Blog · How-to guides · Retail banking
Marketing analytics in banking follows campaign responses through account opening, funding, activation, and recorded revenue. Build a comparable cohort table from approved exports, calculate the measures in a spreadsheet, and reconcile the results before changing a campaign.
Marketing analytics in banking follows campaign spend and responses through account opening, funding, activation, and recorded revenue. Compare groups at the same age and reconcile their outcomes to approved campaign, account, and finance exports. The result supports a campaign performance review, with product suitability and individual financial recommendations outside its scope.
A response, an opened account, and an activated account answer different questions. Marketing owns the campaign and response definitions. Account operations owns opening, funding, and activation status. Finance owns recorded revenue and the cost allocation policy.
Agree those definitions before anyone ranks channels. For this example, activation means the qualifying account funded and met the bank's documented activity condition within its first 60 days. The condition is consistent across the campaign cohorts. It is an internal measurement definition, not a universal banking standard.
Keep campaign acquisition separate from broader customer analytics in banking, which examines established customer relationships. An existing relationship can produce a new campaign response without creating a new customer. Label whether the campaign concerns new customers, new accounts for existing customers, or a mixture.
An assigned campaign source is also an accounting convention for the analysis. Set it before reviewing outcomes. If a customer received several messages, the authorized data owner applies the agreed assignment rule internally. Do not count the same qualifying account in several channel totals.
Start from the campaign spend register, response summary, qualifying account report, and finance's cohort revenue export. Have the responsible owners prepare an approved aggregate table so the comparison does not require names, account numbers, addresses, or transaction descriptions.
| Field | Definition to agree |
|---|---|
| Campaign and cohort | Assigned channel and the intake period |
| Eligible recipients | Distinct recipients within mutually exclusive campaign groups |
| Responses | Qualifying responses under a documented rule |
| Opened | Qualifying accounts associated with those responses |
| Funded | Those opened accounts with funding by the cutoff |
| Activated | Funded accounts meeting the agreed activity condition |
| Recorded revenue (USD) | Finance-approved income for the cohort's observation window |
| Campaign spend (USD) | Direct and allocated campaign costs on the agreed basis |
Apply the same campaign-to-response and response-to-opening eligibility windows before comparing the groups, then follow each qualifying account for the same duration after opening. For the worked example, the authorized source report counts at most one qualifying account per response. Other campaign designs may differ, so document the count unit instead of assuming that customers and accounts are interchangeable.
Balances belong in a separate field. The FDIC's Quarterly Banking Profile covers bank earnings as well as loan and deposit activity; those are distinct aspects of bank performance. Your campaign export should preserve that distinction. The existing retail banking revenue analysis explains the wider income question.
An illustrative US regional bank compares mature campaigns for the same account offering. Each group has completed the same 60-day account observation window. The campaigns have mutually exclusive assigned recipient groups, and finance has supplied recognized cohort revenue under one consistent policy.
| Campaign | Spend (USD) | Recipients | Responses | Opened | Funded | Activated | Recorded revenue (USD) |
|---|---|---|---|---|---|---|---|
| Direct mail | 18,000 | 12,000 | 240 | 96 | 72 | 54 | 4,320 |
| 6,000 | 15,000 | 450 | 135 | 108 | 81 | 6,480 | |
| Branch follow-up | 9,000 | 3,000 | 180 | 90 | 81 | 63 | 5,040 |
| Total | 33,000 | 30,000 | 870 | 321 | 261 | 198 | 15,840 |
With that layout in A:H and campaign rows in 2:4, use these Excel formulas for response rate, opening conversion, activation among opened accounts, and campaign cost per activation:
I2 =IFERROR(D2/C2,"")
J2 =IFERROR(E2/D2,"")
K2 =IFERROR(G2/E2,"")
L2 =IFERROR(B2/G2,"")
Fill down. Microsoft's IFERROR documentation explains how the replacement result works when a calculation returns an error. A blank here means investigate the denominator; it does not mean zero performance. Covirage is not affiliated with Microsoft.
| Campaign | Response rate | Opened per response | Activated per opening | Cost per activation (USD) |
|---|---|---|---|---|
| Direct mail | 2.0% | 40.0% | 56.3% | 333.33 |
| 3.0% | 30.0% | 60.0% | 74.07 | |
| Branch follow-up | 6.0% | 50.0% | 70.0% | 142.86 |
| Total | 2.9% | 36.9% | 61.7% | 166.67 |
Rates are rounded to one decimal place and costs to cents. Email's cost per activation is 6,000 / 81 = $74.07. The combined rate is 33,000 / 198 = $166.67, calculated from totals rather than an average of channel costs. Use =SUM(B2:B4)/SUM(G2:G4) for that total.
This campaign cost ratio is narrower than customer acquisition cost. It excludes any acquisition costs outside the campaign policy and counts activated qualifying accounts rather than automatically treating them as new customers.
For each campaign, opened accounts must reconcile to activated accounts, funded but inactive accounts, and unfunded accounts. These categories use the same cutoff and are mutually exclusive.
| Campaign | Activated | Funded but inactive | Unfunded | Reconciled openings |
|---|---|---|---|---|
| Direct mail | 54 | 18 | 24 | 96 |
| 81 | 27 | 27 | 135 | |
| Branch follow-up | 63 | 18 | 9 | 90 |
| Total | 198 | 63 | 60 | 321 |
Funded but inactive is =F2-G2; unfunded is =E2-F2. The combined identity is 198 + 63 + 60 = 321. If the totals disagree, investigate duplicate mappings, excluded accounts, or a different status cutoff before evaluating the campaign.
Finance reconciles $15,840 to the scoped revenue extract for the same accounts and observation dates. That may span accounting periods, so it should not be compared blindly with one month's whole-bank income. Keep fee reversals and any net interest allocation consistent, and prevent the same income from appearing under several campaigns.
Email has the lowest observed campaign cost per activation. Branch follow-up has the highest response and activation rates. Neither statement proves which campaign would produce the most incremental profit after a spending change; audience selection, staffing, incentives, and operating costs still matter.
Bring approved aggregate campaign, account outcome, and recorded revenue exports with the cohort window and definitions. Covirage maps the columns; its tools compute the comparisons and check the totals, while the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own approved export; creating an account takes about a minute.
See Covirage for retail banking for the supporting income analysis context.
Include eligible recipients, responses, qualifying account openings, funded accounts, activated accounts, campaign cost per activation, and recorded cohort revenue. Define each stage before comparing channels. An opened account alone does not show whether the account funded, became active, or generated recognized income.
No. Keep deposit balances separate from recorded income. Finance should define the revenue included in a campaign cohort, including any approved net interest allocation and recognized fees. Do not substitute balances, projected lifetime value, or gross interest receipts for that accounting definition.
Choose a window that gives the relevant accounts time to fund and activate, then apply it consistently. Label younger cohorts as incomplete. A campaign with longer follow-up can appear stronger because more outcomes have been recorded, even when its underlying performance is unchanged.
It shows revenue recorded for the accounts assigned to the campaign under a stated rule. Some customers might have opened accounts without the campaign. Incremental return needs a separate comparison design and relevant costs; a campaign-to-account mapping alone cannot establish it.