Inbound operations worry about how fast they answer. Outbound operations worry about something else entirely: of all the numbers they dial, how many turn into a useful conversation, and how many of those conversations turn into the outcome the client is paying for. Every outbound metric is a stage in that funnel, and most reporting confusion comes from people quoting a stage without saying what it was divided by.

Start with one day of numbers

Suppose a credit card cross-sell team ends the day with these counts:

StageCount
Dials (attempts)5,000
Connects (a human answered)1,500
Right-party contacts (the intended customer)1,000
Sales closed120

The standard stage-to-stage rates are:

  • Connect rate = connects ÷ dials = 1,500 ÷ 5,000 = 30%.
  • RPC rate = right-party contacts ÷ connects = 1,000 ÷ 1,500 ≈ 66.7%.
  • Conversion rate = sales ÷ right-party contacts = 120 ÷ 1,000 = 12%.

Now the trap. Someone else might report RPC rate as RPCs ÷ dials, which gives 20%, and conversion as sales ÷ connects (8%) or sales ÷ dials (2.4%). None of these is wrong; they are different questions. But if one team leader reports 12% conversion and another reports 8%, they may be performing identically. Every outbound dashboard should state the denominator next to the percentage, or show the raw counts.

What each stage is really measuring

The point of splitting the funnel is that each stage is controlled by different things, so a drop in each one points to a different fix.

Connect rate is mostly about the data and the dialling strategy. Stale numbers, numbers that have been ported or disconnected, calling at hours when people do not pick up, and caller ID that customers have learned to ignore all push it down. Agents have almost no influence here. Improving it means cleaning the list, testing call windows, and using retry rules that space attempts sensibly instead of hammering the same number five times in an hour.

RPC rate is about whether the number belongs to the person you want. Family phones, office landlines and recycled mobile numbers lead to connects with the wrong person. In collections this stage is heavily affected by skip tracing quality. Agent skill matters a little (a polite request for a better time to reach the customer can turn a wrong-party connect into a scheduled callback) but mostly it is a data problem.

Conversion is where agent skill, the offer and the script live. Once the right person is on the line, the result depends on how relevant the offer is, how well the agent opens and handles objections, and whether the targeting sent this offer to people who could plausibly want it.

This is why ranking agents on sales per dial is unfair. An agent given a worse list connects less often and looks weaker, even if she converts right-party contacts as well as anyone. Compare agents on conversion per RPC, and compare lists and campaigns on connect and RPC rates.

How a small gain upstream travels down

Because the stages multiply, an improvement near the top carries through. If the team keeps the same 5,000 dials but a list refresh lifts connect rate from 30% to 36%, connects rise to 1,800. With RPC rate still at two-thirds, right-party contacts rise to 1,200, and at 12% conversion that is 144 sales instead of 120. That is a 20% rise in sales without any change in agent behaviour.

The same arithmetic gives a useful planning figure: in the base case the team needs about 41.7 dials per sale (5,000 ÷ 120). If a client wants 200 sales a day, the operation needs roughly 8,300 dials’ worth of data and dialling capacity, plus a cushion, because a fresh list never stays fresh for long and the best leads are used up first.

The collections version of the funnel

In collections the bottom of the funnel is not a sale but a payment, and there is an extra stage that sales teams do not have:

  1. Promise to pay (PTP) rate: of right-party contacts, how many committed to pay an amount by a date. If 400 of 1,000 RPCs give a PTP, the PTP rate is 40%.
  2. Kept rate: of those promises, how many were actually honoured. If 260 of the 400 pay, the kept rate is 65%.
  3. Amount recovered and roll rates: how much money came in and how many accounts moved to a better or worse delinquency bucket.

PTP rate on its own is easy to inflate. An agent under pressure can accept vague promises (“I will try next week”) that were never likely to be kept. Watching PTP rate and kept rate together exposes this quickly: a high PTP rate with a falling kept rate usually means promises are being logged rather than earned.

Collections calling in India is also bound by conduct rules. RBI’s instructions on recovery agents restrict calling to between 8 a.m. and 7 p.m. and prohibit intimidation or harassment, and lenders remain responsible for the conduct of agencies they outsource to. Funnel targets must be designed so they never pressure agents toward those lines.

Productivity measures that sit beside the funnel

Rates tell you about quality of contact; you also need volume. Common companions are dials per hour, connects per hour, talk time per logged hour and average handle time on RPC calls. A predictive dialler raises connects per hour dramatically because agents stop waiting through ringing and voicemail, but it also introduces abandoned calls when a customer answers and no agent is free. That trade-off between throughput and nuisance calls is a topic on its own, and regulators in several countries cap the permitted abandon rate for this reason.

The practical habit is simple: whenever you see an outbound percentage, ask “of what?” Then look at each stage separately, and fix the stage that is leaking with the lever that actually controls it.

Written by the RP Strativa Team. The campaign figures in the examples are illustrative.

Leave a Reply

Trending