Build a Voice Campaign Funnel from Attempt to Verified Outcome

A voice campaign success funnel shows where a campaign makes progress and where it stops. Counting attempts alone measures activity. Counting answered calls adds connection evidence. Neither tells you whether the intended person had a useful conversation or the business objective was completed.

A good funnel makes each transition explicit. It also separates recipients from attempts, because one recipient may receive more than one attempt. Without that distinction, retrying the same audience can make activity look like growth.

Define the stages for your purpose

Start with eligible recipients, then launched recipients, call attempts, connected calls, relevant conversations and verified outcomes. Add a pending stage if a human or external system must finish the work.

Use a dictionary for every stage. “Relevant conversation” might require that the intended participant understood the purpose and engaged with the request. “Verified outcome” should identify the evidence required for the particular task, not merely a positive-sounding summary.

Twilio's Call resource explains that completed connections can include voicemail or IVR. This is why carrier status should remain a connection layer rather than being relabelled as customer success.

Keep recipient and attempt views separate

A hypothetical pilot might contain 40 eligible recipients and 50 attempts because some recipients were retried. If 24 unique recipients connected, report both the attempt count and the unique connected-recipient count.

Do not calculate a recipient conversion rate using attempts as the denominator. Likewise, do not add multiple successful interactions with the same person and describe them as multiple converted customers. Decide whether the business question concerns conversations, people or completed tasks.

Burki's campaign guide provides operational context for audience and run history. The funnel should be built from the actual reviewed run rather than the campaign's original planned size.

Work through a hypothetical funnel

Suppose the 40-recipient pilot yields 24 connected recipients. Of those, 18 have a relevant conversation, 10 complete the intended supported outcome and 4 create a valid follow-up request. The remaining conversations do not reach either result. These are illustrative numbers, not platform performance claims.

The verified-outcome rate is 10 divided by 40, or 25% of eligible recipients. Among relevant conversations it is 10 divided by 18, approximately 56%. Both figures are valid if labelled; they answer different questions.

The four pending follow-ups should not be added to completed outcomes until their required final state is confirmed. Show them as progress with an owner and review window.

Explain losses at each transition

A low connection rate suggests questions about number quality, timing, caller identity or reachability. A strong connection rate with weak relevant engagement suggests the opening, purpose or audience may be wrong. Many useful conversations without completed outcomes suggest workflow friction or dependencies.

These are investigation directions, not automatic causal conclusions. Compare evidence from the same run and inspect representative calls. Avoid blaming a voice model for a problem caused by an outdated audience.

Keep opt-outs, wrong numbers and uncertain outcomes visible as separate categories. Removing inconvenient rows can make the chart cleaner while preventing the business from learning what happened.

Allow a defined completion window for delayed outcomes. A follow-up finished tomorrow should update the original cohort according to a documented rule, not become an unrelated success in whichever report looks best. Retain the original run identity when reviewing these later results.

Add cost after the funnel is stable

Attach actual campaign costs to the defined period and workload. Cost per attempt, per connected recipient and per verified outcome are different measures. Show their denominators and include separately billed components where relevant.

Burki's cost tracking guide can support that reconciliation. Do not use a favourable per-minute price to imply campaign efficiency without outcome evidence.

After each pilot, choose one transition to improve and state the hypothesis before changing the setup. Keep the audience and purpose comparable where possible. The next useful dashboard is a small, well-defined funnel whose counts can be traced back to recipients, calls and verified results.