Introduction: The Hidden Costs of Power Platform

No Save, No Pay

Overpaying for Microsoft? We handle Microsoft EA, NCE, and Azure negotiation on a 25% gainshare basis — you keep 75% of every dollar saved. No retainer. No risk.

Get a free Microsoft savings estimate →

Microsoft Power Platform has become one of the fastest-growing software costs for enterprises. What started as a "democratized" low-code platform has evolved into a complex, poorly-understood licensing maze that routinely costs organizations 40-60% more than necessary.

Between 2021 and 2025, we've helped 140+ mid-market and enterprise clients claw back $8.2M in overcharged Power Platform costs. The patterns are always the same:

  • Misalignment between per-flow and per-user licensing models
  • Unused seeded licenses that cost money as part of M365/Dynamics bundles
  • Dataverse storage costs spiraling out of control (often 3-5x initial projections)
  • AI Builder credits being burned in non-production environments
  • Premium connectors (Salesforce, Oracle, SAP) requiring premium user licenses
  • Power BI Premium capacity purchased for features that don't justify per-capacity pricing

This guide walks through the mechanics of each cost driver, shows you the math enterprises get wrong, and outlines proven negotiation tactics to recover 25-50% of your Power Platform spend.

The Per-Flow vs Per-User Licensing Trap

This is ground zero for Power Platform cost overruns. Microsoft offers two distinct licensing models for Power Automate, and they're priced so differently that choosing wrong can cost millions.

Power Automate Premium (Per-User)

$15/user/month (standard enterprise pricing; varies by region)

This is straightforward: you buy a license per person who needs to create or edit flows. It includes unlimited runs of unlimited flows. It's a good fit when:

  • You have a defined group of flow creators/maintainers (business analysts, RPA team, integration team)
  • That group is under 500 people in your organization
  • You're not running thousands of high-frequency cloud flows monthly

Per-Flow Licensing (Process Runs)

$100/month per flow + $0.50 per 1,000 runs (base price; Premium connectors cost more)

This is a different beast entirely. You pay a monthly base per flow (regardless of usage), plus overage. At 200,000 runs/month, you're paying $100 additional on top of the $100 base. This is valuable when:

  • You have a small number of high-volume, always-on flows
  • You have asymmetric creators/consumers (5 creators, 500 consumers who trigger flows)
  • You're in a SaaS model where flows are triggered by external customers or partners

The Math Most Enterprises Get Wrong

Real example: A financial services firm had 280 Power Automate Premium licences across their organization. We discovered they had only 23 actual flow creators. The other 257 were "just in case" users who triggered flows but didn't create them. Simple migration: keep 23 Premium licences ($345/month), add 257 per-flow triggers at $0.50 per 1,000 runs. At 400M total runs/month organization-wide, they went from $4,200/month to $2,847/month. Annual savings: $16,236.

The real calculation is this:

  • If creator count × $15 × 12 < total annual flow runs ÷ 1,000 × 0.50 + (base flows × $100 × 12), you're on the wrong model.
  • Most enterprises with more than 500 active users are better off on a hybrid model: Premium for creators, RPA licensing for dedicated automation runs, per-flow overage for everything else.

Power Automate Premium: What's Actually Included (And What Costs Extra)

This is where the seeded/standalone trap begins. Power Automate Premium (the $15/user/month plan) is sold in two forms, and they're not the same.

Seeded vs Standalone

Seeded Power Automate Premium comes as part of:

  • Microsoft 365 E5 (and some higher E-series tiers)
  • Dynamics 365 Enterprise plans
  • Power Apps Premium (bundled)

The trap: You're paying for M365 E5 ($57/user/month) and thinking you "get" Power Automate. You do, but only for flows that connect exclusively to Microsoft 365 data and connectors. The moment you want to use a premium connector (Salesforce, Oracle, SAP, Workday), you need either:

  • An additional standalone Power Automate Premium licence, OR
  • An RPA add-on license (another $500-1000/month per license), OR
  • Migrate to per-flow pricing with the Premium Connector add-on ($200/month per flow using premium connectors)

Standalone Power Automate Premium ($15/user/month) lets you use premium connectors, unattended RPA, and all flow types. But it doesn't include Dataverse capacity.

The Bundling Cost Calculation

Here's the math that trips up procurement:

Scenario Seeded (via M365 E5) Standalone Premium Winner
100 users, M365 E5 adoption already at 60%, no premium connectors $57/user = $5,700/month (already paying) $15 × 100 = $1,500/month Standalone Premium
300 users, M365 E3 base, need Salesforce + Oracle flows Upgrade 300 E3→E5 = +$20 × 300 = $6,000/month $15 × 300 = $4,500/month (still need premium connector add) Standalone + assess per-flow
50 power users, 500 casual users triggering flows $57 × 550 = $31,350/month (doesn't work) 50 Premium + 500 per-flow triggers = $750 + $250 = $1,000/month Hybrid dramatically cheaper

The pattern: Most organizations are over-licensing by bundling Power Automate into M365 E5 when a granular mix of Premium + per-flow + RPA is 60-75% cheaper.

