Weekly Business Review (WBR): How to Run One for a Revenue Team

A weekly business review (WBR) is a recurring meeting where leaders walk the same set of metrics every week, compare each with its target and last year, and discuss only the exceptions. Amazon made it famous. For a revenue team it covers customers, bookings and the forecast, then the inputs that drive them: pipeline created and deal movement.

The pages that rank for this term explain how Amazon runs its WBR, and they explain it well. Almost all of their examples come from retail and consumer apps: page views, in-stock rates, signups. A B2B SaaS revenue team has different problems. Bookings arrive in lumps, most of the quarter closes in the last three weeks, and the pipeline you see on day one is not the revenue you will book. This guide adapts the Amazon method to a revenue team: which metrics belong in it, how to set weekly targets that respect seasonality, how to tell a real exception from noise, how to explain a forecast change, and a free template that does the math.

What is a weekly business review?

A WBR is a fixed weekly review of a fixed set of metrics, where each metric has an owner and the meeting only spends time on what is unusual. The format comes from Amazon, where finance runs it and the metric owners present their own numbers. Commoncog's account of the Amazon WBR, drawing on former Amazon executive Colin Bryar, says the meeting answers three questions in order: what did customers experience last week, how did the business do last week, and are we on track to hit targets. Customers come first on purpose. The same piece reports that Amazon's leadership WBR covers 400 to 500 metrics in 60 minutes, 90 during the holiday season. Former Amazon executive Bill Carr, writing on LinkedIn, puts it at more than 200 metrics in a one-hour meeting. Either way, the count is high and the time is short, which only works because most metrics get no discussion at all.

Four mechanics make that possible:

• Same deck, same order, every week. The Working Backwards WBR page says reports should use standard formats such as the 6-12 chart, and that the format does not change from metric to metric, so people learn what normal looks like.
• Owners, not presenters. Each metric has one owner who explains it. If the week was routine, the owner says so and the meeting moves on.
• Exceptions only. Time goes to unusual movement and missed targets. Strategy debates are taken offline.
• Inputs over outputs. The meeting is built around the controllable input metrics that move the results, not the results themselves.

It does not need special software. The Working Backwards page notes that Amazon started its WBRs in Microsoft Excel, and the template on this page is a spreadsheet for the same reason.

WBR vs forecast call, pipeline review, MBR and QBR

The WBR is the one meeting that looks at the whole revenue system every week, with customers, outputs and inputs on one page. The other meetings in a revenue operating cadence each look at one part of it. The WBR does not replace the forecast call. The forecast call goes deal by deal. The WBR looks at the trend in the forecast and in the inputs behind it, and sends questions to the forecast call or the pipeline review when a number moves. If your WBR turns into a deal-by-deal walk, it has become a second forecast call. The sales operating rhythm entry covers how to keep each meeting to one decision.

Input metrics vs output metrics for a revenue team

Output metrics are the results, such as new ARR booked. Input metrics are what the team controls this week that moves those results, such as qualified pipeline created and deals decided. Amazon's WBR is built around inputs.

The Working Backwards authors explain why with a story quoted by Holistics: "The CEO, and companies in general, have very little ability to directly control output metrics." Their selection metric started as the number of detail pages, then became detail page views, then the percentage of page views where the product was in stock, and finally the percentage in stock and ready for two-day shipping. Each version was closer to what customers actually experienced. Adam Fishman's guide to setting up a WBR gives the short version: "An input metric is one that you or a team has influence on."

For a revenue team, the inputs come from how a quarter actually gets built. In ORM's customer data, only about 20% of the pipeline that carries an in-quarter close date on day one closes in that quarter. Pete's view is that a forecast has three sources of revenue: deals already in the pipeline, deals that will be created and closed inside the quarter, and deals pulled forward from later quarters. A WBR that only tracks bookings and the forecast total sees none of that. The inputs below do.

The movement metrics come from Pete's points on what predicts a miss. A rep changing the close date is the best single signal that a committed deal will slip. In the first half of 2026, many ORM customers had enough pipeline on paper, but fewer deals were being decided: days in stage crept up and cycles stretched while coverage still looked fine. And the change in the CRM often tells you more than its current state. A deal in Proposal is useful to know. Knowing it moved into Proposal yesterday is more useful. That is why four of the twelve metrics count movement, not totals.

