top of page
Lost by trying to out-build AWS and Azure at their own infrastructure game instead of leaning into the network and edge assets CenturyLink actually owned — a cautionary tale in attacking a market on the incumbent's terrain rather than your own structural strength.
1
MODEL
BUSINESS MODEL
Infrastructure Platform
model bm
HOW THEY BUILT IT
- Built primarily through acquisition: CenturyLink acquired Savvis (2011, ~$2.5B) for enterprise hosting and colocation, then Tier 3 (2013) for public cloud/PaaS capability, assembling 'CenturyLink Cloud' as a bolted-together IaaS/PaaS offering to compete with AWS and Azure.
- Positioned as an enterprise-friendly alternative to hyperscale public cloud, leaning on CenturyLink's existing telecom enterprise relationships rather than developer-first adoption.
- Ultimately could not match the capital intensity and pace of innovation of AWS, Azure, and Google Cloud, and CenturyLink began winding down and divesting public-cloud assets in the late 2010s, eventually rebranding the parent company as Lumen Technologies (2020) to refocus on network and edge infrastructure rather than compete head-on in commodity public cloud.
HOW TO ARCHITECT IT
How to architect this model (or avoid this mistake):
1. Before entering a capital-intensive infrastructure race against hyperscalers, honestly assess whether your actual structural advantage (in CenturyLink's case, a nationwide fiber network) is being leveraged, or whether you're just building a smaller, later version of what incumbents already do at scale.
2. If you acquire your way into a category (as CenturyLink did with Savvis and Tier 3), invest as much in integrating the acquired technology into a coherent product as in the deals themselves — a stitched-together cloud offering competing against AWS's unified platform is a structural disadvantage from day one.
3. Recognize the exit signal early: when a legacy telecom's cloud unit can't keep pace with hyperscaler R&D spend, pivoting to a narrower, defensible niche (network/edge, in Lumen's case) beats continuing to burn capital chasing an unwinnable scale war.
DISTRIBUTION MODEL
Enterprise Sales, Direct Sales
dm
HOW THEY OPERATIONALIZED
- Sold primarily through CenturyLink's existing enterprise telecom account relationships, cross-selling cloud infrastructure to customers who already bought network connectivity — a channel advantage in reach but not in developer mindshare.
- Lacked the developer-first, self-serve, API-driven distribution motion that made AWS the default choice for engineering teams, a structural distribution disadvantage against hyperscale competitors.
HOW TO REPLICATE WHAT WORKED
What worked (narrowly): using existing enterprise telecom relationships to get a foot in the door for early cloud infrastructure deals before the market fully commoditized around AWS/Azure/GCP.
Trap if copied blindly: entering a hyperscale infrastructure race without hyperscale capital or developer-first distribution is close to un-winnable — this case is most useful as a cautionary example of misjudging your actual structural advantage rather than a playbook to replicate.
2
MARKET
mkt mt es
MARKET TYPE
Consolidated Market
WHY THEY WON
By the time CenturyLink Cloud was assembled, public cloud infrastructure had already begun consolidating rapidly around AWS, with Azure and Google Cloud as the only credible secondary challengers, each backed by capital and R&D spend legacy telecoms couldn't match. CenturyLink entered a market already too consolidated for a smaller, acquisition-stitched player to compete on equal terms. Transferable principle: entering a capital-intensive, already-consolidating market as a smaller player without a genuinely differentiated structural advantage is a losing sequencing, however strong your existing customer relationships.
ENTRY STRATEGY
Greenfield Entry
EXECUTION
CenturyLink assembled its cloud offering entirely through acquisition (Savvis in 2011, Tier 3 in 2013) rather than organic build, the fastest route to cloud capability for a telecom with no existing hyperscale infrastructure — but one that left the resulting platform less architecturally unified than AWS's ground-up build.
