IT Tech Pulse Exclusive Interview with Chris Ortbals, Chief Product Officer, Tangoe

Chris Ortbals, Chief Product Officer at Tangoe, discusses how AI-driven expense management, technology spend governance, and proactive cost optimization are reshaping enterprise IT financial management.

As IT spending becomes more decentralized and consumption-based — spanning telecom, cloud, SaaS, and now AI how is the discipline of expense management evolving, and what’s the biggest blind spot most enterprises still aren’t accounting for today?

IT spend used to sit in silos, with telecom here, cloud there, SaaS somewhere else, each with its own owner and tools. That’s breaking down. At the more mature enterprises, we’re seeing expense management consolidate into a single discipline that looks a lot like FinOps but stretched well beyond cloud. You can even see that evolution playing out in the industry’s governing bodies, not just in the tools themselves.

For large enterprises, a Technology Expense Management (TEM) platform has become table stakes. Organizations that try to manage this manually end up spending more, seeing less, and putting more strain on IT, finance and procurement than they need to.

The biggest blind spot right now is AI. Companies are getting real sticker shock from AI and token consumption spreading across the business, and most don’t have a way to see it coming, let alone manage it or calculate real ROI. We saw something similar in the early days of cloud a decade-plus ago but I don’t think this follows the same trajectory. AI spend is compounding faster, and the governance tools have a long way to catch up.

Beyond dashboards, what specific capabilities automation, audit tools, allocation engines — will separate the next generation of IT expense management platforms from today’s tools, and which one do you think enterprises are underinvesting in right now?

Dashboards got us this far, but they’re not what separates the next generation of platforms. What matters more is tighter integration between the enterprise and the TEM provider. A real connection into the workflows and systems enterprises already run, instead of just a portal we ask them to log into. We’re seeing more customers building their own AI layers internally and asking for a modern-day API into their ITSM or their own AI interface, so the two systems can work together.

That’s also the piece I think most enterprises are underinvesting in today: it’s less about the dashboards a vendor hands you and more about a hybrid model, where your platforms and your TEM provider’s platform interact directly.

The capability enterprises should watch for is really twofold: new things AI makes possible, like automated invoice and procurement intelligence and proactive cost optimization insights, and AI making the fundamentals such as inventory and asset automation, and service delivery, meaningfully better. The one thing we’re careful about as a provider is that this can’t just turn into “give us your data pipe.” The expertise and the scale a TEM partner brings still has to be the differentiator, which is what has kept Tangoe on top as a leader in this space for over two decades.

Tangoe works with 12 of the Fortune 20. What’s the single most surprising telecom billing error or overpayment pattern you’ve uncovered at that scale, and why does it keep slipping past finance teams undetected?

The most surprising pattern we see at this scale isn’t really a billing error, but the excess or unnecessary spending that’s out in the open and gets missed. Legacy services that no longer fit the business, capacity nobody has re-evaluated in years, stranded credits sitting untouched. It slips through the cracks because the invoice looks reasonable and the charges look familiar and nothing jumps out as wrong. Bad billing from vendors is a small part of the issue, whereas the bigger problem is businesses change faster than their telecom environment. The money is usually found in the gaps between what a company needs today and what it’s still paying for based on decisions made years ago. Oftentimes that money is significant at large enterprises that do infrequent service audits.

At one of our global information services customers, we found legacy DS1 circuits still being billed at high month-to-month rates for services the business no longer needed. Once they confirmed and disconnected them, the realized savings topped $1.6 million. At another large enterprise, usage on an 800 number was being billed outside the client’s contracted pricing because one state hadn’t been added to the contract. The intrastate usage was running around $40 a minute instead of under 10 cents. Discovering and fixing that issue dropped the rate and delivered nearly $1.4 million in savings. Neither charge was fraudulent, or even technically wrong on the carrier’s part since the usage and the legacy services were both real. They just weren’t what the business needed or had negotiated.

Telecom tax audits are notoriously difficult, spanning multiple jurisdictions, service types, and shifting tax treatments. Why do these errors go undetected for so long, and what’s driving enterprises’ renewed appetite for specialized tax recovery audits?

The short answer is that scale hides them well. Individually, these are small errors such as a rate applied to the wrong service or a jurisdiction that should never have been taxed in the first place. At enterprise scale, small errors repeated across thousands of invoices add up to real money, and most organizations simply don’t have anyone whose job it is to check the assumption behind the tax rather than the math itself.

Jurisdictional complexity compounds this over time. Telecom services cross states, counties and municipalities, each with different rates, exemptions and definitions of what’s even taxable. For example, federal USF fees, state sales and use tax, local franchise fees, 911 surcharges, all governed by different, frequently updated rules on a single invoice. Carriers’ billing systems weren’t built to keep all of that current in real time, so an error gets baked in and just repeats, invoice after invoice, for years.

Then there’s what we call “service-type blur.” As telecom shifted from plain voice and data to bundled cloud, UCaaS, SD-WAN and hybrid services, tax code hasn’t always kept pace. Something that used to be clearly “telecom,” and taxed one way, might now be partially an information service or software, taxed differently. Carriers don’t always go back and re-evaluate that classification once it’s set. We’ve seen global customers assessed taxes on facilities sitting in tax-free zones simply because nobody checked.

What’s driving renewed interest is straightforward: CIOs and finance leaders are under real pressure to find budget, and this is savings most of them didn’t know existed. Because the model is contingency-based (no findings, no fee),  there’s effectively no downside to looking. We recently identified $1.4 million in potential recovery for one enterprise customer in about 30 days, and most of these findings aren’t subject to a statute of limitations, so it’s compliance-driven work rather than a one-time cleanup.

