A practical guide to white-labelling for agencies and B2B platforms — what it really means, the features that matter, and which platforms deliver true white-labelling versus token customisation.
TL;DR White-label embedded analytics means delivering dashboards to your customers under your brand, your domain, and your visual identity — not the vendor's. The label "white-label" gets attached to platforms that offer wildly different levels of actual customisation, from full domain control and theming down to a logo swap. For agencies and B2B platforms where dashboards are part of your client-facing brand, the difference matters. Real white-labelling needs custom domain, full theming, no vendor logos, per-client branding, and a hidden-vendor experience. Platforms that don't deliver all five aren't really white-label, regardless of marketing copy.
What white-label embedded analytics actually means
White-label, at its strictest definition, means your customers cannot tell that the analytics are powered by a third-party platform. They see your brand, your URL, your design language. The platform vendor is invisible.
That strict definition matters because the term gets used loosely. Vendors describe products as "white-label" when they let you upload a logo, when they let you choose two colours, when they allow a CNAME-pointed subdomain, or when they offer all of those plus full theming and CSS control. These are different products at different price points, often within the same vendor's catalogue.
For agencies delivering analytics to clients, the difference is the difference between a polished professional service and a clearly-outsourced add-on. For SaaS products embedding analytics as a feature, the difference is between analytics-as-your-product and analytics-as-a-Microsoft-or-Tableau-thing-bolted-on.
This blog covers what real white-labelling requires, what to look for, and which platforms in 2026 actually deliver versus market it.
For the broader context, see our embedded analytics overview and the 10 best embedded analytics platforms in 2026.
The five must-have features for real white-labelling
If a platform doesn't deliver all five of these, it's not genuinely white-label — regardless of marketing language. Use this as a checklist when evaluating.
1. Custom domain (not just a vendor subdomain). Customers access analytics at analytics.yourcompany.com or reports.yourdomain.com, not at yourcompany.vendor.com. CNAME pointing is fine; what matters is that the visible URL is yours. Vendor subdomains are a partial brand experience, not white-label.
2. No vendor branding anywhere in the user experience. No "Powered by [Vendor]" footer. No vendor logo in the favicon. No vendor name in the page title. No vendor watermark on exported PDFs. Customers should never see the vendor's brand unless they're actively looking for it in source code.
3. Full visual theming, not just logo and colours. Your typography, your colour palette across every UI element (not just the header), your spacing, your iconography. Real white-labelling lets you make the analytics feel like part of your product; logo-and-two-colours customisation makes it feel like a vendor's product with your sticker on it.
4. Per-client or per-tenant branding (for agencies and multi-tenant SaaS). Agencies serving multiple clients need each client to see their own brand, not the agency's brand. SaaS products with branded customer organisations need each tenant to potentially have their own theme. One brand per platform is fine for SaaS with a single brand; not enough for agencies or multi-brand platforms.
5. Authentication and email under your domain. Login pages, password reset emails, notification emails — all should come from your domain and look like part of your product. Customers getting emails from the vendor's domain about "your analytics" breaks the illusion immediately.
Platforms that deliver all five include Power BI Embedded with DataTako, dedicated multi-tenant platforms like Sisense and Qrvey, and developer-first platforms like Cube and Embeddable (where you build the UI yourself so white-labelling is implicit).
Platforms that deliver some-but-not-all include Tableau Embedded (custom domain available but vendor branding lingers), Looker Embedded (depends on tier), and Power BI Embedded without a delivery layer (you build the white-labelling yourself).
Platforms that market white-label but deliver token customisation include some lower-tier configurations of various BI tools — read the contract carefully.
What white-label looks like in practice
Two real-world patterns illustrate what good white-labelling delivers:
An agency delivering client reporting. The agency serves twenty clients across financial services, healthcare, and retail. Each client logs in at their own branded subdomain (acme-reports.agency.com, globex-reports.agency.com), sees their own logo, their own colours, and their own dashboards. No client knows that the agency uses Power BI Embedded underneath, and no client sees another client's brand. The agency's brand appears subtly as the platform provider, but the primary experience is the end-client's.
