White-Label vs Bundled Apps for Operators

A mobile operator with a large subscriber base and a shrinking ARPU eventually asks the same question: should we put commerce inside our own app, or bundle somebody else’s?

Both answers are defensible and they are not the same business. This is what actually separates them — who owns the customer, where the security boundary sits, and what each one costs to keep running after the launch press release.

The two shapes

Bundling means partnering. Your subscribers get a partner’s app, or a partner’s service inside a tab in yours, usually with the data allowance zero-rated and the billing going through the partner. You are supplying distribution and, sometimes, a payment rail. The partner supplies the product.

White-label means the marketplace is yours. Your brand, your app, your merchant relationships, your customer records, your take rate. A platform vendor supplies the software underneath and is not a name your subscriber ever sees.

The temptation is to read these as “cheap and fast” versus “expensive and slow”. That is roughly true of the first six months and roughly false of everything after, which is why the decision is worth making on something other than launch cost.

Who owns the customer

This is the question that decides the rest, so it is worth being blunt about it.

Under a bundle, the partner owns the transaction and therefore owns the behavioural data. You learn that a subscriber used the tab. The partner learns what they bought, how often, at what price point, and how they respond to a discount. When the partnership ends — and partnerships end — you keep the subscribers you already had and none of the commerce relationship you spent three years subsidising.

Under a white-label arrangement, the order record is yours. So is the merchant relationship, which matters more than it sounds: the merchants on your marketplace are a second customer base, they have their own churn characteristics, and they are the thing a competitor cannot copy by signing the same partner you did.

There is an honest counter-argument. If your realistic ambition is to reduce churn by a point or two and you have no intention of running a commerce business, owning that data is an asset you will not use. Bundling is the right answer more often than vendors of platforms like to admit.

Where the security boundary sits

This is the part that gets underweighted in the business case and then dominates the second year.

A bundled app is a third party inside your trust boundary, or beside it. If it is a genuinely separate app, the boundary is clean and your exposure is mostly reputational — your subscribers associate the experience with you whether or not your name is on the code. If it is embedded in your app, via an SDK or a webview, then the partner’s release cadence, their dependency tree and their incident response are now inside yours. That is a governance question, not a technical one, and the honest version of it is: can you get an answer out of that partner at 2am on a Sunday, and is it in the contract?

A white-label platform moves the boundary rather than removing it. The vendor is a processor handling your subscribers’ data, so you need the same things you would need from any processor: tenant isolation you can have described to you in specifics, a subprocessor list you can read, a payment integration where the card data does not land in scope you have to defend, and a written answer about what happens to the data if you leave. The difference from a bundle is that these are your controls to specify, rather than a partner’s controls to accept.

Neither shape is inherently safer. What differs is who is accountable when a regulator asks, and under a bundle the answer is more complicated than the slide deck suggested.

What each costs to run

Launch cost is the number that gets compared and the least useful one. Three lines matter more:

Merchant acquisition. A marketplace with no merchants is a directory of nothing, and signing merchants is field sales — people, in cars, in cities. A bundle inherits the partner’s merchant base on day one, which is genuinely the strongest argument for it. A white-label marketplace does not, and an operator who has never run a field sales force routinely underestimates this by a factor that embarrasses everybody.

Operations. Somebody has to run dispatch, handle the order that arrived cold, and answer the merchant whose payout looks wrong. Under a bundle that is the partner’s cost. Under white-label it is yours, and it does not go away at scale — it grows roughly linearly with orders.

The economics of the take rate. Under a bundle you typically receive a share of a partner’s margin on a business you do not control. Under white-label you set the take rate and carry the platform fee. Whether that is better depends entirely on volume, and the crossover is further out than most business cases assume.

How to decide, in three questions

Do you intend to run a commerce business, or to reduce churn? If the honest answer is churn, bundle. The commerce P&L will never justify itself on its own and does not need to.

Can you sign merchants? Not “could we hire someone” — do you have, or can you buy, a field organisation that signs and retains small businesses. If not, a white-label marketplace launches into an empty catalogue and the subscriber experience is worse than the bundle you rejected.

What happens in year four? Write down what you own on the day the partnership ends, or the day you switch platform vendors. If those two answers look the same, the choice matters less than it appears. If they look very different, that difference is the decision.

Where SuperApp sits

SuperApp is a white-label multi-vertical marketplace platform: an operator launches their own branded marketplace across food delivery, retail and supermarkets on one system, with customer, merchant and dispatcher tools included and no commission taken on the operator’s take rate.

That makes it an answer to the second shape and not the first. If you are choosing to bundle, this is not the product, and a comparison page that pretended otherwise would waste your time. If you are choosing to own the marketplace, the questions above are the ones we would expect you to put to us — and the architecture and security post is where the tenant-isolation and subprocessor answers are written down rather than promised.