Business softwareCustom softwarePricingIntegrationSMEs
Power Apps vs a custom app: which one fits your business?
10 min read
Power Apps is the right choice when your users are employees already on Microsoft 365, the app stays internal and fairly simple, and you need something working within weeks. A custom app wins once you need to serve customers or partners, handle large data volumes, deliver a polished user experience, or when per-user licensing starts to cost more than owning the software outright. Everything in between comes down to three numbers: how many users, for how many years, and which licenses you will actually have to buy.
This guide relies on Microsoft's own pricing pages and documentation, checked on November 6, 2026. It is not an argument against Power Apps. For plenty of internal needs, it is an excellent tool. The aim is to help you work out which side of the line your project sits on before you commit.
Power Apps in two minutes
Power Apps is Microsoft's low-code platform. You assemble screens with a visual designer, write Excel-like formulas in a language called Power Fx, and connect the app to data through connectors: SharePoint, Excel, Outlook, Teams, SQL Server, Salesforce and hundreds more. Apps come in two flavors:
- canvas apps, designed screen by screen, with free-form layouts;
- model-driven apps, generated from tables in Dataverse, the platform's own database. They are more rigid but quick to produce for a standard back office.
Power Apps belongs to the Power Platform, alongside Power Automate for workflows, Power BI for reporting and Power Pages for websites that external visitors can sign in to.
Licensing, as of November 6, 2026
Most comparisons skip this part, yet it usually decides the outcome. Prices below are US list prices from the Power Apps pricing page and Microsoft Learn. Microsoft notes that published prices are for marketing purposes; your reseller or enterprise agreement may differ, and prices vary by country.
| Plan | List price | What it covers |
|---|---|---|
| Rights included in Microsoft 365 | Included with some Microsoft 365 plans | Apps using Microsoft 365 data and standard connectors only |
| Power Apps Premium | $20 per user per month, paid yearly | Unlimited apps for that user, premium connectors, Dataverse |
| Premium, 2,000-seat minimum | $12 per user per month, paid yearly | The same, for very large organizations |
| Pay-as-you-go (Azure) | $10 per active user, per app, per month | Anyone who opens the app at least once that month |
| Extra Dataverse capacity | $40 per GB per month, paid yearly | Database storage beyond the included entitlement |
| Power Pages (external users) | $4 per active signed-in user, per website, per month, metered | A site for customers or partners, with data in Dataverse |
Sources: Power Apps pricing, pay-as-you-go meters, Power Platform licensing overview.
Four things to know before you budget:
- The per app plan is being phased out. Microsoft announced that the Power Apps per app license is no longer sold to new customers from January 2, 2026, with exceptions depending on how you buy (official announcement). For a new project, plan on Premium or pay-as-you-go.
- "Premium" is easy to trigger. According to the Power Apps licensing FAQ, an app becomes premium as soon as it uses a premium connector (SQL Server is one), a Dataverse table, an on-premises data gateway or a custom API. Model-driven apps are premium by default.
- Enforcement tightens in February 2027. The same FAQ states that from February 2027, scenarios where licensing was applied leniently (apps in managed environments, model-driven apps) will require a premium license, or users will no longer be able to open the app. If you run Power Apps "for free" today, audit your licenses before then.
- Dataverse has a quota. Each Premium license adds 250 MB of database and 2 GB of file capacity, pooled across your tenant. Beyond that, extra database capacity costs $40 per GB per month. Years of history with attachments can push you past it.
A quick calculation to redo with your own numbers
At list price, 30 employees on Power Apps Premium cost 30 × $20 × 12 = $7,200 per year, or $36,000 over five years, before building the app, maintaining it or buying extra storage. With 150 users, that line reaches $36,000 every single year. Those licenses cover every Power Apps app those users run, so if you already have several, the marginal cost of one more app is zero. That is Power Apps' strongest argument.
A custom app has no per-user cost, but it has a build cost and a maintenance cost. Our guide to custom business application costs breaks those down, and our comparison of a subscription versus an upfront build covers the two ways to pay for it.
When Power Apps is the right call
To be fair: in many small and mid-sized companies, Power Apps is the right answer. It fits when most of these are true:
- your users are employees already on Microsoft 365, and you may already own Premium licenses for other apps;
- your data lives in Microsoft tools: SharePoint lists, Excel files, Teams, Outlook, or Dataverse if you run Dynamics 365;
- the app is an upgraded form: purchase requests, leave approvals, inspection checklists, a small asset register;
- functional design is enough, with no branding requirement and no heavy daily use by hundreds of people;
- someone in-house can maintain it, or a Microsoft partner supports you;
- the lifespan is short or uncertain: you are testing a process and will revisit later.
In that setting, Power Apps delivers a tool within weeks, inside an environment your teams already know, with Microsoft 365 sign-in and security built in. No custom build will be faster or cheaper for that kind of need.
