September 2026. This page is confidential.
Where September lands: what has actually happened so far, plus what we expect for the rest of the month, measured against the 7+5 forecast.
Click + on any line to break it down, or i for a plain-English explanation of what the number means.
| MTD (A) | BOM (B) | Est. total | 7+5 forecast | Variance | % |
|---|
Every line answers the same question: where will September land, and is that better or worse than what we said it would be in the 7+5 forecast?
MTD (A) is what has actually happened between 1 and 22 September (month to date) — real hours, real signed deals, real bank balances. BOM (B) is the balance of the month — what we expect between now and the 30th, built from work already in people's calendars and deals already in the sales system. Est. total adds the two together. 7+5 forecast is the target: the forecast made after seven months of actual results and five months of estimates. Variance and % compare the two.
The coloured dot or pill shows only how far a line is from its forecast — green at or better, amber within 10%, red beyond 10%. It shows distance from forecast only, not whether the result is good or bad. Gross margin is green this month because costs fell at the same rate as revenue.
Brackets like ($104k) mean below forecast. Every dollar figure is in thousands, so $1,983k means $1,983,000.
This is a management view, not a closed set of books. Nothing here has been through accounting cut-off, and every line is stated on the same basis as the forecast it is measured against, so the comparison is like for like. The rest of this page is the technical detail behind each number.
The 7+5 forecast is the September column of the monthly KPI pack (August 2026, dated 17 September) for hours, bookings, leads and cash, and the September column of the 7+5 forecast workbook for revenue and cost of sales. It is fixed and does not move with a refresh. It is not the 2026 plan, which carries 30,872 September hours against the forecast's 21,627.
HubSpot — live, pulled 23 September. SALs are new-business pipeline deals counted on their create date; deals later disqualified or lost still count, because they were accepted. The HubSpot count ties to the KPI pack within three leads a month (August 95 against 95, July 90 against 93, June 136 against 138). Bookings month to date are closed-won deals with a September close date. Balance of month is every open deal with a September close date, weighted deal by deal. New business, existing business and referral are split by HubSpot pipeline, which reproduces the month-to-date split exactly.
Productive — live, pulled 23 September. Billable time on client projects, non-billable services excluded, grouped on each person's entity. Domaine NA here means North America plus Studio, CRM and B2B; the forecast lists those separately and they are folded in to give a three-brand comparator. Hours tie to the KPI pack within 0.6% for August (22,367 against 22,236).
Rippling — live, pulled 23 September. Employees only, all legal entities, on the forecast's own population. Contractor movement and one test record dated 30 September are excluded; one test record dated 30 September sits in Rippling and is excluded here.
Cash — bank transaction file plus the Concero 13-week model. Month to date is now actual cash on hand: the closing running balance of each account in the Concero bank transaction workbook (refreshed 23 September), converted at the FLC model's own rates — EUR 1.15, CAD 0.71 — so cash is stated on the same rates as the cost lines. Pattern's Chase account is excluded. The est. total is the cash forecast's own September month-end position, and BOM is the difference between the two. Bank of America holds $2k and has recorded nothing since 4 May. The RBC balance is as at 17 September, all other live accounts as at 22 September.
Hours. One-time and ongoing are split on the budget the time was logged against: a budget whose name is a retainer, a CRM sprint or a support agreement counts as ongoing, everything else as one-time. That rule puts 50.8% of month-to-date hours in ongoing, and the same mix is applied to the balance of the month. 14,751 logged through 22 September, plus 283 hours of logging catch-up (the completeness curve applied to each week already worked), plus 5,902 scheduled hours at 0.940 realization — the trailing three closed months. Realization is the single most sensitive assumption on the page: each point of it is about 59 hours and $11k of revenue.
Revenue. Those hours at each brand's realised rate, less contractual hourly discounts and the 3% concession the forecast is built on (a combined factor of 0.9640), with two fixed-fee overrides described below. Referral income is carried at the forecast and pro-rated 16/22 working days because it is invoiced rather than modelled from hours — which means that line cannot miss by construction. Total revenue and total hours are both capped at forecast; neither cap binds this month.
