What Dynamics 365 is in 2026
Dynamics 365 in 2026 is Microsoft’s modular CRM-and-ERP suite running on Dataverse, woven into the Power Platform, and increasingly fronted by Copilot agents created in Copilot Studio. It is not one product — it is a family of business apps (Sales, Customer Service, Field Service, Marketing/Customer Insights, Finance, Supply Chain, Project Operations, Commerce, and Business Central) that share a common data layer, a common security model, and a common AI runtime. The marketing term Microsoft prefers is “agentic ERP and CRM,” and after the 2026 Wave 1 release the term is more than marketing — autonomous agents now ship inside Customer Service, Sales, Contact Center, and Business Central as first-class objects.
Compared to Salesforce, Dynamics is the Microsoft-stack play. If your organization runs Microsoft 365, Azure, Entra ID, Teams, Fabric, and Power BI, Dynamics fits without a foreign-DNA tax. Compared to ServiceNow, Dynamics owns the customer-facing and back-office app workloads where ServiceNow owns the operational and IT workflows. Compared to NetSuite or SAP S/4HANA, Dynamics 365 Finance and Supply Chain compete in the upper mid-market and lower enterprise; Business Central competes for the SMB ERP slot.
The 2026 positioning is “agentic by default on Dataverse, governed by Copilot Studio.” Every Wave release this year ships either a new agent, a new AI Builder primitive, a Copilot Studio MCP integration, or a Dataverse capability that makes agents safer to ship. If you run Dynamics today, you are running a CRM that is becoming an agent runtime underneath you.
The product family
Dynamics 365 in 2026 is structured around customer experience, service, finance, supply chain, and SMB:
- Sales — opportunity, account, lead, sequences, sales accelerator, forecast, Sales Copilot, and the new Sales Agent.
- Customer Service — case, knowledge, omnichannel, Customer Service Copilot, Customer Service Agent, unified routing engine.
- Contact Center — standalone contact-center SKU for organizations using non-Dynamics CRMs at the desktop. Voice, IVR, agentic deflection.
- Field Service — work order, scheduling, mobile technician app, IoT integration, asset management, the Field Service Copilot.
- Customer Insights (Data + Journeys) — the unified customer data platform plus real-time marketing journeys; the rebuild of Marketing on top of CDP.
- Finance — general ledger, accounts payable/receivable, budgeting, fixed assets, expense, electronic invoicing.
- Supply Chain Management — inventory, warehousing, manufacturing, planning, procurement, and the IoT-connected production floor.
- Project Operations — project accounting, resource scheduling, time and expense.
- Commerce — POS, e-commerce, call center, merchandising.
- Business Central — the SMB ERP (formerly NAV) with its own agentic ERP and Sales agents shipping in 2026.
- Power Platform layer — Dataverse, Power Apps (canvas + model-driven), Power Automate, Power Pages, Copilot Studio, AI Builder, Power Fx.
- Microsoft 365 Copilot for Sales / Service — the Copilot surfaces inside Outlook, Teams, and Word that read from and write to Dynamics.
If you draw the architecture, Dataverse is the bottom layer, the Dynamics first-party apps and Power Platform sit above it, Copilot Studio orchestrates agents across them, and Microsoft 365 Copilot is the consumption surface that matters to most end users.
2026 release pulse
Microsoft ships two waves a year for Dynamics and Power Platform.
Wave 2 2025 (October 2025) set up Q1 2026 — the Customer Service Agent moved to GA, Sales Copilot got the account-summary upgrade, and Copilot Studio gained the first MCP server connectors.
Wave 1 2026 (April 2026, rolling out now) is the agentic release. Highlights:
- Customer Service Agentic Capabilities went GA — autonomous case triage, summarization, knowledge grounding, and resolution drafting across nine channels.
- The Sales Agent shipped for daily command and mobile context, including in-Teams and in-Outlook briefings.
- Business Central got its agentic ERP and agentic Sales agents — the first major agentic capability in the SMB SKU.
- Copilot Studio MCP integration shipped general availability with patterns for both inbound (agent calls external MCP) and outbound (external agent calls Dataverse via MCP).
- Customer Insights Journeys real-time picked up trigger-based agentic next-best-action.
- Contact Center self-service deflection moved to a model-routing architecture under AI Builder.