The limitations that tip the balance
External users
This is the clearest limit. A canvas app can be shared with outsiders, but the guest access documentation requires each guest to be invited into your Microsoft Entra tenant and to hold a matching license if the app uses Dataverse or premium connectors. For a genuine customer portal, Microsoft steers you toward Power Pages, billed per active user per site. If hundreds of customers sign in every month, the bill grows with your success. A custom app takes one more customer at no license cost. Our article on building a custom client portal goes deeper.
Data volume
Power Apps handles large datasets well as long as every query is delegated to the data source. When a formula cannot be delegated, the delegation documentation explains that the app only processes the first 500 records by default, up to a maximum of 2,000 if you raise the limit. Past that, a filter or an average can return a wrong result with no visible error. It is manageable, but it takes an experienced Power Apps developer who knows the delegation rules of each data source. At that point, you are a long way from "low-code for everyone".
User experience and heavy use
Power Apps produces decent, data-entry screens. If people use the app eight hours a day, in the field, offline on a phone, or if it carries your brand in front of customers, every detail matters: load times, keyboard shortcuts, screen flow. An interface designed around your process saves every user minutes a day.
Business logic and integrations
Complex pricing rules, logic that depends on dozens of parameters, an ERP with no connector, overnight jobs on thousands of rows: all of it is possible on the Power Platform, but by stacking Power Automate, custom connectors and sometimes Azure Functions. The project turns into real software development, with the platform's constraints on top.
Dependence on Microsoft
A Power Apps app only runs inside Power Apps. You can export the solution and your Dataverse data, but you cannot run the app anywhere else. License prices, quotas and compliance rules are Microsoft's call, as the February 2027 change shows. That is fine if your company has committed to Microsoft for the long term. It is a risk if the app becomes the core of your business, or if you might one day sell it to other companies as a SaaS product. Our vendor lock-in exit plan lists what you should be able to take with you, whatever the tool.
Decision table
| Factor | Lean toward Power Apps | Lean toward a custom app |
|---|---|---|
| Users | Employees already on Microsoft 365 | Customers, partners, or a very large workforce |
| Data | SharePoint, Excel, Dataverse, moderate volume | Large volumes, heavy calculations, non-Microsoft data |
| Integrations | Microsoft tools and existing connectors | ERP or line-of-business systems with no connector |
| User experience | Functional, occasional use | Heavy daily use, mobile, customer-facing branding |
| Cost | Licenses already paid for other apps | Per-user cost that outgrows the cost of owning it |
| Lifespan | Short or uncertain | Long, central to how you operate |
| Product potential | None | Could be sold to other companies as SaaS |
If your answers land on both sides, a hybrid works: a custom core (database and API) for the business logic, and a few internal Power Apps screens plugged into it through a custom connector. Keep in mind that the custom connector makes those screens premium.
Replacing MS Access with Power Apps: what migrates and what does not
Many teams ask this when retiring an aging Microsoft Access database. Microsoft provides an Access to Dataverse migration tool, started from Access by exporting tables. It moves tables, relationships and data. It does not turn forms, printable reports, complex queries or VBA code into Power Apps screens: those have to be rebuilt.
Two practical consequences:
- the licensing model changes: with Access you paid for Office; with Dataverse, every user of the app needs a premium license;
- the work is a rebuild, not a conversion. And once you are rebuilding, the Power Apps versus custom question is open again, with the criteria in the table above.
Our dedicated guide on how to convert an MS Access database to a web app walks through the method, from the table inventory to the cutover.
Common mistakes
- Building on a trial license, discovering licensing at launch. The app is built with a trial or developer license, and users get a prompt to buy on day one. Check the app's premium designation in the Power Apps maker portal before promising anything.
- Using SharePoint as a database to dodge Dataverse. It avoids premium licensing, but SharePoint lists have their own delegation and structural limits. For a few hundred rows, it works. For years of history, it is fragile.
- Leaving a critical app to a single citizen developer. When that person moves on, nobody can maintain it. Document it, or hand maintenance to a team.
- Comparing a monthly license with a development quote. Compare over five years, including licenses, maintenance and new features, on both sides.
- Going custom on principle. If you need an internal form for twenty people already on Microsoft 365, Power Apps is probably the sensible choice. Our build vs buy guide covers when not to build.
Outgrown a spreadsheet but not picked a tool yet?
Many Power Apps projects start from a spreadsheet that got out of hand. Before choosing the tool, describe the need: who enters data, who approves, who reads it, how many rows a year, which systems must exchange data. Our guide to replacing Excel with a custom application helps you build that inventory, which is useful whatever you choose.
Where we come in
LPR Studio is not a Microsoft partner and resells no licenses. We build custom business applications and publish our own software. If your project lands in the custom column of the table, we will show you a free prototype of your app over a video call within 5 business days, so you can compare on evidence rather than promises. If your need is clearly a Power Apps job, we will tell you that too. Not sure what a prototype should prove? Our article on prototypes, MVPs and proofs of concept explains the difference.