Twelve metrics is a starting point, not a rule. Commoncog runs its own WBR on 47 metrics. Keep the WBR at team and segment level: the rep-level version of these inputs, weighted into one score for one-on-ones, is the sales scorecard template. Add a metric when an exception keeps coming back and nothing on the page explains it. Retire one when nobody has acted on it in a quarter.

How to set weekly targets for a SaaS quarter

Give each week its share of the quarterly target based on your own seasonality, not one thirteenth of it. A flat weekly target makes every early-quarter week look like a miss and every last week look like a triumph.

In ORM's customer data, the third month of a quarter is always the largest and the first is the smallest, and a large share of business closes in the last week. ORM breaks each customer's quarter into 13 weekly weights. One representative curve from Pete, where 1.00 is an average week:

The weekly target is the quarterly target times the week's weight divided by the sum of the weights (12.99 here). With a $1,300,000 quarterly bookings target, week 4's target is $1,300,000 x 0.81 / 12.99 = $81,062, and week 13's is $227,175. A flat target would be $100,000 every week. A $65,000 week 4 is 35% under the flat target and 20% under the seasonal one, which changes the conversation.

The quarterly targets themselves come from the plan; the annual operating plan template phases the year into quarters by seasonality. Build your own weekly curve from two or more years of closed-won bookings by week of quarter. Pipeline creation usually has a different shape from bookings, often flatter, so give it its own weights. For more on how seasonality distorts pacing, see sales seasonality.

How to tell an exception from noise

Use natural process limits: the average of the last 12 weeks, plus or minus 2.66 times the average week-to-week change. A week outside the limits is an exception worth discussing. A week inside them is routine variation, and the owner says "nothing to see here."

This is the XmR chart from statistical process control. The individuals chart formula sets the upper limit at the average plus 2.66 times the average moving range, and the lower limit at the average minus 2.66 times it. Commoncog describes the WBR as a process control tool, but notes in its essay on becoming data driven that the Amazon WBR "does not use process behaviour charts of any type." Amazon leans on years of practice to know what normal looks like. A revenue team starting a WBR does not have that yet, and the limits give it a fair rule from week one.

Two adjustments make this work for revenue metrics. First, for seasonal metrics like bookings and pipeline created, compute the limits on the percent of each week's target instead of the raw dollars, so a normal week 13 spike does not register as an exception. Second, add the run rule: eight weeks in a row on the same side of the average is a signal even when no single week crosses a limit. Slow drifts, like days in stage creeping up a day at a time, show up there first.

A metric that is inside its limits but off target is a different case. The process is behaving normally, and normal is not good enough. That calls for a change to the inputs, not an investigation of last week.

How to run a weekly business review in 7 steps

Cut the data on Monday morning, let owners prepare on Monday, meet on Tuesday, and close every exception with an owner and a date. In order:

1. Freeze the week. Weeks run Monday to Sunday. RevOps pulls the data Monday morning from the CRM, billing and the forecast submission, and nobody edits it afterward. Two versions of a number in the room end the meeting.
2. Publish the pack before the meeting. One page: every metric with this week, target, last week, same week last year, quarter to date, the limits and the signal. Amazon's company WBR runs midweek, after metric owners review their numbers and departments hold their own reviews.
3. Owners write a note for every exception and every missed target. What happened, the cause if known, what happens next. "We don't know yet" is an acceptable note. A guess dressed up as a cause is not.
4. Read the pack in order. Customers, outputs, inputs, context. Routine metrics take seconds. The Data Analysis Journal's guide to the Amazon WBR suggests 30 minutes for smaller organizations and 60 or more for larger ones.
5. Explain the forecast change as a bridge. Last week's forecast, plus deals added, plus upside moved in, minus slips, minus losses, plus or minus amount changes, equals this week's forecast. A number that moved without a bridge has not been explained.
6. Log every question with an owner and a due date. The same guide says unanswered questions go back to the owner, who follows up by email within a day. Open items start the next meeting.
7. Review the metric set every quarter. Add the metric that would have explained a recurring exception. Retire metrics nobody acted on. Tighten definitions that caused an argument.

