- A $55 buyer at Priority 1 routinely loses to a $42 buyer at Priority 3 once you multiply payout by billable conversion rate and remaining capacity.
- Rank Ringba Target Priority by expected value per call: EV/call = payout × 7-day billable conversion rate × (1 − capacity fill %), not sticker payout.
- In high-repeat verticals like Medicare AEP and ACA OEP, duplicate-call dedupe windows can suppress payouts on the second dial unless you cascade the duplicate to a buyer with a non-overlapping window.
- Set concurrency caps at multiple layers (Target and Buyer at minimum) so a single buyer running four Targets doesn’t get hammered with 4x the calls they can answer.
- Re-rank tiers weekly off the Ringba Call Details report on a 7-day rolling window, with a 75–100 call minimum before touching Priority.
Most pay-per-call media buyers rank Ringba buyer payout tiers by sticker price, then wonder why their revenue per call (RPC) plateaus while buyers complain about quality. The right Ringba call routing optimization for buyer payout tiers ranks Targets by expected value per call: payout times the rate a buyer actually pays, times how much room that buyer still has in their daily cap. Rank on that, not the bid, and the leaks you’ve been blaming on traffic quality start to close.
Switching from payout-ranked to EV/call-ranked Target Priority is one of the highest-leverage changes a media buyer can make. The biggest gains tend to show up where one high-payout buyer was monopolizing Priority 1 and capping out by lunch. This article walks through four leaks worth closing: payout-ranked Priority, single-layer concurrency caps, duplicate-call dedupe mismatches, and missing partial-pay tiers.
Sticker Payout Is a Vanity Metric. Ringba Rewards Whoever Actually Converts.
The buyer with the highest bid is rarely the buyer with the highest revenue per routed call. Payout is a ceiling. EV/call is what you actually collect.
Here’s the trap. Buyer A pays $55 but only marks 22% of routed calls as billable, and they’re already 80% full for the day. Buyer B pays $42, marks 48% billable, and is 30% full. On paper, A wins. In practice, B is worth almost six times more per call you route.
This section sets up four fixes. The next four sections deliver them.
EV/Call Is the Only Ranking That Survives Contact With a Real Account
Use this formula on every Target in your routing plan:
EV/call = Buyer Payout × 7-day Billable Conversion Rate × (1 − Current Capacity Fill %)
Three inputs, all pulled from Ringba:
- Buyer Payout comes from the Target’s payout settings.
- Billable Conversion Rate comes from billable calls divided by routed calls in the Ringba Call Details report, filtered to a 7-day rolling window.
- Capacity Fill % is calls routed to that Target this period divided by the Target’s configured cap.
A worked example where $42 outranks $55
| Buyer | Payout | 7-day Billable Conv. | Capacity Fill | EV/Call |
|---|---|---|---|---|
| A | $55 | 22% | 80% | $2.42 |
| B | $42 | 48% | 30% | $14.11 |
| C | $38 | 35% | 50% | $6.65 |
The ranking that gets you paid is B, then C, then A. The ranking most operators run is A, then B, then C, because that’s the order of the bids. That gap is the leak.
Pulling the inputs from Ringba’s Call Details report
Open Call Logs in Ringba. Filter to the last 7 days. Group by Target. Pull routed calls and billable calls per Target. Divide. That’s your billable conversion rate. Pair it with each Target’s payout and current fill %, and you have everything you need.
Target Priority, Weight, Capacity, and Concurrency Do Four Different Jobs
Most operators confuse two of these four. The rest of the playbook depends on getting them straight.
- Target Priority is the tiebreaker. Higher priority wins when multiple Targets are eligible.
- Weight is the splitter inside a priority tier. Two Targets at Priority 1 with weights 70/30 split traffic that way.
- Capacity is the volume cap. Hourly, daily, monthly call counts or payout totals.
- Concurrency is the simultaneous-call cap. How many live calls a Target can hold at once.
Ringba also offers Predictive routing, which computes an internal revenue-per-call estimate and applies priority bonuses (for new or underperforming targets, for example) on top of your manual settings. Per Ringba’s Predictive Routing documentation, those bonuses sit alongside the values you configure rather than replacing them. The operator move is to override sticker-payout thinking with your own EV/call ranking at the Priority level, then let any predictive logic do fine-tuning underneath.
