Keep a messaging rate sheet that can explain an old estimate
Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.
An estimate approved last month now shows a different total, even though nobody changed the planned volume. The rate cell points to a pricing tab that someone updated yesterday. The new figure may be useful for planning, but the spreadsheet can no longer explain the original approval.
The fix is to keep pricing inputs as dated records rather than replaceable cells. Give each rate version a stable ID, preserve the evidence behind it, and save each estimate with the exact rate versions and assumptions used. When prices change, create a revised forecast alongside the approved estimate—not inside it.
This approach separates three questions: what did we reasonably estimate then, what should we forecast now, and what should the delivered activity have cost? Those questions can legitimately require different versions of the same rate.
1. Give each rate both an effective date and a verification date
A single “last updated” column cannot explain a historical estimate. You need to know when a price applies and when your team learned about it.
- Effective period: the activity dates to which the rate applies, according to the pricing evidence.
- Verification date: when someone checked that evidence and recorded the input.
- Recorded version: the immutable record of what your team understood at that time.
These dates need not match. A new rate might be verified before its effective date. A team might also discover a change after it has taken effect. Without both dates, it is hard to distinguish an estimate based on the evidence available at approval from an estimate that overlooked a known update.
Record whether an effective end date is inclusive or exclusive. For a spreadsheet, inclusive start and end dates are often easier to inspect. Use an explicit value such as “open; no end recorded” rather than leaving a blank that could mean missing data.
Do not rewrite the historical record when an open period later closes. Add a successor record that captures the newly known end date, and mark which record it supersedes. The earlier record remains available to explain estimates made before the update.
2. Make the register describe a price, not just a number
A rate without a charging unit, market, category and currency is not a usable pricing input. It is an unexplained number that happens to fit a formula.
For WhatsApp Business Platform budgeting, Meta pricing depends on delivered messages, message category and recipient market. Exact rates and applicable exceptions need checking against the current official pricing information and rate card. Do not assume every planned or delivered message is billable.
A practical register should contain:
- Stable record ID and cost component: distinguish Meta messaging, provider/platform subscription and AI usage.
- Pricing dimensions: recipient market, category, currency, rate and charging unit, plus any applicable condition that changes the price.
- Timing: effective start, effective end and verification date.
- Evidence: source reference, retained copy reference and the relevant section or row.
- Review information: verifier, status and superseded-record reference.
Keep separate cost components on separate rows or in separate registers with compatible IDs. A monthly platform charge is not a per-message charge. An AI allowance or usage assumption should not disappear inside a blended messaging rate.
A source URL helps a reviewer find the publisher. A retained, dated copy helps explain what the team actually read. Store the relevant rate card, contract schedule or pricing notice in a controlled evidence folder, subject to your organisation's retention and access rules. A link to a page that changes is not enough on its own.
3. A filled fictional rate register
Fictional worked example: every market label, document, date, rate and amount below is hypothetical. These are not current prices, vendor terms or Alactic customer results. “Market A” is a placeholder, and the charging units are assumptions for this example only.
The team initially records an open-ended marketing rate. It later receives a hypothetical notice introducing a different rate from 1 April. Instead of editing the first record, it adds a period-closing successor and a new rate record.
REGISTER: hypothetical example Date convention: YYYY-MM-DD; period endpoints inclusive Currency for all example records: USD Record: META-MA-MKT-01 Component: Meta messaging Recipient market: Market A Category: Marketing Rate: USD 0.040 per qualifying billable delivered message Effective period: 2030-01-01 to open; no end recorded Verified: 2030-03-08 Source: EX-D01, fictional rate schedule, marketing row Retained evidence: pricing-evidence/EX-D01-20300308 Verifier: Finance reviewer Status: Historical; retained exactly as known on 2030-03-08 Record: META-MA-MKT-01B Component: Meta messaging Recipient market: Market A Category: Marketing Rate: USD 0.040 per qualifying billable delivered message Effective period: 2030-01-01 to 2030-03-31 Verified: 2030-03-21 Source: EX-D01 plus EX-D02, fictional change notice Retained evidence: pricing-evidence/EX-D02-20300321 Verifier: Finance reviewer Status: Supersedes 01 for period selection after verification Record: META-MA-MKT-02 Component: Meta messaging Recipient market: Market A Category: Marketing Rate: USD 0.050 per qualifying billable delivered message Effective period: 2030-04-01 to open; no end recorded Verified: 2030-03-21 Source: EX-D02, fictional change notice, marketing row Retained evidence: pricing-evidence/EX-D02-20300321 Verifier: Finance reviewer Status: Available for forecasts covering 2030-04-01 onward Record: META-MA-UTL-01 Component: Meta messaging Recipient market: Market A Category: Utility Rate: USD 0.015 per qualifying billable delivered message Effective period: 2030-01-01 to open; no end recorded Verified: 2030-03-08 Source: EX-D01, fictional rate schedule, utility row Retained evidence: pricing-evidence/EX-D01-20300308 Verifier: Finance reviewer Status: No later change recorded in this example
The utility record shows why category belongs in the selection key. A formula that matches only market and currency could silently select the wrong price. Likewise, the old marketing record and its successor must not both appear as eligible current records for the same activity date.
The folder references above are fictional internal evidence identifiers, not external citations. In a working register, they should resolve to documents a reviewer can actually open.
4. Freeze the approved estimate, including its assumptions
A rate register preserves pricing evidence. An estimate snapshot preserves the decision made with that evidence. You need both.
Do not make an approved workbook depend on a live lookup that always retrieves the newest rate. Save the selected record IDs and the actual input values used. Keep the formulas or calculation description as well, so that the stored total can be recalculated rather than merely trusted.
