Value a telecom company with recurring AI services by forecasting its cash flows and checking the result against comparable businesses—not by applying an assumed “AI premium.” The key question is whether AI services add durable, profitable cash flow after customer-retention costs, service delivery, compute and software expenses, and required investment. Because disclosures may bundle AI into broader service lines, a defensible valuation must distinguish what is reported from what cannot be isolated.
Set the valuation scope before choosing a method
First define what is being valued: a listed operator, a private company, a business unit, or an asset. Specify the geography, reporting period, currency, valuation purpose, and whether the output is enterprise value or equity value. A fiber or tower asset, for example, has a different capital and revenue structure from a service-heavy operator; comparing them without adjustment can mislead.
Use at least two lenses: a discounted cash flow (DCF) analysis, which makes forecasts and assumptions explicit, and a comparable-company analysis, which checks the result against businesses with similar economics. Lumen’s 2025 annual report describes both DCF and market approaches and notes that the appropriate approach depends on the facts and circumstances (Lumen 2025 annual report).
Separate reported business lines and test what “recurring” means
Reconstruct the company’s revenue and costs using its disclosed segments. Useful categories may include consumer and enterprise connectivity, wholesale, infrastructure, digital services, and AI-enabled services. Reconcile adjusted or non-GAAP measures to reported figures where possible. A management label such as “AI” does not by itself show that revenue is recurring, incremental, or profitable.
#1 Best Overall
If AI revenue is bundled into cloud, communications, or managed services, say that public disclosures do not isolate it. Separate subscription fees from usage charges, implementation work, resale, hardware, and pass-through costs only when the company provides enough detail. Do not treat an opportunity pipeline, bookings, or a run rate as recognized recurring revenue without evidence that supports that interpretation.
For each material recurring stream, examine the evidence that customers will continue paying and the costs required to serve them:
Rank #2
- Contract length, minimum commitments, cancellation terms, renewal rates, and actual renewal history.
- Customer concentration, churn, net expansion, and the difference between booked or run-rate amounts and recognized revenue.
- Gross margin and cash conversion after acquisition, integration, support, and retention costs.
- For AI services, whether customers pay a distinct recurring fee or receive AI within an existing contract; who pays for compute, models, licensing, and security; and whether usage-based pricing leaves the provider exposed to rising delivery costs.
A 2020 B2B telecom-sector paper discusses annual recurring revenue (ARR) and customer retention or upsell as analytical measures; it is a conceptual reference, not a source for current trading multiples (Bryan, Garnier & Co. paper). Verizon’s 2019 SEC filing illustrates customer-level inputs such as churn and revenue per user in a telecom DCF example; it should not be used as a source for current market assumptions (Verizon SEC filing).
Build a telecom-informed DCF
Forecast operating cash flow over an explicit period, with assumptions that can be checked against the company’s history, contracts, and investment plans. Telecom economics require more than a top-line growth rate: network maintenance and expansion, spectrum or licensing needs, subscriber and customer growth, churn, pricing or revenue per user, operating costs, and customer acquisition all affect cash flow.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsVerizon’s filing describes a wireless-license DCF example that incorporates subscriber growth, churn, revenue per user, capital investment, acquisition costs, operating costs, and resulting EBITDA margins. That is useful as a list of drivers, not as a current forecast or discount-rate benchmark.
Model AI services separately where disclosures allow. Forecast paid adoption, usage, renewal, service margins, delivery costs, and the software, compute, and network investment required to provide the service. Treat cost savings as a separate operating benefit: distinguish realized savings from announced plans and reflect uncertainty in the forecast. Use downside, base, and upside cases for adoption, pricing, compute costs, and competitive response. Test discount-rate, terminal-growth, and terminal-margin assumptions because small changes can materially affect long-dated cash flows.
Rank #4
Check the DCF against comparable businesses
Select public-company peers or transactions with similar geography, customer mix, growth, network ownership, leverage, regulation, and service mix. Explain why each belongs in the comparison set. Enterprise value to EBITDA is often a useful cross-check; revenue or cash-flow measures may also help when their use is justified by the businesses being compared. Interpret each measure in light of differences in capital intensity, spectrum, infrastructure ownership, and accounting definitions.
Do not lift a broad industry multiple or an old sector figure and apply it as if it were current and specific to a hybrid telecom-and-AI business. Lumen’s market-approach description uses publicly traded companies with comparable services and recognizes estimation uncertainty. The sources cited here do not establish a current, universal “AI telecom” multiple.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Decide whether AI changes the valuation evidence
AI can affect both revenue and costs. McKinsey describes possible operator opportunities in differentiated services and network APIs that expose network capabilities to developers and enterprises, potentially enabling usage-based monetization. It also discusses AI-driven network economics and automation (). That disclosure does not establish that all of the revenue is AI-derived, recurring, or sufficiently profitable to warrant software-company multiples.
BCG’s 2026 report gives sector context: it reports about 9% median annualized total shareholder return for 63 telcos over 2021–2025 and $616 billion in net value creation for the study’s telcos over five years ( The Tool Desk
