Treat a three-tier route offer as a price list and you’ll end up subsidizing the customers who cost you the most to serve. Treat it as a way to sort customers by what a failed call costs them, and the tiers start paying for themselves.
Reselling voice termination in tiers means buying routes at different quality levels from an upstream carrier and packaging them as distinct products for your own customers, usually premium, standard and economy. Each tier carries a different promise about caller ID, route stability and path directness, and a different price. The skill isn’t finding the cheapest route. It’s matching each customer’s traffic to the tier whose failure mode they can live with.
This post works from first principles, because most tiered offers I’ve seen were built backwards: someone took three buy rates, added a markup, and called it a product.
A tier is a promise about behavior, not a price point
The useful definition of a tier is the set of things a customer can expect when a call goes out on it. Price follows from that, not the other way round.
A premium tier typically means CLI delivered unchanged, direct or short paths to the terminating operator, and stable routing that doesn’t shift with load. A standard tier might still deliver CLI on most destinations but tolerate more transit hops. An economy tier makes few promises about CLI and optimizes for cost. Those definitions vary by upstream carrier, which is exactly why you need to write your own.
Once you think of tiers this way, one consequence is obvious. Two routes at the same buy rate can belong in different tiers, and two routes with buy rates far apart can belong in the same one. The tier is defined by behavior. If the cheapest route happens to deliver clean CLI to a destination, it can sit in premium. Lucky you.
The organizing idea: what does a failed call cost this customer?
Every customer segment has a different cost of failure, and that cost is what they’re paying to protect against.
A contact center running outbound collections loses agent time on every call that doesn’t get answered, and answer rates depend heavily on the displayed number. For them, a missing CLI isn’t a quality blip. It’s idle headcount. Our post on CLI and non-CLI routes explains the mechanism.
A fintech sending verification calls loses a user at the login screen. The failure is rare but expensive, and it lands on their brand.
An SMB on a UCaaS plan makes a few international calls a week. A dropped call is annoying and gets redialed. They’ll accept most things except a surprising bill.
A consumer calling app selling cheap international minutes has customers who are price-first and tolerant, as long as the call connects and the audio holds.
Put those side by side and the tiers almost design themselves. Premium for the first two. Standard for the third. Economy for the fourth, with clear disclosure. The mistake is letting the customer’s budget decide the tier instead of their failure cost.
Pricing: why percentage markup quietly rewards the wrong traffic
The default reseller pricing model is a flat percentage markup on the buy rate. It’s easy to explain and it’s usually wrong for tiered products.
Here’s an illustrative example for one destination. The buy rates are invented; the structure is what matters.
| Tier | Buy rate per min | 25% markup sell rate | Gross margin per min |
|---|---|---|---|
| Premium | $0.0200 | $0.0250 | $0.0050 |
| Standard | $0.0120 | $0.0150 | $0.0030 |
| Economy | $0.0070 | $0.00875 | $0.00175 |
On paper, every tier earns 25%. In practice, the margins aren’t comparable, because the cost of serving a minute isn’t the same across tiers.
Economy traffic generates more trouble tickets: CLI complaints, route changes, the odd audio issue, the customer who bought economy and is now upset it behaves like economy. Someone on your team investigates each one. Suppose, again illustratively, that support and operations cost you $0.0003 per minute on premium, $0.0005 on standard and $0.0010 on economy.
| Tier | Gross margin | Support cost | Net margin per min |
|---|---|---|---|
| Premium | $0.0050 | $0.0003 | $0.0047 |
| Standard | $0.0030 | $0.0005 | $0.0025 |
| Economy | $0.00175 | $0.0010 | $0.00075 |
Economy now earns about a sixth of premium per minute. And because it’s the cheapest tier, percentage pricing makes it the most attractive one for customers to move toward. You’ve built a pricing model that pulls volume into your least profitable product.
The alternative: price each tier on its own economics
Better models set a margin floor per tier in absolute terms, then adjust. One simple approach is a fixed cents-per-minute margin that covers the tier’s support cost plus a target contribution. Another is to widen the percentage on economy and narrow it on premium, which also shrinks the visible price gap between tiers and makes premium easier to sell.
There’s a strategic reason to keep that gap modest. The customers who most need premium are often the ones comparing on price. If premium looks three times the cost of economy, they’ll gamble on economy, have a bad month, and churn. If it looks 40% more, they’ll pay for the protection.
Don’t forget increments and connection fees when you model any of this. A tier with a better rate but coarser billing may not be better at all for short-call customers; the billing increments breakdown shows how much rounding can move the effective price.
Tier leakage: the failure that eats resellers
Tier leakage is what happens when traffic lands on a tier that doesn’t match its failure cost. It’s the most common way a tiered offer loses money.
