Here’s the rewritten content with all HTML tags preserved as is:
Analysts believe Salesforce’s choice to stop offering new enterprise contracts for Heroku likely indicates a strategic shift away from the former leading PaaS, as the company now focuses on AI-driven expansion.
Salesforce has indicated a significant strategic change for its veteran cloud platform, Heroku, by ceasing new Heroku Enterprise contract sales and placing the service into a “sustaining engineering” phase, primarily focused on maintenance.
“As of today, Heroku is moving to a sustaining engineering model, prioritizing stability, security, reliability, and support… We will no longer be providing Enterprise Account contracts to new clients,” stated Nitin T Bhat, Heroku’s chief product officer, in a recent blog post.
Industry experts interpret Heroku’s altered status as a precursor to the eventual retirement of the formerly prominent Platform-as-a-Service (PaaS), as Salesforce redirects its efforts towards growth powered by AI.
Sanchit Vir Gogia, chief analyst at Greyhound Research, explains that sustaining engineering is seldom a permanent state; instead, it’s a temporary measure designed to make a future integration or discontinuation less problematic for the parent firm, Salesforce, in this instance.
Gogia further elaborated that once a product or platform enters the sustaining engineering stage, engineering efforts pivot from creating new features to mitigating risks. Its momentum diminishes, and resources across product development, sales, partnerships, and talent often redeploy. In rapidly changing cloud environments, such a platform gradually becomes less pertinent for developer tasks, and recovering from a loss of internal support is rare.
Historically, similar patterns have often signaled managed decline scenarios throughout the industry, observed Pareekh Jain, principal analyst at Pareekh Consulting.
“Numerous established examples exist where vendors transitioned products into sustained engineering. IBM moved Bluemix into a maintenance state as it firmly shifted focus to Red Hat OpenShift, and VMware put Pivotal Cloud Foundry into sustain mode before integrating it entirely into Tanzu. Although Google App Engine’s standard environment is still technically operational, its development slowed significantly after Google realigned its strategy around Kubernetes and GKE,” Jain elaborated.
Jain cited additional instances, mentioning Microsoft Silverlight, which remained in a “supported but static” condition for years prior to its eventual discontinuation, as well as Oracle’s Solaris, Atlassian’s Fisheye, and Adobe’s Flash—all of which were either abandoned or saw their strategic importance diminish.
What led to Heroku’s diminished importance?
The primary factor contributing to Heroku’s reprioritization seems to be its difficulty in adapting to the evolving demands of cloud platforms, including the competitive market and economic aspects.
Chandrika Dutt, research director at Avasant, identifies two key structural hurdles.
Firstly, the current competitive environment features alternatives such as Render, Railway, Fly.io, Vercel, and Supabase, which offer greater agility, modularity, and cost efficiency for contemporary development approaches.
Secondly, the foundational Postgres ecosystem has expanded considerably, with dedicated hosted Postgres and backend services lessening the distinct value proposition of Heroku’s unified stack.
Dutt highlighted, “The convergence of reduced innovation, increased relative expenses, and the escalating opportunity cost of engineering efforts probably influenced Salesforce’s choice to redirect investments towards areas with greater growth potential, specifically AI-focused services and wider cloud integrations.”
Conversely, when Salesforce acquired Heroku in 2011, it offered benefits like straightforward application deployment and hosted Postgres, which Salesforce leveraged to attract developer-driven cloud projects and expand its ecosystem, Dutt further explained.
Gogia from Greyhound concurred with Dutt, stating that Heroku initially acted as Salesforce’s gateway to the wider developer community. However, Salesforce eventually outgrew the initial challenge Heroku was designed to address, strategically moving its attention from drawing developers to managing and profiting from enterprise AI solutions.
Gogia observed that as Salesforce leaned into substantial enterprise agreements, platform unification, and AI-driven distinctiveness, Heroku’s role grew ambiguous—it was too autonomous for deep integration within Salesforce, yet too branded to function as an impartial developer platform.
Gogia remarked that attempts to “enterprise-grade” Heroku enhanced its compliance and networking features but diminished its initial appeal to developers. Subsequent modernization efforts failed to halt its ecosystem’s decline. He concluded that by the time Salesforce began emphasizing AI and data platforms in its strategy, Heroku had lost its central importance.
The path ahead for Heroku
Nonetheless, Salesforce clarified that the sustaining engineering phase does not signify an immediate discontinuation.
“Customers currently utilizing Heroku will experience no alterations. Both existing and new customers who process payments through credit cards via the Heroku dashboard can proceed with using Heroku without any changes to pricing, invoicing, service, or daily operations,” Bhat stated in his blog post, providing reassurance of ongoing support and even prompting contract renewals.
Despite these assurances, analysts advise enterprise clients to carefully examine this transition when developing their long-term strategies.
Dutt highlighted that a lack of strategic investment substantially elevates long-term platform risk. She further advised customers against initiating new strategic development on Heroku, as the indicators align with a platform approaching obsolescence rather than undergoing a revitalization.
Gogia also cautioned that CIOs ought to begin viewing Heroku as a legacy component of their infrastructure.
“Avoid assuming that runtime compatibility with the wider ecosystem will continue indefinitely. Catalog your dependencies, particularly data services and integration points. Make sure backups and exports are standard practice, not merely a goal,” Gogia advised.
Gogia further explained, “Develop a prototype for at least one feasible alternative deployment route to grasp the migration effort accurately, rather than speculating. The most significant danger isn’t a sudden cessation of service. Rather, it’s complacency, where teams eventually realize that the expense, labor, and organizational resistance involved in transitioning have become far more substantial than necessary.”
Likewise, Dutt highlighted that businesses beginning migration planning early will possess an advantage, as they will retain operational control rather than being dependent on a vendor or Salesforce until Heroku’s eventual discontinuation is declared. From Salesforce’s viewpoint, Jain suggested that Heroku’s Postgres, appearing to be its most enduring component, will probably be integrated into its Data Cloud as a managed service, while the dyno-based compute layer will be gradually retired as Salesforce guides AI and agentic development via a combination of Agentforce and the Data Cloud.