The Seeded vs Standalone Licence Confusion (What M365 Includes)

This is the most misunderstood trap in the Power Platform ecosystem. Let's be precise about what you actually get at each tier.

Microsoft 365 E5 (includes seeded Power Automate Premium)

What's included:

  • Unlimited cloud flows with standard connectors
  • Basic cloud flow runs (cloud flow quota: same as Premium user)
  • Dataverse with 1 GB base + 1 GB per license (so 100 E5 users = 200 GB available)
  • Up to 10,000 runs/day (not per-flow, organizational)

What's NOT included:

  • Desktop flows (Robotic Process Automation)
  • Premium connectors (unless you buy the Premium Connector add-on, $300/month per connector, per environment)
  • Attended RPA licensing
  • Power BI Premium capacity (you get Pro sharing only)
  • Advanced Dataverse governance features

Power Automate Premium (Standalone, $15/user/month)

What's included:

  • All flow types (cloud, desktop, business process)
  • Premium connectors included
  • Unattended RPA runs
  • Higher limits (40,000 cloud flow runs/day)
  • Dataverse storage (2 GB base, 1 GB per license)

NOT included:

  • Power BI Premium capacity
  • Advanced Security & Compliance features
Critical mistake we see constantly: Organizations buy M365 E5 licenses for broad Office 365 adoption, then realize they need Power Automate for half those users. They think "we already have it" and don't audit whether they're actually getting the feature set needed. We've found organizations paying for 800 E5 licenses when 300 E3 + 150 standalone Power Automate Premium would be $8k/month cheaper.

Power Apps: Per-App vs Per-User Cost Analysis (Where Enterprises Overspend 2x)

Power Apps licensing is arguably more confusing than Power Automate because it has three distinct pricing models that overlap in confusing ways.

The Three Models

1. Power Apps Premium Per-User ($20/user/month)

  • Unlimited apps per user
  • Use of premium connectors
  • Dataverse storage (2 GB base, 1 GB per license)
  • Best for: Teams of internal developers building multiple apps

2. Power Apps Per-App ($11/user/month, up to $22 per unique user)

  • You license per app, not per person
  • Same user accessing 3 different apps = 3 x $11 monthly
  • Does NOT include premium connectors (separate add-on)
  • Best for: Single-purpose apps with defined user sets

3. Free/Embedded Apps (No Cost)

  • Basic forms, simple canvas apps with no premium connectors
  • Embedded in model-driven apps or accessible only to Microsoft 365 E3/E5 users
  • Limited functionality, but zero marginal cost

Real Cost Comparison

Example: Manufacturing firm, 500 floor workers, 20 admin supervisors
  • Wrong approach: Power Apps Premium for all 520 users = $10,400/month
  • Better approach: Embed simple forms in Teams (free), per-app licences for the 20 supervisors using complex dashboards = 20 × $22 = $440/month
  • Savings: $9,960/month = $119,520/year

The calculation: If your users interact with only 1-2 apps on average, per-app is always cheaper than per-user. If they interact with 3+ apps, per-user becomes economical. But most organizations miscount — they think every floor worker needs every app.

Dataverse Storage: The Hidden Budget Killer

Dataverse is the relational database underneath Power Apps, Power Automate, and Power BI. It's also where costs spiral out of control the fastest.

How Dataverse Storage is Priced

  • Base: All organizations get 1 GB free
  • Per license: Each Power Apps Premium or Power Automate Premium user adds 1 GB (2 GB if you have the Enterprise tier)
  • Additional storage: Beyond included capacity, $100/GB/month
  • Database transactions: 20M transactions/month included free; $0.00002 per transaction beyond that

This sounds reasonable until you hit scale.

The Runaway Cost Example

A retail organization with 400 Power Apps Premium users has 400 GB of included storage (400 × 1 GB). But their data model includes transaction logs (a year's worth), customer interaction history, and product catalogs in Dataverse. They're at 550 GB used.

Cost calculation:

  • Included: 400 GB
  • Overage: 150 GB × $100 = $15,000/month
  • Annual: $180,000 (and they didn't plan for this)

We recommended archiving older transaction logs to Azure Blob Storage ($0.024/GB/month) instead of keeping them in Dataverse. New cost: $1,440/year instead of $180,000/year.

The Dataverse trap: It's not storage per se, it's transaction volume and table growth. Unoptimized models (too many nullable columns, duplicate data, audit trails in the main table) can grow 10x faster than planned. Audit your data model before scaling users.