FOOTHOLD STRATEGY
fs
Beachhead Strategy
The intended beachhead was existing CenturyLink enterprise telecom customers who could be cross-sold cloud infrastructure alongside their network contracts — a reasonable initial wedge given the existing relationship, but one that never expanded meaningfully beyond that captive base into the broader developer-driven cloud market that ultimately determined category leadership.
GROWTH CAMPAIGN
CAMPAIGNS THAT WORKED
Savvis acquisition (2011): the foundational asset for enterprise hosting and colocation capability.
Tier 3 acquisition (2013): added public cloud/PaaS capability, forming the combined 'CenturyLink Cloud' brand.
Divestiture and Lumen rebrand (2020): the eventual retreat from public cloud competition, refocusing the parent company on network and edge infrastructure where it retained genuine structural advantage.
KEY LEARNING
If you're assessing whether to enter a capital-intensive, consolidating infrastructure category, ask honestly whether your existing assets (network, relationships, data) give you a genuine structural edge in that specific category, or whether you're simply late to a race the leaders are already winning — if the latter, a narrower pivot to your actual strength may be the better long-term bet, as Lumen's later refocus on network/edge illustrates.
gc
3
MONEY
money rev pri
REVENUE MODEL
Usage-Based, Subscription
PRICING MODEL
Competitive Pricing
WHY THEY WON
Priced similarly to competing IaaS/PaaS offerings — usage-based compute, storage, and network consumption plus subscription tiers for managed services — bundled where possible with CenturyLink's existing enterprise network contracts.
Priced to compete directly with AWS and Azure's published rates rather than establish a premium or differentiated pricing position, targeting enterprise IT buyers already engaged with CenturyLink for network services who wanted to consolidate vendors rather than shop for the cheapest possible compute.
TARGET AUDIENCE
CUSTOMER BUYING BEHAVIOUR
tg cb
Enterprise IT departments already buying CenturyLink network/telecom services (buying vendor consolidation convenience); mid-market companies wary of AWS/Azure complexity (buying a simpler, more relationship-driven cloud vendor).
Sales-led and relationship-driven, leveraging existing account manager relationships from CenturyLink's telecom business rather than the self-serve, credit-card sign-up model that drove AWS and Azure's developer-first growth.
4
MOTION
GROWTH EXPANSION MODEL
COMPETITIVE STRATEGY
motion ge cs
Vertical Integration
Fast Follower
HOW THEY EXPAND
CenturyLink attempted vertical integration by combining network connectivity, colocation (Savvis), and public cloud compute (Tier 3) into a single enterprise offering, but ultimately reversed course — divesting cloud assets and rebranding as Lumen Technologies in 2020 to refocus specifically on network and edge infrastructure rather than continue competing across the full cloud stack.
HOW THEY COMPETE
CenturyLink Cloud positioned itself as a fast-follower alternative to AWS and Azure for enterprise customers wary of hyperscaler complexity, a defensive strategy that ultimately proved unsustainable once it became clear the follower lacked the capital intensity to keep pace with continuous hyperscaler innovation.
GROWTH ENGINE
GTM
ge n gtm
Paid Acquisition Engine
Growth relied on cross-selling into an existing customer base rather than a self-reinforcing viral or network-effect loop, meaning growth was capped by the size of CenturyLink's existing enterprise telecom relationships rather than expanding independently — a structural ceiling that became apparent once hyperscaler competitors captured essentially all new cloud-native demand.
Cross-sell motion through existing enterprise telecom account managers, targeting CenturyLink's installed base of network customers rather than building a new developer-first acquisition channel.
SUSTAINING MOATS
Switching Costs, High Customer Lock-In, Brand Power, Technology Advantage (complex enterprise scenarios)
moat
CenturyLink Cloud's only real moat was distribution through existing telecom relationships, and that advantage proved insufficient against AWS and Azure's combined capital scale and developer-first network effects — a clear example of a moat that looks real on paper (existing customer relationships) but doesn't hold up once the market's underlying mechanism (developer self-serve adoption, hyperscale R&D) shifts beneath it.
bottom of page