Layer concurrency caps where Ringba lets you set them
Ringba’s Target settings expose concurrency capping on the Target, with related capacity controls available at the Buyer and Campaign level. Most operators set concurrency at the Target only. That’s a problem when four Targets share a single Buyer with a real call-center capacity ceiling. The Buyer rejects overflow calls and Ringba marks them as routed-but-failed.
The fix is to layer caps where the platform allows them. Set Target-level concurrency for fairness across Targets in the same group. Set Buyer-level concurrency to match the buyer’s actual contracted simultaneous-call limit. Use campaign-level capacity to protect the vertical from running past your team’s QA capacity. Multiple nets, not one.
When to flatten with Weight instead of stacking with Priority
If two Targets have EV/call within ~10% of each other, don’t stack them in Priority. Split them with Weight. A 60/40 weighted split inside one priority tier keeps both buyers fed, prevents one from starving, and gives you cleaner disposition samples to compare next week.
Use strict Priority when EV/call gaps are wide (say, 30% or more). Use Weight when they’re close enough that you want parallel data.
Duplicate Call Settings Are Where Repeat-Caller Revenue Quietly Dies
This is the leak almost nobody talks about. In high-repeat-caller verticals (Medicare AEP, ACA OEP, debt, final expense), duplicate call settings can silently suppress payouts you already earned.
Here’s the mechanic. Buyer A has a 30-day dedupe window. A caller dials your number on day 3, gets routed to Buyer A, and gets paid. The same caller dials again on day 18. Ringba sees the repeat and, by default, either suppresses the second payout or sends the call back to Buyer A unpaid. Either way, you don’t get paid for a call you bought twice.
Cascading repeat callers to a secondary buyer
The fix is a routing-plan branch for duplicates. Configure the plan so a caller flagged as duplicate to Buyer A cascades to Buyer B: a buyer with a shorter dedupe window, a non-overlapping window, or no overlap with Buyer A’s exclusion period at all.
This takes one rule in the routing plan and it turns suppressed traffic into billable traffic. In Medicare and ACA accounts during the AEP and OEP windows, where repeat dial rates run hot, this is one of the highest-leverage single settings in the account.
Auditing your current dedupe settings in 15 minutes
Open each Buyer in Ringba. Note the dedupe window setting and the duplicate-call action. Build a small grid: Buyer, window length, action on duplicate. Any window over 14 days in a high-repeat vertical is a candidate for cascade routing. Any duplicate action set to “suppress” without a fallback Target is leaking money.
Capacity Overflow and Sub-Threshold Calls Need Their Own Tiers, Not a Hangup
Two fallback architectures are doing real work in mature accounts. Both are missing from most.
Design the Priority 2/3 fallback by EV, not by bid
When Priority 1 hits capacity at 11am, where does the spill go? Most plans default to whichever Target had the next-highest bid. That’s the same sticker-payout mistake at the fallback level.
Rank your fallback tiers the same way you rank Priority 1: by EV/call. The Target that catches overflow should be the one with the next-best expected revenue per call, not the next-best headline payout. Otherwise you’re routing the back half of your day into your worst-performing buyer at the highest price.
Partial-pay tiers: monetizing the 30 to 60 second call
Most top buyers require 60 to 120 seconds of call duration before they pay, depending on vertical. Calls that connect but drop at 35 seconds get nothing. Building a partial-pay tier means routing those sub-threshold calls to a buyer that accepts shorter calls at a lower payout.
You’re not chasing huge dollars per call. You’re refusing to throw away calls you’ve already paid to generate.
Schedule overlays for time zones and surge windows
Buyer hours of operation rarely match your traffic pattern. East Coast buyers reset their caps at midnight ET while your West Coast publishers are still sending volume at 9pm PT. Medicare buyers staff up for AEP surge hours but go thin at lunch.
Use schedule-based routing to overlay buyer hours on top of your EV ranking. A buyer with high EV/call but limited staffed hours should only sit at Priority 1 during those hours, and a different Target should take Priority 1 the rest of the day.
Reshuffle Tiers Weekly Off Disposition Data, Not Monthly Off Gut Feel
Routing decay is real. Buyer conversion rates drift. Capacity changes. New buyers join. A routing plan you set in January is leaking by March if you haven’t touched it.