This pattern requires per-client branding (different visual themes per tenant), tenant isolation (Row-Level Security plus visual separation), and a delivery layer that handles all of this automatically — typically Power BI Embedded with DataTako, or a purpose-built embedded platform like Qrvey.
A SaaS product embedding analytics. The SaaS product is an HR platform with 200 customer organisations. Each customer logs into the HR product at customer.hrproduct.com and sees workforce analytics integrated into the navigation as a tab. The dashboards look like part of the HR product — same typography, same colour palette, same navigation patterns. End users don't realise a third-party analytics platform is rendering the dashboards.
This pattern requires full visual theming (matching the SaaS product's design system), no vendor branding (the HR product's brand is the only visible brand), and authentication integrated with the SaaS product's existing identity layer. Power BI Embedded with DataTako delivers this; Tableau Embedded delivers a less complete version where vendor branding sometimes lingers.
In both patterns, the platform underneath is invisible. That's the bar.
What "white-label" usually means in vendor marketing
Vendors describe products as white-label when they offer varying levels of customisation. Here's what to actually expect from the marketing language:
"Custom branding" usually means a logo upload field and two-three colour pickers. It's better than nothing but it's not white-label — the platform's UI structure, typography, and remaining UI elements still feel like the vendor's product.
"White-label option" usually means a higher-tier subscription where you get a custom domain plus the branding options. The custom domain alone is a big upgrade; whether it's truly white-label depends on whether vendor branding actually disappears (often it doesn't, completely).
"Fully white-label" is the marketing claim that most needs verification. Sometimes it's true (the vendor really has removed all branding and theming hooks). Sometimes it's marketing language for "you can hide our logo but the URL bar still says our domain." Verify the five must-haves above before trusting the claim.
"Custom embedding" usually means the vendor exposes APIs for you to build whatever UI you want around their analytics engine. This is genuinely white-label by default — because you build the UI, the vendor's brand never appears. Examples: Cube, Embeddable. The trade-off is that you do the building work yourself.
The clearest test: ask the vendor for screenshots of three different customers' deployments. If the screenshots all look like the same vendor product with different logos, it's not white-label. If they look like three different products with no apparent vendor connection, it is.
Platform-by-platform white-label reality check in 2026
Power BI Embedded (Microsoft). Out of the box, Power BI Embedded provides embedding APIs but no white-label portal. To get true white-labelling — custom domain, full theming, no Microsoft branding, per-client variation — you either build it yourself (4-6 months of engineering work) or use a delivery layer like DataTako. With DataTako, all five must-haves are checked. Without it, you're building the portal infrastructure yourself. See Power BI Embedded.
Tableau Embedded (Salesforce). Strong embedding APIs and customisation options at higher tiers. Custom domains are supported. Tableau's branding is generally well-hidden in embedded mode. Per-tenant branding is harder — Tableau's multi-tenant patterns work but per-tenant theme isolation requires careful configuration. White-label in spirit but the implementation is heavier than dedicated solutions.
Looker Embedded (Google). Higher-tier plans include white-labelling features. Custom domains supported. LookML-based modelling means the developer experience is strong but the visual customisation is more code-driven than UI-driven. Good for engineering-led teams; less straightforward for non-engineering customers.
Sisense. Purpose-built for embedded scenarios with strong white-label primitives. Custom domains, full theming, per-tenant branding all supported natively. Strong choice for agencies and multi-tenant SaaS specifically wanting white-label.
Qrvey. Multi-tenant SaaS specialist with white-labelling built in. Per-tenant branding, custom domains, no vendor branding in the user experience. AWS-native deployment lets you put the analytics entirely under your own infrastructure boundary, which some compliance-heavy industries require.
