When big digital programs fail, it’s rarely for lack of software. It’s for lack of an operating model.
When marquee digital programs flame out, it’s rarely because teams chose the “wrong features.” It’s because they anchored on a product, not an operating model. Take Hertz. The rental giant hired a global SI to deliver a “transformed” web experience built on Adobe Experience Manager (AEM). Seduced, in part, by the bells-and-whistles promise of a full-blown DXP.
After spending roughly $32 million, the site never went live.
Court filings and contemporaneous reporting detail AEM components implemented poorly, code that was hard to maintain, and a platform that couldn’t be upgraded cleanly. An expensive lesson in what happens when the slideware outpaces the spine.
Across the world, Australia’s myGov rebuild followed a similar arc. The Digital Transformation Agency tapped Deloitte as prime and selected AEM among “core technology components” before running a competitive process or making a public business case.
Contracts to Deloitte snowballed to about $45 million; separate deals to Adobe topped $30 million.
When the revamped portal finally appeared, even friendly coverage called it “largely cosmetic,” with major capabilities still “coming soon.” The controversy wasn’t that AEM is “bad software”; it’s that choosing a DXP for its feature buffet - rather than designing a small, durable experience spine - locked the program into cost and complexity long before outcomes were proven.
The pattern is consistent across sectors, buying a DXP as a SKU creates theater (licenses, logos, and launch parties) while the actual outcomes (latency, lift, LTV, service completion) lag. If these programs, backed by the world’s best tools and talent, can stall under the weight of their own DXPs, the lesson isn’t “buy a different suite,” it’s to stop treating DXP as a SKU and start treating it as an operating model: a thin, pragmatic, durable experience spine for content, events, policy, and orchestration, so implementation and change stay cheap and measurable.
What the analysts are signaling
Gartner still defines DXP as an integrated suite to manage and optimize experiences, but the direction is clear: composable architectures you can deploy independently. Several 2025 summaries highlight composability, headless/hybrid delivery, and API-level integration as table stakes. One widely quoted stat from Gartner: by 2026, most orgs will be mandated to acquire composable DXP tech, i.e., modularity isn’t optional.
Forrester splits the conversation: its DXP Wave (Q4’23) benchmarks platform breadth, but parallel research on Experience Optimization (Q4’24/Q1’25) says the real game is instrumentation and real-time decisioning: collect → analyze → deliver thousands of tailored experiences continuously. Translation: if you’re not optimizing in flow, you’re not “doing DXP”; you’re doing software assembly.
IDC leans into intelligent, composable apps and digital business platforms: buyers will pay more for GenAI when it drives measurable outcomes; orchestration and shared contracts (data, events, identity) are the enablers. In other words, platform value shows up when the operating layer exists.
Real Story Group (RSG) has been blunt for years: “There is no such thing as a DXP” as a single product. The practical frame is capability maps and hybrid-composable architectures, especially around data and disciplined governance. Buyers should ignore theater and buy outcomes.
Even when analyst grids spotlight vendors, the subtext is to compose around a durable spine and prove outcomes, in real time.
What vendors are really selling
The suite-first camp, such as Adobe and Sitecore, promises one pane of glass and a roadmap that “covers everything.”
In practice, the bundle’s gravity nudges teams to adopt too much, too early. Adoption is concentrated on a few features. The rest turns into shelfware that still consumes budget and attention. Integration debt doesn’t vanish, it moves to the edges where the suite stops and your business begins.
The composable crew, such as Contentful and Contentstack, sells agility: “assemble your DXP from best-of-breed parts.” The dark side is a mosaic of tools with no shared language for content, events, identity, or governance. Procurement becomes product management. The integration layer quietly becomes the most important (and underfunded) product.
But both camps miss the point. The differentiator isn’t how many boxes you own or swap. It’s whether you provide a reliable spine that makes change cheap.