The weekly review: which reports, which fields, which window
Every Monday, pull the Ringba Call Details report for the prior 7 days. Group by Target. For each Target, compute:
- Routed calls
- Billable calls
- Billable conversion rate (2 ÷ 1)
- Average payout
- Current capacity fill %
- EV/call (payout × billable conv. rate × (1 − fill %))
Re-rank Target Priority to match the new EV/call order. Adjust Weights inside priority tiers where EV/call is close. Touch nothing else.
The 75 to 100 call minimum before you touch Priority
Small samples lie. A new buyer with 18 routed calls and a 60% billable rate looks amazing, until call 50 brings them back to 35%. Don’t change Priority on any Target with fewer than 75 to 100 routed calls in the window. This is a working rule of thumb, not a Ringba-published standard, but it keeps you out of small-sample whiplash.
New buyers go in a probationary tier with a capped daily volume. They earn their EV ranking after they’ve seen enough calls to produce a stable rate. Until then, route to them on Weight, not Priority.
When EV/call rankings lie: sample bias and day-parting
A buyer that converts at 45% in the morning and 18% in the evening doesn’t need a Priority change. They need a schedule overlay. Pull conversion by hour-of-day before you reshuffle. If the swing is wide, fix it with hours of operation, not Priority.
Dispositions also lag. Buyers sometimes dispute or update call status 24 to 72 hours after the call. Anchor your weekly review to the same day of the week every week so the disposition data has settled consistently.
FAQ
How do you calculate expected value per call in Ringba?
EV/call equals buyer payout times 7-day billable conversion rate times (1 − current capacity fill %). Payout comes from the Target settings, conversion rate from the Call Details report filtered to a 7-day rolling window, and fill % from the Target’s cap and current volume. Recompute weekly per Target and re-rank Priority accordingly.
What’s the difference between Target Priority, Weight, Capacity, and Concurrency in Ringba?
Priority is the tiebreaker between eligible Targets. Weight splits traffic inside a single priority tier. Capacity caps total volume (hourly, daily, monthly). Concurrency caps simultaneous live calls. Set Priority by EV/call, Weight when two Targets are within ~10% EV, and concurrency at every layer Ringba exposes it (Target and Buyer at minimum).
How often should I reshuffle Ringba buyer tiers?
Weekly, on a fixed day of the week, off the Call Details report’s 7-day rolling window. Monthly is too slow, because buyer conversion rates drift and capacity changes faster than that. Daily is too noisy and pulls you into small-sample decisions.
Why is my Priority 1 buyer hitting cap by 11am every day?
Because Priority 1 is concentrating traffic into a single Target that doesn’t have enough capacity to absorb it. Either raise the Target’s capacity (if the buyer allows), or split Priority 1 across two Targets with Weight, or move the buyer to Priority 2 during peak hours via a schedule overlay. The right fix depends on whether the buyer wants more volume or you’re saturating their actual capacity.
How do duplicate call settings in Ringba suppress revenue?
When a buyer has a dedupe window (say, 30 days) and a caller dials in twice within that window, Ringba’s default behavior either suppresses the second payout or returns the call unpaid. In high-repeat verticals like Medicare AEP, ACA OEP, and debt, that quietly kills otherwise-billable revenue. Fix it by configuring the routing plan to cascade duplicates to a secondary buyer with a non-overlapping dedupe window.
How small a sample is too small before I change Priority on a Target?
Fewer than roughly 75 to 100 routed calls in the 7-day window is too small. Conversion rates on small samples swing widely, and changing Priority on noisy data creates more churn than it fixes. Park new buyers in a probationary tier with a capped daily volume and route to them on Weight until they hit the threshold.
Should concurrency live at the Target level or the Buyer level?
Both. Target-level concurrency keeps individual Targets from monopolizing live capacity inside a group. Buyer-level concurrency matches the buyer’s contracted simultaneous-call limit so you don’t trigger buyer-side overflow rejections. Single-layer caps are the most common concurrency mistake in audits.
Want a second set of eyes on your Ringba routing plan? Book a call with Elevarus and we’ll walk through your Call Details report, compute EV/call across your current Target stack, surface the dedupe and concurrency leaks, and hand back a re-ranked routing plan you can ship the same week.