Who attends: the CRO, the heads of sales, marketing, customer success and RevOps, and finance. At Amazon, finance runs the meeting and audits the metrics so they are hard to game, which is a good model for a revenue team too: the person who owns the forecast should not be the only one who checks it. Pete's habit for the RevOps lead who owns the forecast is to talk to the sales leaders every day and always before the weekly forecast call, because the model's assumptions have to be challenged with what the field is seeing.

Worked example: week 4 of a Q4

Illustrative example. A SaaS company near $14M ARR, Q4 bookings target $1,300,000, Q4 pipeline creation target $3,900,000. Report week Monday October 19 to Sunday October 25, 2026, week 4 of the quarter. These are the numbers in the template's worked example file.

The WBR pack flags 3 exceptions and 4 routine-but-off-target metrics. Five metrics get "nothing to see here".

What the meeting covers, in order:

• Bookings are behind but normal. $65,000 against a seasonal week 4 target of $81,062. Quarter to date is $251,000 against $273,210, 8% behind. Inside the limits, so no investigation of this week. The question is whether the inputs will carry the back half of the quarter.
• The forecast fell $48,000. The bridge: $1,262,000 last week, plus $38,000 new deals, plus $22,000 upside moved in, minus $71,000 slipped to next quarter (three deals), minus $24,000 lost, minus $13,000 in amount changes, equals $1,214,000. The forecast range is $1,090,000 to $1,330,000, so plan is still inside the range but above the base case.
• Pipeline created is the real exception. $165,000 is 55% of the week's target, below the lower limit of 73%. Quarter to date is $1,063,000 against $1,200,000. This hurts the in-quarter deals that close in weeks 11 to 13 and next quarter's starting pipeline. The owner's note: two partner webinars moved to week 6, and inbound demo requests fell. Marketing confirms by Thursday whether paid search spend paused.
• Deals are stalling. 24 close-date pushes against an average of 12, and median days in stage at 41, the eighth week in a row above its average. Deals decided, at 9, is still inside its limits, but the three movement metrics point the same way. This goes to the pipeline review with a list of every push and the buyer's stated reason.

Five metrics took no time, including churn, renewal risk and stage advances. That is the point of the format: the hour went to pipeline creation and deal velocity, the two things that decide whether weeks 11 to 13 deliver.

The WBR template, tab by tab

The workbook turns a weekly data table into the one-page pack and the 6-12 chart, with every formula visible. It works in Excel and imports into Google Sheets with no changes. Yellow cells are inputs. The 6-12 chart is the Amazon format: the trailing six weeks and the trailing twelve months side by side, each against last year. Six weeks show what just happened. Twelve months show whether it is a trend or a blip. Fishman uses a similar pair of windows, trailing 8 to 12 weeks and trailing 14 days. The box scores next to the chart give week over week, year over year, quarter to date against target, and year to date against last year.

Common mistakes

• Starting with bookings and the forecast, and stopping there. Outputs tell you what happened. The WBR exists to look at the inputs early enough to change them.
• Flat weekly targets. Every week 2 looks like a crisis and every week 13 like a win. Use a seasonal curve.
• Discussing every metric. If routine weeks get airtime, the meeting doubles in length and exceptions get the same weight as noise.
• Changing definitions without saying so. A redefined pipeline stage breaks the trend line. Note the change in the metric dictionary and keep the old series for a quarter.
• A WBR that becomes a deal review. Specific deals belong in the forecast call and pipeline review. The WBR sends them there.
• Building the pack by hand every Monday. The Data Analysis Journal piece notes that the Alexa WBR Excel workbook took half a day to update. Automate the data pull before you add metrics.

Where ORM fits

The hardest part of a revenue WBR is the movement metrics: close-date pushes, stage advances, days in stage and the forecast bridge. A CRM stores the current state of each deal, so measuring what changed last week means comparing snapshots. ORM snapshots each customer's pipeline every day, so it sees what moved and when, and uses those changes to produce a revenue forecast by week, with a range, that targets 95% accuracy. Gartner reports that only 7% of sales teams reach 90% forecast accuracy or better, with a median of 70% to 79%. A WBR built on a forecast that moves for explained reasons is a shorter meeting. To see it on your own data, see how ORM works.