Continuing the hypothetical example, the team approves an April budget on 10 March, before it verifies the price change:
ESTIMATE: EST-APR-01 Status: Approved baseline; do not overwrite Approved: 2030-03-10 Planned activity: 2030-04-01 through 2030-04-30 Pricing evidence cutoff: 2030-03-08 Selected rate record: META-MA-MKT-01 Assumed qualifying billable delivered messages: 12,000 Saved unit rate: USD 0.040 Meta messaging estimate: 12,000 x 0.040 = USD 480 Provider/platform subscription allowance: USD 200 Evidence: EX-P01, fictional subscription schedule AI usage allowance: USD 30 Basis: EX-A01, fictional usage worksheet Approved total: USD 710 Assumptions: Market A marketing activity only; all 12,000 messages in this calculation assumed to qualify for charging; no currency conversion; taxes excluded from this example. Approver: Budget owner Approval evidence: approvals/EST-APR-01
The billable-volume assumption is deliberately explicit. A real estimate should explain how expected delivered activity becomes expected chargeable activity, using verified charging rules and applicable exceptions. Planned sends alone do not establish the billed quantity.
Preserve the market and category mix, delivery assumptions, activity dates, fee inclusions, currency conversion basis if any, and treatment of taxes or adjustments. Otherwise, reproducing the arithmetic may still fail to reproduce the reasoning.
5. Choose the version according to the decision
“Always use the latest rate” is too crude. The newest published rate might apply next month, while the activity being reconciled occurred last month. Conversely, a historical approval rate may no longer be suitable for a forward forecast.
- Explaining an old approval: use its frozen inputs and evidence cutoff. Show later information separately.
- Preparing a current forecast: use verified rates effective for the forecast activity dates, including known future changes.
- Requesting revised approval: create a new estimate version using the updated assumptions. Keep the original approval and link the revision to it.
- Reconciling delivered activity: use the rates and charging conditions applicable to the activity being billed, supported by the relevant pricing evidence and billing records.
For a forecast spanning a price-change date, split the expected volume across the effective periods. Do not apply the newer rate to the whole period just because it is easier. If the activity timing is uncertain, show that allocation as an assumption or scenario.
In the fictional example, the team creates EST-APR-02 on 21 March. It references META-MA-MKT-02 and keeps the same assumed volume:
- Meta messaging: 12,000 × USD 0.050 = USD 600.
- Unchanged subscription and AI allowances: USD 230.
- Revised forecast: USD 830.
- Increase against the approved baseline: USD 120, attributable entirely to the hypothetical rate change.
The original USD 710 remains the approved historical baseline. The USD 830 becomes the current forecast, pending whatever revised approval the organisation requires. Neither number should masquerade as the other.
6. Reconcile without erasing the price-change story
A useful reconciliation separates rate variance from volume variance. Otherwise, a lower-than-forecast invoice can hide a price increase, and a higher invoice can be wrongly blamed on pricing.
Suppose, in the same hypothetical example, the billing evidence supports 11,000 qualifying billable delivered marketing messages in April. Assume the subscription and AI charges equal their allowances and there are no other adjustments.
- Messaging charge: 11,000 × USD 0.050 = USD 550.
- Total reconciled cost: USD 550 + USD 200 + USD 30 = USD 780.
- Variance against the original USD 710 approval: USD 70 higher.
Explain that USD 70 as two movements: a USD 120 rate increase at the original assumed volume, offset by USD 50 less cost from 1,000 fewer qualifying messages at the revised rate. This preserves the link between the original decision, the update and the outcome.
If the invoice does not match, keep the difference visible. Check activity dates, market and category allocation, billed quantities, exceptions, fee lines, rounding and currency treatment before changing the register. An invoice discrepancy is a reason to investigate, not evidence that a convenient replacement rate is correct.
7. Handle missing evidence and conflicting versions explicitly
The register will sometimes be incomplete. The important control is to expose uncertainty instead of converting it into false precision.
A historical rate was overwritten
Look for the approved workbook, retained rate documents, approval attachments and billing evidence. Reconstruct the input only as far as the evidence supports it. Label the result “reconstructed,” record the evidence used and identify unresolved assumptions. Do not backdate a verification or present a remembered rate as an original record.
Two records appear valid for the same activity
Check whether they differ by category, market, charging condition or effective period. If they truly conflict, flag the selection for review rather than silently choosing the cheaper or more recently entered row. A useful spreadsheet check counts eligible matches and flags anything other than one.
A source has changed or disappeared
Use the retained copy to explain the earlier estimate. Obtain fresh evidence for new work. If no reliable historical copy exists, preserve that gap in the record; a current page cannot prove what it said earlier.
There is a trade-off between simplicity and control. One shared spreadsheet may be sufficient for a small rate set, provided editing is restricted and approved snapshots are protected. More markets and pricing conditions increase the value of structured validation and independent review. These are recommended operating controls, not claims about any product's built-in features.
8. Test the next estimate before trusting the process
Start with one recently approved estimate rather than rebuilding every historical spreadsheet. Assign its selected rates stable IDs, retain the source evidence and save a frozen calculation with its assumptions.
Then ask a colleague who did not prepare it to answer four questions:
- Which exact pricing evidence supported the approval?
- Which effective period and charging unit were used?
- Can the approved total be recalculated from the saved inputs?
- Can a current forecast be produced without changing the approved baseline?
Record how many answers required asking the original estimator. For that estimate, aim for all four to be answerable from the saved record alone. If one fails, add the missing evidence, assumption or version link before extending the register to more work.
The practical test is not whether the pricing tab looks tidy. It is whether another person can explain an old number and calculate a new one without confusing the two.
Put the next conversation to work.
Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.
Explore Communications