Your domain, your logo, your product

Why the logo on the dashboard is a commercial decision

Agencies are often slightly embarrassed to care about white labelling. It sounds like vanity — the work is the work, the client knows you use tools, why pretend otherwise. That framing misses what is actually being bought and sold.

A client who logs into a vendor's dashboard learns three things over time. They learn the vendor's name. They learn the vendor exists as a purchasable thing. And they learn, implicitly, that a component of what they pay you for is a subscription somebody else operates. None of that matters at all until the day a finance director runs a cost review, types the vendor's name into a search engine, finds a public price list, and asks the obvious question.

That question — "why are we paying you £2,000 when this appears to cost £299" — is unanswerable in the moment, even though the answer is straightforward and good. The agency is being paid for strategy, campaign management, creative and the human being who answers when something breaks; the software is a small input. But you are now defending your margin rather than discussing results, and that conversation has a meaningfully worse expected outcome than the one you were having before.

The same dynamic plays out when a client's in-house marketer arrives, or when a new operations director starts and reviews every supplier in their first month. Both are routine events, both are moments when an agency's value has to be legible in about ninety seconds, and both go better when what the newcomer finds is a platform with your name on it rather than a login to something they can price.

White labelling does not deceive anybody. Clients know agencies use software; most would assume you built or licensed something regardless. What it does is stop the client's attention from landing on a specific price list at a specific moment, and keep the conversation about what you deliver rather than what you buy.

The larger effect is on how the agency is perceived. An agency with a platform at their own domain, with their branding, that clients log into and receive reports from, looks structurally different from one that emails spreadsheets. It supports higher prices, it survives comparison with larger competitors, and — the part nobody mentions — it is materially harder to leave. Not because the client is trapped, but because the relationship has an artefact in it rather than being a series of invoices.

What "white label" actually covers here

The term gets used loosely, so here is the specific list. Anything not on it is not claimed.

Multi-tenancy is the part that has to be true

Branding is cosmetic. What sits underneath it is not, and it is where most "white-label" platforms are quietly weakest.

The requirement is simple to state and easy to get wrong: a login belonging to one client must not be able to reach another client's data, no matter what request it sends. Not "must not see a link to it" — must not be able to reach it by changing an identifier in a URL, by replaying an old token, by calling an API endpoint directly, by requesting an export, or by receiving a webhook meant for somebody else.

The common failure is to enforce isolation in the interface. The dashboard shows the client only their own accounts, so nobody notices that the underlying endpoint will happily return any account it is asked for. That system behaves correctly for every user who uses it as intended and fails completely for the first person who opens the network tab. For an agency reselling to competing businesses in the same town, that is not a theoretical risk — it is the end of the agency.

Here, the tenant scope is applied in the query rather than the view. Every read carries the agency and client it is permitted to see, and a request for something outside that scope returns nothing rather than returning data the interface would have hidden. The same rule applies to background jobs, to scheduled reports, to data exports and to inbound webhooks, because each of those is a path that bypasses the interface entirely.

The stakes are higher for an agency than for a single business, and specifically higher in the case agencies most often build towards. A vertical specialist — the agency that works only with dentists, or only with personal injury firms — ends up with a client base of direct competitors, frequently in the same city, sometimes bidding on each other's brand terms. A leak between two of those tenants is not an incident report; it is the end of the agency and quite possibly a legal matter. Specialisation is the most reliable way to grow a local agency, and it raises the cost of weak isolation from embarrassing to existential.

It is worth asking any vendor a specific question about this: if a client's login sends a request for a record ID belonging to a different client, what happens? A good answer describes server-side scoping. A vague answer about roles and permissions usually means the check happens somewhere it can be skipped.

Setting it up

The whole thing is an afternoon, most of which is waiting for DNS.

1. 1

   Point a subdomain at the platform

   A CNAME on a subdomain you control. The certificate is issued and renewed automatically, so there is nothing to remember in eleven months.

2. 2

   Upload your branding

   Logo, favicon, primary and accent colours, product name. It applies across the interface, the reports and the emails at once rather than needing to be set in three places.

3. 3

   Verify your sending domain

   SPF and DKIM records so mail from your address authenticates properly. This is the step agencies skip and then wonder why client reports land in spam.

4. 4

   Create your clients

   Each becomes an isolated tenant with its own users, numbers, domains, data and reports. A client is not a filter on a shared dataset; it is a boundary.

5. 5

   Decide what each client can see

   Role tiers control whether a client user can view reports only, manage their own leads, or configure settings. Most agencies start read-only and loosen selectively.

The three ways agencies charge for this

All three work. Which one fits depends on your client mix and how much variance there is in usage between accounts.

Sub-agencies, and reselling to other agencies

The hierarchy goes deeper than agency and client, which matters for two situations that come up more often than people expect.

