Heroku Enterprise carries pricing that buyers consistently under-budget. The dyno economics, add-on costs, and Private Spaces premium combine into a TCO profile materially heavier than the published rate card suggests.
Heroku occupies a specific position in the Salesforce portfolio. It is the platform-as-a-service offering inherited from Salesforce's 2010 acquisition, sold both standalone and as part of the broader Salesforce platform contract. Heroku Enterprise sits at the upper tier of the offering, with private network isolation (Heroku Private Spaces), enterprise SSO, dedicated support, and the commercial terms that enterprise buyers expect.
This article examines the Heroku Enterprise cost structure, the components that drive most of the spend, the negotiation levers that move pricing in the buyer's favor, and the strategic question of whether Heroku is the right home for the workloads it runs.
Heroku Enterprise pricing breaks into four primary components. Each has distinct economics and distinct negotiation dynamics.
The first is dyno spend — the compute capacity that runs the customer's applications. Heroku dynos are priced per dyno per hour, with several performance tiers. Performance-M dynos run at approximately $0.347 per hour ($250/month); Performance-L dynos run at approximately $0.694 per hour ($500/month). Larger Performance-XL and Performance-2XL tiers are available on Private Spaces only. An enterprise workload running 20 Performance-M dynos plus 6 Performance-L dynos consumes approximately $8,000 in dyno spend per month, or $96,000 annually.
The second is the Heroku Private Spaces premium. Private Spaces — the network-isolated runtime environment required for most enterprise compliance postures — adds a flat fee of approximately $1,500 per month per Private Space, plus the dynos running inside it. Most enterprise deployments use two to four Private Spaces to separate production, staging, and other environments, adding $36,000 to $72,000 annually before any dyno spend.
The third is the add-on economics. Heroku's add-on marketplace provides databases, caches, search, messaging, observability, and dozens of other capabilities, each priced separately. Enterprise add-on spend frequently exceeds dyno spend, particularly on Postgres database tiers, Redis caches, and observability add-ons. A typical enterprise stack spending $96K on dynos often spends $140K to $260K annually on add-ons.
The fourth is the platform fee. Heroku Enterprise carries a platform fee that varies by contract size and includes the enterprise SSO, audit logging, dedicated support, and account management. The platform fee typically runs $30K to $80K annually depending on the enterprise tier.
| Component | Typical enterprise spend | % of total |
|---|---|---|
| Dyno spend | $96K–$320K/year | 25–35% |
| Private Spaces premium | $36K–$96K/year | 10–15% |
| Add-on spend (Postgres, Redis, etc.) | $140K–$520K/year | 40–55% |
| Platform fee | $30K–$80K/year | 8–12% |
| Total enterprise TCO | $300K–$1M/year | 100% |
Heroku Postgres is the largest single add-on line item on most enterprise Heroku contracts. The pricing tiers escalate sharply as capacity requirements grow. Standard-0 ($50/month) and Standard-2 ($200/month) tiers cover development workloads; production workloads typically require Premium or Private tiers. Private-4 lists at approximately $3,500/month; Private-8 at approximately $7,000/month; Private-10 at approximately $10,000/month.
The Postgres tier required by an enterprise workload is determined by data volume, query concurrency, and high-availability requirements. Buyers who size Postgres tiers based on initial workload often face capacity-driven upgrades within the first 18 months as data accumulates. The procurement practice is to project 24-month data growth during initial sizing and negotiate price-lock on the projected tier rather than the initial tier.
Heroku Postgres list pricing scales steeply across tiers. A workload growing from Premium-4 to Premium-10 across 24 months sees database cost grow from $42K to $120K annually — a 186% increase that buyers frequently fail to anticipate in initial budgeting.
Three negotiation levers move Heroku pricing in the buyer's favor.
The first is commitment-based dyno discounting. Heroku offers dyno discounting for committed annual spend, with discount levels typically running 10–18% off list rates for $250K+ annual dyno commitments. The procurement practice is to forecast dyno consumption across the contract term and negotiate the committed-spend tier accordingly.
