The RFP went out in late 2023. Three names on the shortlist: SLB, Halliburton, Baker Hughes. From the outside, it looks like a simple choice between three industry giants. The reality is a lot messier when you're the one signing the purchase order.

I'm handling procurement for a mid-stream compression project, and I've been in this game long enough to have made (and documented) more expensive mistakes than I care to admit. This decision, for a critical gas turbine package worth north of $2.5 million, was starting to feel like one of those moments where you could get it right or get it really, really wrong.

The Surface-Level Reading

People assume the lowest quote means the vendor is more efficient. What they don't see is which costs are being hidden or deferred. Our evaluation committee had spreadsheets comparing SLB, Halliburton, and Baker Hughes line by line. On paper, Halliburton came in slightly lower on the base package. SLB had the slickest digital services pitch. Baker Hughes was... well, they were the ones who kept saying 'we don't do that in-house, but here's who does it better.'

That didn't feel like a strength at first. It felt like a weakness.

The First Red Flag

The SLB presentation was textbook. They talked about their integrated service model, their digital twins powered by the DELFI environment, their ability to handle 'the full well-to-pipeline lifecycle.' The rep made it sound like if you bought their compressor package, you got a free ticket to an oilfield utopia. Halliburton's pitch focused on their track record in unconventional plays and their 'Landmark' software suite. Both presentations made the same implicit promise: we can do everything for you, and it'll all work seamlessly together.

I bought it. For about three weeks.

Then I started digging into the references. I called a project manager I knew who'd used SLB's integrated gas compression package on a similar project in the Permian. His response was, in a word, qualified. 'The turbine itself is solid,' he said. 'But the integration? It took us four months to get the control system talking to our existing SCADA. And the aftermarket support for the non-core components was outsourced to third parties anyway.'

The question I should have asked sooner: what does 'integrated' actually mean when the integration points are brittle?

The Baker Hughes Meeting That Changed Everything

About a month into the evaluation, I had a meeting with Baker Hughes's local team. They were different. The sales engineer started by walking through their NovaLT™ gas turbine portfolio—their core offering. He talked about efficiencies, fuel flexibility, and their experience running turbines in similar environments. Solid presentation. Nothing earth-shattering.

Then came the twist.

'For the power generation controls,' he said, 'we'd recommend a specialist partner. Our team doesn't have the same depth in that specific HMI/PLC ecosystem. We can spec the interface points, but the implementation is better done by people who eat, sleep, and breathe that platform.'

I literally stopped writing notes. A vendor telling me they weren't good at something? That doesn't happen. You get the 'one-stop-shop' pitch until you sign, and then you find out the hard way that their 'one-stop' is actually a maze of subcontractors you didn't vet.

I asked him to clarify. 'You're saying you can't do the whole install?'

'We can do it,' he said. 'But the question isn't can we. It's should we. For this project, with these specific control system requirements, a specialist will give you a better outcome. Our job is to deliver a reliable turbine package with a guarantee. Their job is to make sure the controls meet your exact specs. We both win.'

The vendor who said 'this isn't our strength—here's who does it better' earned my trust for everything else.

The Real Decision Point

After that meeting, I went back and re-evaluated everything. The comparison between SLB, Halliburton, and Baker Hughes stopped being about 'who has the biggest catalog' and started being about 'who can deliver the critical path items without a hidden handoff.'

I started asking each vendor a specific question: 'Show me a project where your integration with a third-party supplier broke down. What happened? What did you learn?'

SLB's rep deflected. Halliburton's rep gave a generic answer about 'lessons learned.' The Baker Hughes engineer gave me a specific example from 2022 where a control system handoff went sideways because of a communication gap, and explained exactly how they'd changed their interface protocols since then.

That was the moment.

I still kept second-guessing after I made the recommendation to the steering committee. What if the specialist subcontractor Baker Hughes recommended dropped the ball? What if the coordination overhead ate up all the savings? The three weeks between placing the order and the first project review meeting were stressful. Didn't relax until I saw the project timeline: the turbine package was delivered on schedule, the integration points were clearly documented, and the controls specialist was already mobilizing.

What I'd Do Differently (and What I'd Do Again)

If I could go back and talk to my Q1 2024 self, I'd tell myself three things:

  1. Trust the vendor who admits a boundary. A company that says 'we don't do that' is more trustworthy than one that says 'we do everything.'
  2. Ask about the non-core components. The turbine is the star, but the controls, the valving, the piping interfaces—that's where the delays hide. Baker Hughes was transparent about that. SLB and Halliburton weren't.
  3. Check the references for integration failures. Everyone will give you a happy reference. Ask for a project that struggled with integration, and watch how they talk about it.

Even after choosing Baker Hughes, I kept second-guessing. What if their quality wasn't as good as the samples? The two weeks until delivery were stressful. But the project came in on time and on budget. The turbine package is running at 98.3% availability as of last month's report.

I'd rather work with a specialist who knows their limits than a generalist who overpromises. That's the lesson I learned from this procurement, and it's one I'll carry into every decision going forward.