It happens in both directions. The obvious one is a customer buying economy for traffic that needs CLI, then filing tickets, disputing invoices and eventually leaving. The less obvious one is overflow: premium traffic spilling onto a cheaper route when the premium route is congested, because someone configured least-cost routing across tiers instead of within them. The customer paid for premium and got economy for part of the day, without being told.
A few controls prevent most leakage:
Route within tiers, never across them silently. Your LCR logic should pick the best route inside the tier the traffic belongs to. Overflow from premium should go to another premium route, or fail and alert, not drop down a tier.
Separate tiers at the signaling level. Many wholesale operators use tech prefixes (a short code prepended to the dialed number) or separate trunks per tier. Either way, the tier travels with the call, so you can route, bill and report on it independently.
Tag traffic by use case, not just by customer. One customer may have verification calls that need premium and bulk reminder calls that are fine on standard. Let them send both, tagged, instead of forcing one tier per account.
Report quality per tier. Track answer rates and call durations per tier per destination. If economy suddenly looks like premium, or premium looks like economy, you want to know before the customer does. Our guide to reading ASR, ACD and PDD covers the patterns worth alerting on.
Write tier definitions customers can hold you to
A tier your customer can’t understand is a tier they’ll dispute. Write the definitions down, and make them about behavior you can verify.
Good tier definitions describe CLI handling, routing approach (direct or via transit), how changes are communicated, and what the customer should use the tier for. They avoid promises you don’t control. If your upstream carrier doesn’t commit to a delivery or quality level, you can’t pass one on. Describe what the tier is built for and how you monitor it, not a performance guarantee you’d be underwriting yourself.
Include a “not suitable for” line. “Economy is not suitable for verification calls or outbound campaigns where the displayed number affects answer rates.” It sounds like underselling. It’s the sentence that saves you the most support time.
Where the upstream carrier fits
A tiered offer is only as clean as the routes underneath it, and you need an upstream supply that already separates route quality in a way you can map to your own tiers.
That’s the practical reason many resellers buy from carriers that tier their own routes. IDT Express, for instance, runs voice termination in 160+ countries across three quality tiers, with 1,000+ CLI-certified routes, so a reseller can map premium, standard and economy onto supply that’s already differentiated rather than trying to infer quality from price. Whoever your upstream is, ask how they define each tier and how they test it, then decide whether their definitions fit yours or need adjusting.
For the mechanics of how SIP termination works in a reseller setup (interconnects, codecs, authentication), our explainer on SIP termination for resellers and telecom businesses is the right starting point, and our guide to the mechanics of voice termination ties the whole cluster together.
A short build order
If you’re building or rebuilding a tiered offer, this sequence tends to work:
- Segment your current customers by failure cost, not revenue.
- Write tier definitions around behavior, including a “not suitable for” line each.
- Map upstream routes into tiers per destination, based on testing rather than buy rate.
- Model net margin per tier, including support cost and increments.
- Set prices that keep the premium gap small enough to sell.
- Configure routing so tiers are separated at the signaling level and never overflow downward silently.
- Report quality per tier monthly and review tier assignments per customer each quarter.
The last step matters more than it looks. Customers’ traffic changes. The SMB that made five calls a week launches an outbound campaign. Your tiers should follow what the traffic needs, not what the customer bought two years ago.
Frequently asked questions
How do voice termination resellers make money?
Voice termination resellers buy routes from upstream carriers at wholesale rates and sell minutes to their own customers at a higher rate, keeping the difference as margin. Profitability depends on net margin per minute after support and operations costs, not just the gap between buy and sell rates.
What is the difference between premium and standard voice routes?
Premium routes typically deliver caller ID unchanged and use direct or short paths to the terminating operator, while standard routes may use more transit hops and make fewer guarantees. The exact definitions vary by carrier, so resellers should write their own based on tested route behavior.
Should resellers use percentage markup or fixed margin pricing?
Fixed or tier-specific margins usually work better than a flat percentage markup. Percentage markup gives economy traffic the smallest absolute margin while it often carries the highest support cost, which pulls volume toward the least profitable tier.
What is tier leakage in voice termination?
Tier leakage is traffic ending up on a quality tier that doesn’t match its needs, either because a customer bought a cheaper tier than their traffic requires or because routing overflows premium calls onto cheaper routes. It’s prevented by routing within tiers, separating tiers at the signaling level, and reporting quality per tier.
How do tech prefixes work in wholesale voice?
A tech prefix is a short code added to the front of a dialed number that tells the carrier which route class or product to use. It lets one trunk carry several tiers while keeping routing, billing and reporting separate for each.
Talk to our wholesale team
Building a tiered voice offer for your customers? Talk to the IDT Express wholesale team about mapping your tiers to our route supply.