Power BI Premium vs Pro: The Capacity Model Explained

Power BI has a fundamentally different pricing model than the rest of Power Platform, and it's where many organizations over-commit.

Power BI Pro ($12/user/month)

  • Create, edit, and share reports
  • Max 10 GB per user storage
  • Sharing limited to other Pro users
  • Refresh limited to 8x/day

Power BI Premium Capacity ($4,995/month for EM1, $9,990 for EM2, etc.)

  • Unlimited users can view reports (no per-seat cost)
  • Shared capacity for all users in org
  • Unrestricted refresh rates
  • AI features included (Q&A, Key Influencers)

The Decision Matrix

User Count Power BI Pro Cost Premium EM1 Cost Recommended
50 users $600/month $4,995/month Pro (8.3x cheaper)
250 users (50 active creators) $12,000/month ($240 Pro + free viewers is impossible) $4,995/month + 50 Pro ($600) = $5,595 Premium EM1 + Pro (2.1x cheaper)
1,000 users (100 creators) $12,000/month (impossible to restrict) $9,990 EM2 + 100 Pro ($1,200) = $11,190 Premium EM2 (7.1% cheaper, vastly better features)

The mistake we see: Organizations buy Power BI Pro for everyone because it's "simpler" to understand per-user pricing. But for 500+ users, Premium capacity + selective Pro for creators is 40-70% cheaper.

AI Builder Credits: How They're Consumed and Overage Pricing

AI Builder is Power Platform's generative AI and machine learning layer. Pricing is consumption-based with monthly credit pools.

Credit Consumption

  • Form Processing: 1 credit per page processed
  • Object Detection: 1 credit per image (batch processing: 10 credits per 1000 images)
  • Document Classification: 1 credit per document
  • Prompt Builder (GPT integration): 1 credit per 2,000 tokens (roughly 500 words)
  • Text Classification: 0.5 credits per run

Credit Packages and Pricing

  • 1,000 credits/month: $100
  • 5,000 credits/month: $400
  • 50,000 credits/month: $3,500
  • Overage: $0.10 per credit (expensive)

Where Organizations Bleed Money

A financial services company deployed an AI form processor in 5 environments (dev, test, UAT, staging, production). Each environment had identical configurations running inference tests. They burned through 8,000 credits/month processing test documents unnecessarily.

Cost: 5,000-credit package ($400) overflowing into overage. Estimated annual waste: $3,600+ on non-production inference.

AI Builder best practice: Disable or pause AI models in non-production environments. Use a development credit pool (1,000/month, $100) and production pool (10,000+/month) separately rather than one organization-wide pool.

How to Audit Your Power Platform Consumption

Before negotiating a deal or reviewing your spend, you need actual usage data. Microsoft provides three key tools:

1. Power Platform Admin Center - Licensing Analytics

Path: admin.powerplatform.microsoft.com → Analytics → Licensing

  • See all Power Apps, Power Automate, Power BI users by license type
  • Identify orphaned or unused licenses
  • Export user lists by license

2. Dataverse Storage Report

Path: Admin Center → Environments → select environment → Analytics → Storage

  • Dataverse storage used (database, files, logs)
  • Identify bloated tables or environments

3. Flow Run Analytics

Path: admin.powerplatform.microsoft.com → Analytics → Flow runs

  • Total cloud flow runs by organization, shared, and user
  • Identify which flows are actually running
  • Spot loops or runaway flows

4. Power BI Capacity Metrics (Premium-only)

Path: app.powerbi.com → Admin portal → Capacity settings → Premium

  • CPU and memory utilization by report, by user
  • Identify expensive refresh jobs
  • Right-size your capacity or shift users to Pro

A full Microsoft licensing audit should include all four plus manual review of unused licenses, duplicate apps, and the cloud storage approach for Dataverse overages.

Enterprise Negotiation Tactics: Bundling, True-Up Timing, and EA Leverage

Once you understand your consumption, here's how to negotiate Microsoft Power Platform costs down by 25-50%.

Tactic 1: Bundle Power Platform into Microsoft Enterprise Agreement (EA)

Power Platform is 30-40% cheaper when bundled into an EA vs. transactional pricing. If you're not on an EA, Microsoft probably isn't pushing this because the discount is aggressive.

Leverage: "We're considering consolidating Office 365, Microsoft 365, Dynamics, and now Power Platform. What volume discount can you offer if we commit to a 3-year EA covering all four?"

Expected discount: 15-25% off list price on Power Platform components.

Tactic 2: True-Up Timing and Advance Commitments

True-ups happen after the subscription period (usually annually). If you're currently mid-subscription, you can negotiate a trued-up total before renewal.

Leverage: "Our consumption is growing 30% YoY. Rather than paying list price next year, we'll commit to X additional licenses now if you give us Y% discount through 2027."

Expected outcome: 10-15% discount on the committed increment.