As devices increasingly rely on multiple SIMs, carriers, and connectivity types, reconciling inventory and billing gets messier fast. What breaks down first when enterprises manage this manually, and how should teams rethink governance here?

This is a simple idea that gets very complicated at enterprise scale since many newer mobile devices are enabled to support multiple eSIMs. Multi-SIM breaks what inventory systems have relied on for years, that one line maps to one device. Once you allow multiple SIMs, carriers, and connectivity types to attach to a single device, or a single line to move across multiple devices, that one-to-one relationship doesn’t hold anymore. Inventory breaks down first and that is where teams lose the ability to confidently say which SIM is active, which device it’s tied to, who has it, and whether it’s compliant with policy.

Manually, this becomes a spreadsheet nightmare that doesn’t scale. There’s also a regional wrinkle worth noting, because in the U.S., multi-SIM plans often share one primary line with add-on fees for extra devices, while in Europe and the U.K. each SIM typically runs as its own line under a shared account. A single reconciliation approach doesn’t work globally.

The governance rethink is to stop treating line and device as a fixed pair, and instead build lifecycle management around the SIM itself. Provisioning, suspending, reallocating and retiring it as employees join, change roles or leave, with real-time visibility across every carrier and region in one place rather than disconnected portals. That’s the direction we’ve taken within our Tangoe One Mobile platform to support newer multi-SIM capabilities.

Gartner projects LEO satellite spend will hit $14.8 billion in 2026, up 24.5% year-over-year. Given how immature satellite governance still is compared to fixed and mobile telecom, what should enterprises prioritize first to catch up?

LEO is new, but it doesn’t need a new playbook. It needs the same discipline enterprises already apply to fixed and mobile telecom, just extended to a newer asset type before it sprawls. We see four things separate the enterprises managing satellite well.

First, they treat satellite as a standard WAN access type from day one, tracked alongside circuits and mobile assets in the same system and not in a spreadsheet or a carrier portal that only one person happens to have access to. Second, they validate satellite invoices against contracts, usage patterns and inventory records before approving payment, the same discipline applied everywhere else in telecom. Third, they build cost allocation rules while the satellite footprint is still small, because retrofitting those rules across a large, distributed environment later is a much heavier lift. Finally, they manage terminals and other portable hardware with the same lifecycle discipline as any other physical asset so they know where everything is, who has it, and what happens when a site closes or a contract ends.

Scale is really the crux of it. Four terminals across two sites is a simple spreadsheet problem. Fifty terminals across 40 sites, with variable billing and hardware constantly changing hands, is not. DIY processes can be unreliable as these things grow, and it’s a lot easier to put the right tools and controls in place beforehand than after the accounting gets messy.

AI spend touches providers, infrastructure, employee tools, and even autonomous agents. Of the four governance gaps discovery, allocation, control, and optimization which is proving hardest for enterprises to crack, and what does solving it actually look like?

Honestly, it’s close between allocation, control and optimization. Discovery is the one gap that isn’t the hard part anymore. The data itself is readily available; this is a modern technology stack, so in that sense it’s more like cloud than legacy telecom ever was.

Where I’d focus is allocation and chargeback, with control right behind it, because that’s where the value a TEM partner brings is clearest. The same standards that have applied for decades to expense management are exactly what’s needed here. We’re piloting AI cost management right now with a couple of enterprise customers, and both are driven primarily by allocation and control, with an eye toward ROI. They want to say, “this department or this project consumed this much AI spend, here’s what we allocated to you, and here’s the return on it.” Nobody has fully cracked control at a systematic level yet, but the ROI question is the one leadership keeps coming back to.

That’s a gap in how the four categories usually get framed: discovery, allocation, control and optimization don’t explicitly include ROI, but in practice it’s the thread running through all of them. Solving allocation without being able to answer “was this worth it” isn’t solving the problem enterprises actually have.

Looking three to five years ahead, how do you see enterprises redefining accountability for technology spend as AI agents begin making autonomous purchasing and consumption decisions on their own, without direct human approval each time?

I’d push back on the premise that we’re 3-5 years from agents making six- or seven-figure purchasing decisions with zero human involvement, and if we get there, it’s going to need significant oversight. The scale of the mistakes that are possible here is significant, especially with so much of technology being consumption-based. Without proper monitoring and controls in place, which AI itself can help provide, that risk multiplies. I don’t know that “human in the loop” is the permanent answer so much as a candid description of where we are today, but some layer of approval and accountability has to exist.

On the “who owns it” question, I don’t think it centralizes. Today, telecom spend mostly sits with finance and procurement, cloud and AI infrastructure mostly sits with IT and engineering, and SaaS is everywhere. AI is so easy to consume and stand up that I’d expect it to follow the SaaS pattern and decentralize further, rather than roll up under one owner.

Agents will absolutely be part of everyone’s day-to-day at some point. What I don’t think we’re close to is an agent autonomously sourcing services across a mobile fleet and booking commitments on its own. It might technically be possible to do, but what comes back is likely to have all kinds of problems and nuances that humans can see but agents are blind to. What’s more realistic in the near-term is smaller, well-bounded actions, such as an agent catching a billing error and flagging it for resolution.  That’s a genuinely useful use case, but a longer way from autonomous purchasing.

Thank you, Chris, for taking the time to share your insights with us.

Write to us [⁠wasim.a@demandmediaagency.com] to learn more about our exclusive editorial packages and programmes.