- Dataverse Elastic Tables matured for high-volume telemetry workloads.
Wave 2 2026 (expected October) is where Microsoft is signaling broad availability of inter-agent orchestration patterns, deeper Copilot Studio governance, and the long-promised on-prem-to-online migration tooling improvements that Dynamics CE customers have been waiting on for two years.
The pattern: Wave 1 each year is the “make it real” wave, Wave 2 is the “wire it together” wave. Plan your roadmap accordingly.
Who actually buys it
Dynamics 365’s buyer profile in 2026 is more diverse than Salesforce’s because it spans both CRM and ERP.
- Segment: SMB through enterprise. Business Central anchors the lower end (50-500 employees). The CE apps (Sales, Customer Service, Field Service) play strongest in mid-market through enterprise. Finance and SCM target $500M-$10B revenue companies.
- Company size: Median Dynamics CE customer is around 1,500 employees. Median Business Central customer is around 150 employees. Finance/SCM customers average 5,000+.
- Industry concentration: Manufacturing, professional services, distribution, retail, healthcare, government, and the entire Microsoft-friendly mid-market. Field Service skews to telco, utilities, and equipment OEMs.
- Average ACV: CE apps average $60-150K ACV. Finance and SCM deals routinely run into seven figures. Business Central averages $20-60K.
- Logos: Walgreens, Coca-Cola Bottling, Chevron, Toyota Material Handling, Lowe’s, the State of Texas, Mercedes-Benz, BP, the U.S. Air Force, and tens of thousands of mid-market companies you’ve never heard of running Business Central.
Pricing reality
Microsoft’s price card is public but the per-app, multi-app, and Copilot add-on math is where buyers get surprised.
| SKU / Tier | List (per user / month) | Notes |
|---|---|---|
| Sales Professional | $65 | Limited customization, no marketing |
| Sales Enterprise | $105 | Full customization, sequences, accelerator |
| Sales Premium | $150 | Includes some Copilot capacity |
| Customer Service Professional | $50 | Limited channels |
| Customer Service Enterprise | $105 | Full omnichannel base |
| Field Service | $105 | Includes scheduling and mobile |
| Customer Insights (Data + Journeys) | $1,700 / tenant / month | Plus per-profile and per-interaction overage |
| Finance | $210 | First app; subsequent users $30 |
| Supply Chain Management | $210 | First app; subsequent users $30 |
| Business Central Essentials | $70 | Includes financials, sales, purchasing |
| Business Central Premium | $100 | Adds manufacturing, service mgmt |
| Copilot for Sales (M365 add-on) | $50 | On top of M365 license |
| Copilot Studio messages | $200 / 25,000 messages / month | Consumption pricing for agents |
| Copilot for Service | $50 | On top of M365 |
The gotchas that surprise buyers:
- First app vs subsequent app pricing: The first Dynamics app at full price, the second app at $30 — which sounds great until you realize the “qualifying first app” definitions narrowed in 2024 and many customers don’t actually qualify.
- Copilot Studio messages: Each autonomous agent action consumes messages. Triage agents in Customer Service can burn 5-10 messages per case. A 50,000-case month is 250,000-500,000 messages.
- Customer Insights consumption overages: The base tenant SKU includes profiles and interactions; busy B2C deployments routinely 3-5x the included allowance.
- Storage: Dataverse storage is separately metered and Customer Insights ingest can fill it fast.
- AI Builder credits: AI Builder for forms, prediction, and OCR runs on a credit pool sold separately from Copilot Studio messages. Two separate budget lines.
- On-prem-to-online migration: Costs Microsoft doesn’t list — partner labor, license duplication during cutover, integration retrofits.
Strengths and weaknesses
Strengths
- The Microsoft stack integration is the deepest in the category. Entra ID, Teams, Outlook, Word, Excel, Power BI, Fabric, and Azure all behave like first-party. No other CRM ships a Teams app this good.
- Dataverse plus Power Platform plus Copilot Studio is a genuine low-code platform. Customers ship custom apps in days, not quarters.
- The combined CRM + ERP + SMB ERP footprint is unmatched. You can run a single Microsoft contract from a 50-person Business Central deployment to a 50,000-person Finance + Sales + Service rollout.
- Copilot Studio MCP integration is genuinely useful — the agent extensibility model is more open than Salesforce’s at present.