The second is add-on bundle negotiation. The Heroku add-on marketplace presents itself as a la carte, but enterprise customers can negotiate bundle discounts on the largest add-ons — Postgres, Redis, observability — when combined with dyno commitment. Bundle discounts typically run 8–15% off list rates and require the negotiation to be conducted with Heroku's enterprise sales team rather than through the standard add-on purchase flow.
The third is multi-Private-Space pricing. Buyers deploying three or more Private Spaces should negotiate a multi-space discount that reduces the per-space platform fee. Discount typically runs 15–30% on the second through fifth Private Spaces.
Heroku Enterprise's competitive position has weakened materially through 2022–2026 as buyers increasingly evaluate alternatives. The competitive alternatives that move pricing pressure in the buyer's favor include:
AWS Elastic Beanstalk and AWS App Runner for buyers willing to operate their own runtime configuration. The cost reduction relative to Heroku is typically 40–65% for comparable workloads. The operational overhead is higher, but for buyers with mature DevOps capability, the savings are substantial.
Google Cloud Run for container-based workloads. Cloud Run's per-request pricing model is fundamentally different from Heroku's per-dyno-hour model and produces lower cost for bursty traffic profiles. The migration effort from Heroku to Cloud Run is meaningful but not prohibitive.
Render and Railway, which target the same developer-experience positioning as Heroku at lower price points. These platforms have matured significantly through 2024–2026 and are credible enterprise alternatives for new workloads.
Vercel and Netlify for frontend-heavy workloads. The competitive overlap with Heroku is partial — Vercel and Netlify focus on JAMstack and serverless patterns — but the migration economics are favorable for workloads that fit the model.
500+ engagements · $420M+ in client savings · 34% average reduction.
Contact Us →Buyers considering migration off Heroku face meaningful migration cost. The cost varies by application complexity, add-on dependency, and the chosen destination platform. Typical migration costs run:
| Destination | Migration cost | Annual savings | Payback period |
|---|---|---|---|
| AWS Elastic Beanstalk | $80K–$280K | $120K–$420K | 8–14 months |
| Google Cloud Run | $120K–$340K | $140K–$380K | 10–16 months |
| Render / Railway | $40K–$160K | $80K–$240K | 6–12 months |
| Vercel (frontend workloads only) | $60K–$220K | $60K–$180K | 12–18 months |
The migration calculus depends on workload fit, internal operational capability, and the strategic posture toward platform consolidation. For buyers running Heroku alongside a broader cloud-native estate, the consolidation argument frequently dominates the per-workload economics. For buyers running Heroku in isolation, the per-workload migration economics may not justify the effort.
Heroku purchased standalone has weaker commercial terms than Heroku bundled into a broader Salesforce contract. The bundle dynamics produce 8–15 percentage points of additional discount on Heroku when negotiated concurrently with platform license or other Salesforce purchases. Buyers running both Heroku and Sales Cloud should always negotiate the two concurrently rather than as separate transactions.
The bundle conversation also opens commercial terms not available on standalone Heroku contracts. These include co-termed renewal cycles, unified credit pools that can be redirected across products, and named Salesforce architects who can support the Heroku-Salesforce integration patterns that often drive the workload requirement in the first place.
The largest negotiation lever on Heroku is rarely the rate; it is the strategic question of whether Heroku is the right home for the workloads it runs. Many Heroku workloads were placed on the platform 5–10 years ago when the developer-experience advantage was material; in 2026, that advantage is narrower and the competitive alternatives are stronger. Buyers running disciplined platform-portfolio reviews frequently identify that 30–60% of their Heroku workloads have better homes on other platforms.
The procurement practice is to conduct a workload-by-workload review every 24 months, identify the workloads with credible migration alternatives, and use that analysis to either migrate the workloads or to drive Heroku rate concessions on the workloads that stay. The combined effect of selective migration and rate negotiation routinely produces 25–50% TCO reduction across the Heroku estate.
Heroku Enterprise is a capable platform with a premium price point. Buyers who treat it as the default platform without rigorous TCO comparison routinely overpay; buyers who run the competitive analysis and negotiate with credible alternatives consistently produce materially better commercial outcomes. The procurement effort required is meaningful — workload review, migration cost modeling, alternative platform evaluation — and the cost savings on a $400K+ annual Heroku spend repay that effort many times over. The discipline is available to any procurement team willing to invest the analytical work; the savings, on engagement after engagement, are substantial.
