Compare Chargeflow with your historical results by reviewing recovered dollars, eligible resolved-case rates, unresolved exposure and operating work on explicitly labeled cohorts. Then account for case mix and service costs. The comparison can describe what happened before and after adoption; it cannot establish that the service caused the entire difference.
The practical outcome is a value review that finance and operations can explain together. Start with the merchant’s records and definitions rather than a headline improvement multiplier.
Define the Two Populations Before Reading the Difference
Insights 2.0 uses two different cohort rules:
View
Who is included
What to check
Before Chargeflow
Disputes created before the calendar month in which Chargeflow first managed a dispute
Historical completeness, payment accounts and the earlier process
With Chargeflow
Disputes on which Chargeflow performed an action, such as submission
Which actions and cases are represented, unresolved exposure and eligibility rules
For example, if the first managed dispute was in August, the historical cohort ends with disputes created through July. Do not put every case created after a mid-August activation into the managed group merely because it occurred later.
The global cohort control changes the recovered-revenue and recovery-rate hero cards. Other widgets retain their own or section-level scopes; the managed performance table does not become a historical table just because the hero-card toggle changes. Check each widget’s label before copying values into the review.
Explain the Eligibility Difference
Chargeflow-managed recovery and win rates exclude lost disputes flagged as intentional fraud from the denominator. Before Chargeflow retains those losses. That asymmetry can materially change the apparent gap.
Keep the displayed product result and the comparison qualification together. If you build a supplementary analytical view with harmonized eligibility, label it separately and document the underlying records. Do not silently alter the dashboard’s definition or suggest a harmonized calculation is already a native feature.
Also keep money and cases distinct. A higher dollar recovery rate does not necessarily mean a higher case win rate. The win-rate guide provides the base definitions.
Show the Effect of the Exclusion Before Explaining Uplift
Hold outcomes constant in a hypothetical arithmetic check: $400 won, $100 eligible lost and $500 lost with an intentional-fraud flag. Every amount is resolved and uses USD. This is a demonstration of denominator policy, not an actual historical-versus-managed cohort.
Historical eligibility calculation on the same hypothetical outcomes: $400 ÷ ($400 + $100 + $500) = 40%.
The 40-percentage-point difference in this controlled example comes entirely from eligibility, with no change in wins. Actual historical and managed cohorts also differ in dates and membership. In your review worksheet, record excluded lost count and value beside each rate before attributing a gap to execution.
Use a Comparison Worksheet That Exposes the Differences
Record the following before presenting an improvement statement. A missing field is an investigation item, not a reason to replace it with an assumed value.
Build a historical comparison you can reuse
Enter aggregate USD values and keep the cohort labels with the result. Download the completed review as CSV, or copy it into your own review document. Required fields are marked *.
Historical value recovery rate—
Managed value recovery rate—
Observed rate difference—
Observed rate ratio—
Unresolved value, reported separately—
Keep the comparison qualified. Different periods, populations and eligibility rules can explain a gap. This worksheet does not establish causal uplift, incremental recovery, realized payroll savings or ROI. Blank pending values remain unknown; a zero denominator does not become a 0% rate.
For large differences, inspect the cases that account for the most disputed value. A new processor, a change in fulfillment policy or a small group of large losses can matter more than a blended average.
A Worked Historical Comparison
Hypothetical example: the figures illustrate the arithmetic. They are not customer results, and the periods are not controlled experimental groups.
Measure
Historical cohort
Managed cohort
Won value
$1,200
$3,100
Eligible resolved value
$10,000
$10,000
Value recovery rate
12%
31%
Other material context
Different historical period
Product eligibility exclusions and any changed case mix must be disclosed
Accurate description: the managed cohort’s observed rate was 2.58 times the historical rate, with the stated cohort and eligibility differences.
A claim that Chargeflow caused all 19 points would require more evidence. Likewise, multiplying the gap by another period’s dispute total does not establish incremental recovered dollars. The comparison has not shown what the same disputes would have recovered under the previous process.
Review Value Beyond the Outcome Rate
Inspect submissions, remaining manual work and response timing. These measures can help a new account assess operational activity before enough disputes resolve to support a stable outcome review.
Insights 2.0 Time Saved uses a fixed model of 40 minutes per submitted case at $24 per hour. Label it as an estimate. If finance uses a different handling-time or labor-cost assumption, present that as a separate business-case calculation and include retained manual work.
Keep the financial bridge explicit: won-dispute value, processor credits, service costs and a defined net result are different records. The finance-team framework explains how to reconcile them. Released capacity is useful even when payroll does not fall, but it should not be described as realized cash savings without evidence.
Handle Missing History and Zero Baselines Honestly
If earlier records are absent, show the period and accounts actually available. Do not replace missing outcomes with losses or describe an empty historical panel as a 0% rate. A new merchant can still review coverage, submissions, pending status and documented actions.
If a valid historical rate is zero, a multiplicative rate ratio is undefined. Report the underlying numerator and denominator and, when appropriate, the percentage-point change. If the managed cohort has too little resolved value to interpret, describe the review as preliminary rather than inventing a universal minimum sample.
For recent periods, look at both resolved and pending value. The final rate can move in either direction; an unresolved case is neither a guaranteed win nor an assumed loss.
Finish With a Decision and an Evidence Gap
State the observed result. Include the period, metric and relevant exclusions.
Explain the operating effect. Describe work completed, remaining effort and financial reconciliation.
Name what could change the conclusion. Missing history, unresolved cases or a changed case mix should remain visible.
Choose the next action. Continue the process, investigate a specific loss driver or correct a data problem.
Set the follow-up. Revisit the same definitions after additional outcomes or an operational change.
A useful review does not require every result to be favorable. An unexplained gap should lead to case-level investigation and a clear owner.
Frequently Asked Questions
Does a before-and-after comparison prove Chargeflow caused an improvement?
No. The historical and managed populations can differ in timing, mix and eligibility. The comparison describes observed performance and needs those qualifications.
Can I compare a monthly rate with an all-time rate?
You can display them with clear labels, but they are not the same observation window. Prefer the closest defensible comparison and explain the remaining differences.
What if my account has no historical disputes?
Use coverage, workload, pending status and managed outcomes without a fabricated earlier baseline. An empty historical panel is not evidence of a 0% prior recovery rate.
Should recovery-rate improvement be called ROI?
No. A recovery-rate ratio compares outcome rates. ROI requires a defined investment, costs and a defensible baseline for incremental benefit.
Why might my numbers differ from the processor?
Dates, account scope, status updates, eligibility and USD normalization can explain differences. Trace sample cases before deciding whether a difference is expected or a data issue.
%201.svg)
600+ reviews
.png)
