Sales pacing, and why the monthly forecast review is three weeks too late

Every lender has a monthly sales number and a monthly forecast. Most compare them at the end of the month, in a meeting, where the only available action is an explanation.

The useful version compares them every day, pro-rated, and separates the three things that can cause a gap. Done properly it turns a month-end post-mortem into a week-two decision.

This is the third in the KPI series, after first payment default and the recovery funnel.

Pro-rata is harder than it looks#

The naive version divides the monthly target by the number of days and compares. It is wrong in three predictable ways, and each one produces a false alarm in the first week.

Working days, not calendar days. Lending volume follows the working week. A month starting on a Saturday will look catastrophically behind on day three and fine by day six. Pro-rate against working days elapsed.

Payday shape. Consumer credit demand is not flat within a month. In most markets it clusters around payday and around the end of the month. If your pro-rata line is straight and your demand is not, you will be red for the first ten days of every month and green at the end, and people will learn to ignore the report. Weight the curve using your own history, per market.

Which loans count. Refinance and top-up volume flatter the number without representing new demand. Report new issuance against a new-issuance target, and top-ups separately. If commercial targets are set on a blended figure, that is a target-setting problem worth fixing before it becomes a reporting one.

Get those three right and the pacing line becomes trustworthy in the first week, which is the only week where it is worth anything.

Three causes of a gap, and they need different people#

A shortfall against pro-rata is always one or more of:

Demand. Fewer applications arrived. This belongs to marketing, and it is visible immediately at the top of the funnel.

Acceptance. The same applications arrived and fewer were approved. This belongs to risk, and it is usually a rule or score change, sometimes a bureau outage.

Conversion. The same approvals happened and fewer drew down. This belongs to product, and it is usually a broken step, a slow bank connection, or an offer that got less attractive.

A report that shows a gap without splitting it is a report that starts an argument. A report that splits it sends one team to one problem.

The value of daily pacing is not knowing sooner that you are behind. It is knowing sooner which of three teams can do something about it.

Define it before you measure it#

• Issued means paid out, not approved and not signed. Pick the event and use it everywhere.
• Currency at a fixed rate for the reporting period, published on the report. Otherwise FX movement shows up as commercial performance.
• The forecast is versioned. When commercial revises a target mid-month, both the old and new versions stay in the model with valid-from dates. Otherwise your historical variance is unreconstructable.
• Suppress the first two working days. The variance is too wide to mean anything and it burns credibility.

The dimensional model underneath#

• fct_loan at issue grain, carrying issue timestamp, amount, currency, FX rate applied, product, market and whether it is new or top-up
• fct_application at attempt grain, so demand and acceptance are countable independently
• dim_forecast, versioned, at market-month grain with the intra-month weighting curve attached
• dim_date carrying working-day flags per market, because the calendar differs by country
• Metrics for issued_eur, pro_rata_target, acceptance_rate, drawdown_rate

The versioned forecast dimension is the part that gets skipped and the part that makes the whole thing auditable.

The agent#

SignalProjected month-end for one market lands at 63% of target, outside the tolerance band, on working day 9.

ReadsReads fct_loan and fct_application at market-day grain against dim_forecast (version in force on the 1st), with dim_date working-day weighting applied.

Holds fireWorking day 9 of 21. Sample is wide enough for a projection; the agent would have stayed silent on day 2.

CheckIs demand down? Applications 4% below the four-week average — small, not the story.

CheckIs acceptance down? Yes — acceptance rate fell from 4.9% to 4.0% on the 6th.

CheckDid a rule change on the 6th? Yes: a bureau payment-default rule was tightened.

CheckIs conversion down? Drawdown rate flat at 71%. Not the story either.

VerdictRoughly 80% of the projected gap is the acceptance change on the 6th, not demand and not conversion. At current pacing the market closes about 37% short.

ActionBrief to the commercial and risk leads together, with the rule, the date, the projected shortfall and the approvals it represents. Suggests a joint call rather than proposing a change on its own.

Notice the last line. The agent does not recommend loosening a credit rule, because that is a decision with a risk appetite behind it and it is not the agent's to make. It puts the two people who own the trade-off in the same conversation with the same numbers. That boundary — where the agent stops — matters more than anything it does before it.

What good looks like#

• A pacing view that is trustworthy from working day three, weighted to your own demand curve
• New and top-up reported separately, against separate targets
• Every gap split into demand, acceptance and conversion automatically
• Forecast versions retained, so variance can be reconstructed months later
• Month-end projection with a stated confidence range rather than a point estimate

The test is simple. If your commercial team finds out about a miss in the last week of the month, your pacing report is decoration.

Next in this series: acquisition source quality, and the gap between what a channel costs and what it is worth.