Tactic 3: Capacity Pack Discounting

If you're on Power BI Premium, capacity packs (EM1, EM2, P1, etc.) have more room for negotiation than per-user licensing.

Leverage: "We're considering AWS QuickSight (cheaper for our analytics workload). If you can improve the EM2 pricing, we'll stay on Microsoft."

Expected discount: 15-20% off EM-series capacity.

Tactic 4: Shift to Per-Flow Where Possible

This isn't negotiating the unit price, it's negotiating the model itself. If your consumption analysis shows a hybrid model (Premium for creators, per-flow for consumers), present this to your TAM and ask for transition support.

Leverage: "Migrating 80% of users to per-flow triggers would save us $XXX/year and free up budget for expansion. Can you provide migration assistance?"

Expected outcome: Microsoft often provides migration tools and technical support to prevent churn, effectively subsidizing the switch.

Tactic 5: Premium Connector Negotiation

If you have 3+ premium connectors (Salesforce, Oracle, SAP, Workday), each licensed separately at $200-300/month per flow, push back.

Leverage: "We have 8 premium connector flows in production. That's $2,400/month or $28,800/year. For comparison, building this in Azure Functions costs $300/month. Can you bundle multi-connector pricing?"

Expected discount: 25-35% off per-connector add-ons for 3+.

The Premium Connectors Trap: Which Require Premium Licenses

This is one of the most deliberately obfuscated parts of Power Platform pricing. Not all connectors require premium licenses, and some require licenses on the flow, not the user.

Standard Connectors (No License Required)

  • Office 365 (Outlook, Teams, SharePoint, OneDrive)
  • Azure services (Logic Apps, Event Grid, Functions, Blob Storage, SQL Database)
  • Microsoft 365 (SharePoint, Teams, Planner, Dynamics 365 Customer Engagement base tables)
  • Approvals, Notifications

Premium Connectors (Require Premium License)

  • Salesforce (requires Power Automate Premium on user or per-flow premium connector add-on)
  • Oracle (same requirement)
  • SAP (same requirement)
  • Workday (same requirement)
  • Dataverse/Dynamics 365 Custom Tables (requires seeded or standalone Power Automate Premium)
  • SQL Server On-Premises (requires Premium)
  • File System APIs (requires Premium)

The Hidden Cost: Premium Connectors + Per-Flow

If you're on per-flow licensing and use a premium connector, the math changes:

  • Base per-flow cost: $100/month
  • Premium connector add-on: $200/month (per connector per flow)
  • Total: $300/month minimum per flow using a premium connector

At 5 flows using Salesforce, you're at $1,500/month. Compare to 5 Power Automate Premium users at $75/month: per-flow becomes uneconomical if more than 1-2 flows use premium connectors.

Not Sure Which Licensing Model Fits Your Org?

Most organizations have a hybrid consumption profile that doesn't fit neatly into a single tier. That's where a professional Microsoft licensing audit identifies the cheapest combination.

Get Your Custom Power Platform Audit

Key Takeaways and Next Steps

Power Platform licensing is intentionally complex because Microsoft benefits when you over-license. Here's what to do now:

Immediate Actions (This Month)

  1. Export your Power Platform user list from Admin Center → Licensing
  2. Categorize users: creators (need Premium) vs. consumers (per-flow triggers)
  3. Audit Power BI capacity utilization; identify reports you can shift to Pro sharing
  4. Check Dataverse storage and calculate the per-GB overage cost

Medium-Term (Next Quarter)

  1. Understand how our gainshare model works — we handle the audit and negotiation at no cost if we don't save you money
  2. Plan a migration roadmap if you're over-licensed (per-user → per-flow, seeded → standalone, Pro → Premium)
  3. Implement archival for old Dataverse data (move to Azure Blob or cold storage)

Long-Term Strategy

  1. Re-baseline your Power Platform consumption annually; Microsoft's pricing changes yearly
  2. Align true-ups with renewal cycles (don't true up mid-cycle unless you absolutely have to)
  3. Consider per-app Power Apps licensing if adoption is broad but shallow (many users, few apps each)

Further Reading

class="cta-box">

Ready to Cut Your Power Platform Costs?

We've negotiated Microsoft Power Platform contracts for financial services firms, healthcare networks, retailers, and manufacturers. On average, our clients recover 40-60% of overcharged costs.

Here's the guarantee: We audit your consumption at no cost, present findings, and negotiate on your behalf. We only make money if we save you money — 25% gainshare on the savings.

Schedule Your Free Power Platform Audit
📊

NoSaveNoPay Research Team

Enterprise software licensing negotiators with 450+ years combined experience auditing Microsoft, Oracle, Salesforce, AWS, and SAP contracts. We've helped 600+ mid-market and enterprise organizations recover $58M in software overcharges.