The first is the agency with distinct brands. A group running a trades-focused brand and a professional-services brand does not want one logo across both, and does not want the two client bases sharing anything. Each brand is its own tenant with its own domain, branding and isolation, sitting under one commercial relationship.

The same structure covers the agency that has acquired another and is running both names during a transition, and the one operating a regional brand alongside a national one. In every case the requirement is identical: separate presentation, separate data, shared administration.

The second is reselling to other agencies. A larger agency, or a marketing consultant with a network, can operate their own sub-agencies — each with their own branding, their own clients, and no visibility of each other. The parent sets what each can do and sees usage across the whole tree; the sub-agency sees only its own world and its clients see only theirs.

This is genuinely useful and it is worth being clear about what it demands. Reselling means becoming somebody's support desk. The sub-agency will ask you why a number is not swapping and expect an answer, and if you are not prepared to be that answer, the model produces unhappy partners rather than revenue. Agencies who succeed at it treat it as a product line with its own onboarding and documentation rather than as passive margin.

The technical side is straightforward because the isolation model is the same at every level. A sub-agency is a tenant in exactly the way a client is, enforced by the same server-side scoping, which means the depth of the tree does not introduce new places for data to leak.

What stays visible to you and invisible to them

The dividing line matters. Your clients should see their marketing. They should not see the machinery, and in some cases seeing it would be actively harmful.

• Platform configuration is yours alone. Telephony credentials, provider settings, infrastructure configuration and cost data are never exposed to a client tenant, regardless of role.
• Your costs stay your costs. What you pay for minutes, numbers and transcription is not visible to a client, so an itemised platform fee is not silently accompanied by your margin.
• Other clients do not exist. A client tenant cannot enumerate, name or count other tenants. Not in the interface, not in an API response, not in an error message.
• Your agency sees everything it owns. Cross-client views, usage per account, health across the base — the things you need to run the business rather than one campaign.
• Role tiers are granular. Read-only client viewer, client manager, agency staff, agency owner. Most agencies keep clients read-only and that is the sensible default.
• Errors do not leak existence. Asking for a record belonging to another tenant returns not-found rather than forbidden, because the difference tells an attacker the record exists.

The margin arithmetic, done honestly

The commercial case for a white-labelled platform is usually presented as pure upside, and it is not — it is a fixed cost against a variable benefit, and it is worth doing the sums before the pitch rather than after.

The fixed side is straightforward: a monthly platform subscription, plus telephony and transcription usage that scales with how busy your clients are. The variable side is harder to pin down and considerably larger. It includes the reporting hours you stop spending, the retainers that survive a review because the client could see value, and the accounts you win against competitors who email spreadsheets.

On a ten-client agency the arithmetic is usually decisive. Reporting alone, at ninety minutes per client per month, is fifteen hours — call it two full days of somebody's time. That exceeds the entire platform cost before you count a single retained account. Below about five clients it is genuinely marginal, which is why the Starter plan exists without white labelling: at that size the branding is not yet the thing earning its keep.

The number that decides whether the model keeps working is usage variance. One client receiving four hundred calls a month can consume more telephony and transcription than five ordinary accounts combined, and if the platform is absorbed into a flat retainer, that account is quietly subsidised by the others. Per-client usage metering is what makes that visible — and visible is enough, because the fix is usually a conversation rather than a system change.

The mistake worth avoiding is pricing the platform as a cost to recover rather than a capability to sell. Agencies who add £40 a month to each invoice and call it a "reporting fee" have created a line item for the client to question. Agencies who raise their retainer because they now deliver tracked, attributed, live-reported results have sold something. Same money, entirely different conversation.

What white labelling does not do

It does not stop a client from finding out that software exists. Anybody sufficiently curious can work out that an agency did not build a call tracking platform from scratch, and no serious client believes otherwise. The goal is not concealment, it is that the client's attention is on your work rather than on a vendor's pricing page at the moment they are deciding whether to renew.

It does not make you a software company. The platform is a delivery mechanism for services you provide; agencies who start describing themselves as a technology business on the strength of a rebranded dashboard tend to attract questions they cannot answer.

It does not remove your obligations to your clients' data. Under white labelling, you are the party the client has a relationship with, which means data requests, deletion requests and questions about where information is held come to you. The tooling exists to answer them — per-client export and deletion are built in — but the accountability is yours and it is worth being deliberate about that rather than discovering it during a request.

It does not, on its own, justify a higher price. A rebranded dashboard attached to mediocre work is a rebranded dashboard attached to mediocre work, and clients establish that faster than agencies expect. What white labelling does is make good work legible — it gives competence somewhere to be visible. That is worth a great deal and it is worth nothing without the competence underneath it.

And it does not do anything for a client who never logs in. A branded dashboard nobody opens is a branded dashboard nobody opens. The value comes from the client actually using it, which is a function of whether the reporting answers their questions — which is why white labelling and client reporting are two halves of the same argument.

Common questions