Dyno consumption rarely runs at constant utilization across the workload day. Most enterprise applications exhibit traffic patterns that vary 4–12x between peak and off-peak windows. Heroku's autoscaling capability adjusts dyno count dynamically in response to traffic, with the consumption billing reflecting actual dyno-hour usage rather than maximum provisioned capacity.
The autoscaling economics produce two distinct cost patterns. Workloads with sustained high utilization — typically those running at 60%+ of peak load throughout the business day — produce predictable consumption that maps cleanly to the dyno-hour billing. Workloads with bursty patterns — long idle periods punctuated by brief traffic spikes — produce consumption patterns where the average is materially lower than the peak. For bursty workloads, autoscaling is the architectural feature that prevents dramatic over-provisioning.
Buyers who configure autoscaling thresholds aggressively, with scale-down policies that release dynos within minutes of demand reduction, routinely achieve dyno consumption 35–60% below their initial sizing estimates. Buyers who configure autoscaling conservatively, holding dynos in reserve against possible traffic spikes, achieve smaller reductions but maintain larger operational headroom. The trade-off should be made explicitly against application reliability requirements.
Beyond core platform spend, the operational tooling around Heroku deployments adds meaningful cost. Heroku's native observability — log streams, basic metrics, app status — is sufficient for early-stage workloads but inadequate for enterprise operations. Most enterprise Heroku deployments add commercial observability products: Datadog or New Relic for application performance monitoring, Logz.io or Sumo Logic for log aggregation, Sentry or Rollbar for error tracking.
| Observability layer | Annual cost (enterprise) |
|---|---|
| APM (Datadog, New Relic) | $30K–$140K |
| Log aggregation | $20K–$80K |
| Error tracking | $10K–$40K |
| Synthetic monitoring | $8K–$30K |
| Total observability layer | $68K–$290K |
The observability spend is rarely captured in the Heroku TCO comparison against alternatives, but it is genuinely platform-related. Cloud-native alternatives frequently include comparable observability at lower cost — particularly AWS CloudWatch and Google Cloud Operations Suite — which strengthens the migration economics for buyers willing to operate the native cloud observability stack.
For Heroku deployments that integrate heavily with Salesforce data, Heroku Connect is the standard integration mechanism. Heroku Connect maintains a bidirectional sync between Heroku Postgres and Salesforce objects, with the synchronization handled transparently by the platform. The capability has substantial cost: Heroku Connect lists at approximately $1,500/month per connection for the standard tier, with higher tiers for high-volume or low-latency synchronization.
The Heroku Connect economics frequently dominate the Heroku-Salesforce integration TCO. A typical enterprise deployment with 2–4 Connect mappings spends $36K to $144K annually on Connect licensing alone, before counting the Postgres capacity and dyno spend that the Connect-driven workload requires. Buyers should evaluate whether the operational simplicity of Heroku Connect justifies its cost relative to building the integration directly against the Salesforce API. For workloads with simple synchronization requirements, direct API integration is materially cheaper.
Heroku spend should be reviewed quarterly with dyno utilization, add-on consumption, and Private Spaces usage all evaluated against budget. The quarterly cadence catches consumption drift before it compounds into material overspend. Buyers who run annual rather than quarterly reviews routinely discover by year-end that consumption has drifted 20–40% above budget through gradual accumulation rather than discrete events. The quarterly discipline is modest in effort and consistently produces better cost outcomes than less frequent review cycles.
Heroku Enterprise remains the right platform for specific workloads — particularly those with deep Salesforce data dependencies, modest scale, and a development team that values the platform's productivity. For broader enterprise workloads, the competitive alternatives are increasingly compelling, the migration economics frequently favorable, and the disciplined procurement evaluation routinely produces meaningful cost reduction. Buyers who treat Heroku as a strategic line item rather than as a default platform consistently produce better outcomes across the multi-year horizon. The procurement discipline required is modest; the savings, on engagement after engagement, are substantial.
One field-tested negotiation tactic per month. No vendor pitches.