Cube and Embeddable. Developer-first platforms where you build the UI yourself. White-label by default because the vendor's UI doesn't exist in your product — only the analytics backend does. The trade-off: you build the dashboards, charts, filters, and interactivity yourself.
Domo. Marketing claims white-label support but the implementation is closer to branding customisation than true white-label. Vendor branding tends to persist in unexpected places (export PDFs, certain UI elements). Evaluate carefully if white-label is critical.
Metabase. Open-source self-hosted gives you full white-label control because you control everything. Paid plans (Pro, Enterprise) add white-label features to the cloud offering. Good fit for cost-sensitive SaaS that wants white-label without enterprise platform costs.
ThoughtSpot Embedded. Embedding customisation has improved but white-label depth varies by tier. The search-driven UX pattern (rather than traditional dashboards) can feel non-standard for end-users — independent of branding.
The pattern: dedicated embedded platforms (Sisense, Qrvey, Cube, Embeddable, Metabase) generally deliver real white-label naturally. General-purpose BI tools (Power BI, Tableau, Looker, Domo) deliver it less natively, requiring either higher tiers or delivery layers to achieve real white-labelling.
White-label patterns for agencies specifically
Agencies have white-label requirements that consumer SaaS products don't:
Per-client branding, not just one brand. Each client gets their own logo, colours, and possibly subdomain. The agency's brand appears as the platform provider but each client's experience is dominated by their own brand.
Client-visible reports under client domains. When a client receives a scheduled PDF report or shares a dashboard link with their own team, the URL and branding should be the client's, not the agency's. This requires per-client subdomains or domain delegation.
Reseller pricing transparency. Some agencies want to white-label so completely that they hide the underlying platform from clients entirely, charging their own pricing. Others want to be transparent that they use, say, Power BI Embedded underneath. Both are valid; the platform should support both.
Client-specific configuration. Different clients may want different default dashboards, different access patterns, different export settings. Per-client configuration is part of agency-grade white-labelling.
Branded login experiences. When the agency's client logs into the analytics portal, they should see their own brand, not the agency's "powered by" attribution. Authentication branding matters more than people expect.
For agencies on Power BI Embedded, DataTako handles all five of these natively. For agencies on other platforms, capabilities vary — see the platform-by-platform breakdown above.
See the BI agency playbook and marketing agency playbook for the specific patterns each industry uses.
What white-labelling costs
A few honest cost notes:
True white-labelling lives in higher tiers. Almost every embedded analytics platform has white-label features behind paywalls. The cheapest tier rarely includes them; the middle tier sometimes does; the enterprise tier almost always does. Budget for the white-label tier from the start, not the cheapest available.
Custom domains have configuration costs. Pointing a CNAME and setting up SSL is straightforward but not zero work. Per-tenant custom domains (each client at their own subdomain) is more complex and typically requires platform support.
Per-tenant branding is more expensive than single-tenant. Platforms that support different visual themes per tenant generally charge for the capability. For agencies serving multiple clients with multiple brands, this is unavoidable.
Building white-label yourself is far more expensive than buying it. A custom white-label portal on top of Power BI Embedded or Tableau Embedded is typically 3-4 months of engineering work — $50,000-$100,000 in salaries alone. Compared to €200-€500 per month for a platform that includes white-labelling, the build path makes sense only when analytics is your core product.
For the cost dynamics specifically on Power BI, see the hidden cost of Power BI Pro for BI agencies and the optimise Power BI licence costs page.
How DataTako handles white-labelling on Power BI Embedded
DataTako exists specifically to solve white-labelling for Power BI Embedded. Microsoft provides excellent embedding APIs but no white-label portal — that's what DataTako is.
What DataTako provides on the white-label dimension:
Custom domain by default. Customers log in at analytics.yourcompany.com (or reports.yourcompany.com, or whatever subdomain you choose). No Microsoft URLs visible.