- Per-user pricing on the second app onward is a real cost advantage versus Salesforce when you’re standing up multiple workloads.
Weaknesses
- The Customer Insights rebuild on the new CDP foundation has been bumpy. Marketers who used the legacy Marketing app have spent two years complaining about parity gaps that are now mostly closed but still cited in losses to HubSpot.
- The model-driven app UI, while stable, feels dated next to the latest LWR-based Salesforce experience and HubSpot’s Smart CRM cards. Power Pages helps but doesn’t replace the modern web app feel.
- Documentation is fragmented across Microsoft Learn, the Power Platform docs, and the Dynamics customer engagement docs, and the search experience across them is inconsistent.
- Plugin development, the Web API surface, and the FetchXML-vs-OData decision still confuse experienced devs coming from cleaner platforms.
- Solution layering, dependency management, and ALM are non-trivial. The “dependency hell” failure mode has cost organizations entire deployment cycles.
The implementation truth
A Sales + Customer Service rollout for 500 users runs 16-22 weeks with a competent partner. Customer Insights Journeys is a 4-6 month project on top of source-system integration. Finance and Supply Chain are 12-24 month programs that look more like ERP consulting than CRM rollouts. Business Central deployments for SMBs run 8-16 weeks if scope holds.
Partner ecosystem: the top tier is the Big Four plus Avanade (the Microsoft-dedicated JV with Accenture). Below that, Microsoft specialists — HSO, Hitachi Solutions, RSM, DXC, Sikich, Western Computer for Business Central — do most mid-market work. The long tail of smaller Dynamics CE partners varies sharply in quality. Vet hard.
The most common failure modes I see on rescue projects:
- Customer Insights deployed without identity resolution rules figured out. Profiles split, the “single customer view” diverges, marketing loses confidence inside two quarters.
- Plugins written instead of flows. Custom code for what Power Automate could do creates support debt nobody wants to inherit.
- Solution layering ignored. Managed solutions get patched in production, layers get inverted, deployments fail and rollback is non-trivial.
- Copilot Studio agents shipped without grounding tests. Hallucinations slip into customer-facing journeys, trust erodes, the project gets paused.
- Storage and Dataverse capacity not modeled. Customer Insights ingest fills the tenant, costs spike at renewal.
// The Power Automate vs Plugin vs JavaScript decision is the single most important
// architecture call in any Dynamics deployment. The honest decision tree:
// - Real-time + transactional + needs server-side validation → Plugin
// - Async + can tolerate seconds → Power Automate flow
// - User-experience-only logic → Form script (and only if a Business Rule won't do)Decision framework
| Choose Dynamics 365 if… | Skip it if… |
|---|---|
| You run Microsoft 365 and Azure as your primary stack | You run Google Workspace and AWS exclusively |
| You need CRM + ERP under one vendor and one contract | You only need CRM and don’t care about ERP fit |
| You want low-code via Power Platform alongside CRM | You have a strong custom-dev team that prefers code-first |
| Your business model fits Business Central or Finance + SCM | You’re a B2C with needs only Marketing Cloud or Klaviyo solve |
| Copilot in Outlook and Teams is core to user adoption | Your users live in Slack and Notion, not Microsoft tools |
| You can budget Copilot Studio message consumption | You need a fixed-per-user agentic price model |
Where to go deeper
- Dynamics 365 2026 Wave 1 Overview
- Dynamics 365 2026 Wave 1 Deep Dive
- Customer Service Agentic Capabilities
- Sales Agent Daily Command
- Business Central Agentic ERP
- Copilot Studio MCP Integration
- Customer Insights Journeys Real-Time
- Dataverse Data Model
- Dynamics 365 Power Platform Integration
- Solutions ALM
- On-Prem to Online Migration 2026
- Plugin vs Flow vs JS Web Resource Decision
What to do this week
If you already run Dynamics, audit your Copilot Studio message burn for the last 30 days, count your Customer Insights profiles against your contract, and confirm your Wave 1 2026 update slot. If you’re evaluating, get a Microsoft partner to run a one-day Sales + Customer Service envisioning session against your real data — Dynamics’s value shows up the moment you see your Outlook, Teams, and Excel surfaces in the demo. The platform is at its strongest when the integration with Microsoft 365 is not theoretical.
