asia-Pacific or the Middle East.~41:20
2. The IPA wall: owners can access IPA’s benchmarking; consultants can’t. And contribution cuts both ways: “if you’re not contributing all the data, that means other organizations are not contributing all the data. And then what does that mean?”~42:30
3. Mixed AI results in analysis so far (no platform good enough yet), but real leverage in report writing.~43:20
4. An “AI-driven” early-design bolt-on under evaluation turns out to be… mostly rules. The label test: is this learning, or a decision tree in a new outfit?~45:20
5. Orion’s parallel from the cloud era (telling exchange-server admins “you are a cloud engineer now”), and the Salesforce-homepage-overnight-AI-rebrand bit.~46:40

Breaking Into the Front End

The advice act. For students and mid-career engineers who heard the first forty minutes and thought “that’s where I want to work,” Roger lays out the paths in (from the owner side and the contractor side), plus the skill stack colleges rarely assemble for you, and the resources he uses daily.

“Being a generalist in the front end of projects is very important, and having that lifecycle experience… Push your manager to give you things that are outside of your comfort zone, maybe those not-so-desirable assignments, so you can draw on those experiences and ultimately get involved in the front end.”Roger Farish · ~51:00

1. Owner-side on-ramp: process improvement on the stage-gate machinery itself: younger engineers can test and improve governance without knowing every project in and out.~48:50
2. Contractor-side on-ramp: process engineers are involved earliest; and non-process engineers can get there via startup assignments, where you absorb process knowledge by necessity.~49:30
3. Orion’s college note: engineers can graduate without ever meeting the time value of money. Being taught the finance models beats learning them on the go.~53:20
4. “I would add statistics to the list as well”: for risk, and even for the economic models.~54:00
5. The book pitch: “Even though it’s called Capital Projects, it’s all about the front-end planning of projects… geared towards executives, but very accessible to anybody interested in capital projects.”~57:10

The Close & the Cannonball

The wrap: when to call Roger (earlier than you think), how Roman CG engagements tend to snowball from a small capital project into full front-end stewardship, and Orion’s closing metaphor, which lands the episode’s influence-curve thesis in a single image.

“The less you’re spending, the earlier in the project, the higher outcome of success, kind of like dialing in a cannonball before you shoot it. You want to make sure the aim is correct before you apply all the force: energy, money…”Orion Matthews · ~61:10

1. “The earlier we get in touch, usually the better,” and how a sustaining-capital engagement becomes PDRI reviews, risk reviews, then managing the FEL-1 contractor when the owner is short-staffed.~58:20
2. Beyond owners: EPCs without front-end capabilities, and OEMs who own most of a developer’s scope and need help framing the project.~59:30
3. Where to find him: very active on LinkedIn, plus the contact form at romancg.com, and at industry conferences.~60:20
4. Orion’s takeaway: the influence curve is inverted from the spend curve: by the time the money is flowing, your ability to change the outcome is already draining away.~61:00

The Big Ideas

Three ideas carry the episode. Sketched here the way a front-end war room would whiteboard them.

Quote Bank

The best on-the-record lines, gathered in one filterable place. “Highlight” marks the lines most worth scrubbing back for; timestamps are approximate.

Bookshelf & Links

Everything recommended in the episode, plus where to find the people and the show.