Full visual theming. Logo, colours, typography, navigation, error pages, login screens — all themed to your brand. No "Powered by Microsoft" or DataTako attributions in the customer-facing experience.
Per-client branding for agencies. Each tenant in your DataTako instance can have its own branding. Twenty agency clients can each have their own logo, colours, and visual identity.
Branded authentication. Login pages, password reset emails, notification emails all come from your domain and match your brand. Customers never see "Microsoft" or "DataTako" in the authentication flow.
Multi-tenant isolation with branding consistency. Row-Level Security ensures each tenant sees only their own data; branding configuration ensures each tenant sees their own brand. The two work together so tenants never accidentally see another tenant's experience.
The setup time: hours, not months. For agencies running Power BI Embedded who tried to build their own white-label portal, the typical outcome is 4-6 months of engineering followed by perpetual maintenance. With DataTako, the same outcome is configuration, not engineering.
See how DataTako works and the MeerMetData case study for concrete examples.
Frequently asked questions
What does white-label embedded analytics actually mean? White-label means delivering analytics to your customers under your brand, your domain, and your visual identity, with no visible third-party vendor branding. Strict white-label requires custom domain, no vendor branding, full theming, per-tenant branding (for agencies), and branded authentication. Many platforms describe themselves as white-label but only deliver some of these.
Which platforms truly deliver white-label embedded analytics? Dedicated embedded platforms generally do (Sisense, Qrvey, Cube, Embeddable). General-purpose BI tools (Power BI, Tableau, Looker) require either higher tiers or delivery layers to achieve real white-label. Power BI Embedded with DataTako delivers full white-label; Power BI Embedded alone does not.
How much does white-label embedded analytics cost? Platform tiers with white-label features typically run €500-€2,500 per month depending on platform and audience size. Power BI Embedded with DataTako runs €500-€2,000 monthly for unlimited viewers including white-labelling. Tableau and Looker white-label tiers run significantly higher.
Can agencies use white-label embedded analytics for multiple clients? Yes — this is one of the primary use cases. Agencies need per-client branding (each client sees their own brand), per-client subdomains or domain delegation, and per-client configuration. Power BI Embedded with DataTako handles all of this natively, as do dedicated multi-tenant platforms like Qrvey and Sisense.
What's the difference between white-label and embedded? Embedded means analytics live inside your application rather than in a separate tool. White-label is a specific quality of embedded — your branding rather than the vendor's. Embedded without white-label still shows the vendor's brand; white-label embedded hides the vendor entirely.
Do customers need to know my analytics are powered by Power BI? In true white-label deployments, no — they don't know and don't need to. The Power BI Embedded engine renders the dashboards, but customers see only your brand, your domain, and your design. Some agencies do choose to mention "powered by Microsoft Power BI" as a credibility signal; others hide it entirely. Both are valid.
Can I white-label Power BI Embedded myself? Yes, but it's significant engineering work. You need to build a custom portal that wraps Power BI Embedded's APIs, handle authentication, theming, per-tenant branding, and operational infrastructure. Typically 4-6 months of work. DataTako provides this layer so you don't have to build it.
What about white-label PDF exports? A frequently overlooked white-label dimension. Exported PDFs should carry your branding, not the vendor's. Check this specifically — some platforms add vendor watermarks or branding to exports even when the UI is fully white-labelled.
Is white-label necessary for B2B SaaS? It depends on positioning. If your product positions itself as a complete platform and analytics is a feature, white-label keeps the experience consistent. If you market analytics as a separate component or third-party integration, full white-label matters less. For most B2B SaaS, white-label is worth the investment.
Does DataTako only support Power BI Embedded, or other platforms too? DataTako is specifically built on Power BI Embedded. If you've chosen another platform (Tableau, Sisense, etc.), DataTako isn't relevant. If you're choosing Power BI Embedded or are open to it, DataTako provides the white-label delivery layer.
