Avadhoot Gadgil, Specialist in resilient, petabyte-scale cloud environments, delivering infrastructure innovation across North America, Australia and Europe
TL;DR: Object based storage is the standard way to store unstructured data at scale, but most hyperscaler pricing hides its real cost behind egress fees that only show up when a partner tries to move or serve that data back out. For resellers, that’s not a line-item surprise, it’s a margin problem baked into someone else’s billing model. We break down how object storage pricing actually works, where tiering strategies quietly fail, and what ThinkOn’s channel-only, no-egress-fee model changes for partner economics.
Nobody signs a reseller agreement because they’re excited about object storage. They sign it because a client needs somewhere to put data that isn’t hot enough for primary storage but still has to be retrievable — and because whatever storage line they resell has to leave room for margin underneath it.
The problem is that most object storage pricing is built to make the sticker price look good and the exit look expensive. Ingress is free or cheap almost everywhere. Egress is where the model changes shape: the moment a partner’s client needs to retrieve, migrate, or replicate data out of a hyperscaler’s object store, the bill includes a fee that was never part of the original conversation. Multiply that across every client doing a DR test, a cloud migration, or a routine audit, and the “savings” a partner sold on day one erode every time a client actually uses the data they’re paying to store.
That’s not a pricing detail — it’s a structural conflict. A reseller’s margin depends on predictable, quotable costs. A hyperscaler’s egress model depends on data staying put. Those two incentives don’t line up, and partners are the ones absorbing the gap when a client calls asking why last month’s bill doesn’t match the quote.
Glossary
Object storage — A storage architecture that manages data as discrete objects (data plus metadata plus a unique identifier) rather than as files in a directory tree or blocks on a disk, built for scale and unstructured data.
Object storage tiers — Pricing/performance levels within an object storage platform (e.g., hot, cool, cold, archive) that trade retrieval speed for lower per-gigabyte storage cost.
Egress fees — Charges billed when data is moved or transferred out of a storage platform, typically priced per gigabyte and separate from the storage cost itself.
Ingress fees — Charges (where they exist) for moving data into a storage platform; most providers price ingress at or near zero to encourage adoption.
Hyperscaler — A large-scale, self-service cloud infrastructure provider offering compute and storage at global scale.
Channel-only model — A rare go-to-market structure in which a vendor sells exclusively through partners (MSPs, VARs, resellers) rather than direct to end customers, so the vendor never competes with its own channel for the same account.
Reseller margin — The difference between what a partner pays a vendor for a service and what the partner charges its own client; compressed by any cost the partner can’t predict or pass through cleanly, such as variable egress billing.
What is object based storage, and why does it matter for reseller margins?
Object based storage manages unstructured data as objects, and its real cost to resellers is set by egress pricing, not the storage rate.
Object based storage is the default architecture for unstructured data — backups, media, logs, archives — because it scales without the directory-structure limits of traditional file storage. For end users, the pitch is simple: cheap, scalable, durable storage. For partners reselling that storage, the pitch only holds if the total cost of ownership is predictable, and predictability is exactly what variable egress pricing breaks. A quoted per-gigabyte storage rate means very little to a reseller’s margin if the platform’s real cost structure only becomes visible when a client tries to use their own data.
How does object storage pricing actually work?
Storage cost is the visible number; egress and API request fees are the ones that erode margin later.
Most object storage pricing has three components: a per-gigabyte storage rate (tiered by how “hot” or “cold” the data is), a per-request charge for reading or listing objects, and a per-gigabyte egress charge for data leaving the platform. Providers compete hard on the first number because it’s the one prospects compare. The second and third numbers are where actual spend concentrates over the life of a contract — and where a reseller’s quoted margin gets rewritten by someone else’s billing model.
What’s different about cloud based object storage versus traditional file storage?
Cloud based object storage scales flatly and cheaply for unstructured data; file storage is built for structured, frequently accessed workloads and doesn’t scale the same way.
Traditional file storage organizes data hierarchically and is optimized for fast, frequent access — think active project files or databases. Cloud based object storage is built for volume: petabytes of backups, media, and archival data that don’t need a folder structure, just durable, addressable storage. For partners building backup, archive, or DR offerings, object storage is usually the right architecture — the open question is which provider’s pricing model that architecture sits on top of.
| Feature | Hyperscaler (Hot/Cool) | ThinkOn Warm Object Storage |
| Egress Fees | Variable (The “Cloud Tax”) | $0 or Predictable Flat Rate |
| API Call Costs | Per Request (Get/Put) | Included/Predictable |
| Data Access | Instant to Delayed (4–12 hrs) | Instantaneous |
Why does object storage tiering fail to control costs?
Tiering lowers the storage rate but does nothing to fix egress pricing, which is where the uncontrolled cost actually lives.
Tiering strategies move colder data into cheaper storage classes to reduce the per-gigabyte rate — a sound idea on paper. But tiering doesn’t touch the egress side of the bill. A client who tiers their backups into “archive” storage to save money can still get hit with a disproportionate retrieval and egress charge the one time they actually need that data back, during a DR event or a migration. Tiering optimizes the number partners see monthly; it doesn’t fix the number that shows up when a client actually needs their data.
What should MSPs look for in object storage solutions?
The advertised per-gigabyte rate is the least useful number for evaluating an object storage solutions provider as a partner. What matters more: whether egress and ingress are flat-rate or zero, whether the vendor sells direct (competing with the partner for the same client) or is channel-only, and whether the pricing model is stable enough to quote a client once and not revisit every time they touch their own data.
Find out what egress is really costing your margins.
ThinkOn is channel-only with no direct sales, no fees, ingress, egress or otherwise, and margin you can quote once and keep.
Send us a typical client’s storage profile, and we’ll show you what it costs on ThinkOn next to what you pay today. You’ll get a quotable, flat price with no egress or API fees and no direct competition from us. It’s a 30-minute conversation, and you’ll leave with real numbers.
Talk to our partner team or drop us an email today.