Bookings. Each open deal takes its statistical forecasted probability where HubSpot holds one (17 deals) and its forecast probability where it does not (64 deals). Stage conversion rates are deliberately not used: they are derived from new-business closed-won history, and most of this book is renewals. $9,740k of open pipeline weights to $3,494k, a blended 36%. One deal with neither probability set weights to zero. No manual adjustment is applied, so the line reflects HubSpot as it stands.
Cash. Month to date is the closing balance of each bank account at 22 September, $2,931k, with Pattern's Chase account excluded. The est. total is the September month-end balance the cash forecast carries, $2,585k. The balance of month is the difference between those two figures, a net cash outflow of $346k. Nothing in the balance-of-month column is modelled separately.
Cost of sales. Salaried delivery $1,581k, fully loaded, from the FLC model (23 September build), pro-rated for movers: joiners starting mid-month carry only the part of the month they worked and leavers are credited for the days they did — without that adjustment the model over-charges three joiners by $22k and under-charges five leavers by $17k. Salaried cost is split 16/22 and 6/22 by working day, because a salaried person costs the same whether the hour bills or not. Contractors are the only volume-variable piece: 3,393 hours logged plus 1,375 to come (their 24% share of the confirmed schedule at realization, plus their share of the logging catch-up), at $69.30 an hour. Other direct cost $75k, flat, split the same way.
Productive carries notional costing rates against fixed-fee budgets, which are not what the client pays, so valuing those hours at the notional rate understates the work. Two books are corrected. Estée Lauder Phase 2B is revalued from $195 to $340 across 842 hours worked, 16 hours of attributable logging catch-up and 361 scheduled hours at realization — $170k of revenue, of which $118k sits in the actual column. Pattern is revalued to $300 across all 940 of its hours, which adds $70k. Together the two overrides are $241k of the $3,969k — strip them and gross margin reads 46.7%, below forecast rather than marginally above it. The $300 is corroborated by Pattern's own realised ABR rather than by the forecast alone: the business-unit review puts August at $314.53 and year to date at $286.92, and the KPI pack's ABR bridge gives June $301.30 and August $308.69. The forecast itself sits at the same level — the KPI pack carries Pattern at $165.7k over 540 August hours and $244.2k over 800 September hours, implying $306.86 and $305.25. A $233.34 Pattern ABR appears on the business-unit review but does not reconcile; treat the KPI pack as the source.
EMEA is not corrected and still carries the understatement. Fixed-fee budgets are 43% of EMEA hours against 10% for Domaine NA, so EMEA revenue and EMEA margin are both understated here by an amount nobody has sized. This is the largest unquantified item on the page and is worth an hour of somebody's time.
The contractor rate. $69.30 is August's realised cost per hour from the KPI pack ($379.3k over 5,474 hours). The 7+5 forecast assumes $46.95 for September ($354.9k over 7,560 hours) and Productive's own booking cost for the scheduled contractor hours implies about $46. At $46.95, gross profit is $2,089k and margin 52.6%. Volume is 37% below the forecast and rate is 48% above it.
The salaried cost snapshot. The $1,581k is a point-in-time FLC snapshot pro-rated for movers. August actual was $1,618k.
The open deal book. $569k of weighted value, 16% of the balance of month, sits on deals whose close date has already passed, including a $200k deal carried at 90% in a stage where every other deal weights to zero. Hunter Douglas alone is $679k, 19% of the weighted balance, and the top five deals are 42%.
Referral income. $154k is carried at the forecast and pro-rated rather than modelled from hours, so that line lands on forecast by construction. It also drives the 259% on the bookings sub-line.
Realization and the logging catch-up. 0.940 and 283 hours are both trailing averages, based on how the last three closed months behaved.
Figures are rounded to whole thousands at display, so an expanded subtotal can differ from the sum of its parts by $1k. The underlying arithmetic is unrounded. Profit and loss lines carry better/(worse), sign-flipped for costs; the cash line carries higher/(lower) with no judgement implied. A/R and A/P have been removed from this view: the balance-sheet receivables forecast and the 13-week cash model do not reconcile to each other, and the aging detail invited a conversation about individual invoices that this dashboard is not the place for. Cash is the line to read; collections work happens elsewhere.