A practical playbook · Revenue Command Centre
Will we hit the number, and if not, what do we do about it?
A revenue plan is a hypothesis. It's set in January from last year's numbers plus some optimism, and at some point in the year it stops being true. This playbook shows how to see that moment early, find the lever that broke, and act while there's still time.
Book a free growth reviewThe problem
Teams see the missed number, not the reason behind it
I've sat in a lot of monthly reviews where I could tell at first glance we'd miss the quarter. The forecast was built on numbers that no longer matched what was happening, so carrying on at the same rate meant arriving short.
The plan itself doesn't drift. The assumptions behind it do. Targets are modelled on last year's deal size, win rate and cycle length, and when those move, the results follow within weeks. Most teams only act at the end of the quarter, when the pipeline that could have closed the gap needed to exist months earlier.
Part of the reason is how the numbers travel. Senior leaders hold the full picture. The operators in marketing, BDR and sales see the missed number, but rarely the drift behind it. And the weekly meeting goes on reconciling versions instead of deciding what to do. Results don't feed back into the ICP, the tiers or the scoring either, so the same accounts get the same attention quarter after quarter.
Who it's for
One view, used in multiple rooms
Fixing that starts with everyone looking at the same numbers. Senior leaders use the same view in board prep. Marketing, BDR and sales use it in their weekly sessions, and RevOps keeps the definitions behind it clean. Access follows role, so each person sees what they need to act on.
| Role | Why it matters |
|---|---|
| Executive team | Decide early whether to hold the target or reset it, with evidence instead of an end-of-quarter revision. |
| Sales leaders | See which deals and tiers deserve rep time this quarter, and which to stop opening. |
| Marketing | Tie campaigns, scoring and budget to what actually closes, not to what generates activity. |
| RevOps | Keep one set of definitions clean, so every team trusts the same numbers. |
In practice
What happens when activity is the KPI
In one audit I ran, the data showed most of the sales team's activity going into the wrong tier of accounts. When I spoke to a few sales leads, it emerged that the half-year review had told the sales team two things: they hadn't closed enough to hit target, and they hadn't generated enough activity. That led to a spike in lower-tier opportunities being opened, because those accounts were far more available. When I looked into it, nobody was gaming anything. Opening deals was a KPI, and Tier 3 accounts were the quickest way to show activity. The team was responding rationally to what it was measured on.
Working with the team, we then found the same pattern with SDRs, in demos booked against trials actually started, because the same message about activity had reached them too.
The demo shows the same shape. Tier 3's share of new opportunities climbs every quarter, from 31% in Q1 to 58% in Q3. Over the year it takes 49% of rep time and returns 13% of new ARR won. Once that's on the screen, the conversation changes. It's no longer about whether the team is busy. It's about whether that time will ever get the team to the number.
Prerequisites
What you need first
Spotting a pattern like that early depends on the data underneath it. This is what needs to be in place first.
- 01
A CRM like HubSpot, with reliable company, contact, deal, stage and outcome data, enriched with fresh company and contact details, and a deal type on every opportunity.
- 02
Call intelligence like Gong, with recordings linked to the right deals.
- 03
Defined targets and forecasts, plus marketing KPIs, channel costs and account tiers.
- 04
A consistent record of why deals are won and lost.
- 05
One person who owns the forecast.
If deal types or close dates are unreliable, fix those first. Every rate downstream inherits the error.
How it works
One shared data layer, four parts on top
With those in place, everything reads from one data layer. The parts are built in this order, and each one has a clear test for when it's done.
System diagram
How the command centre fits together
- Trigger
- Data collection
- Agent
- Dashboard
Dashboard
Weekly view for every team
Dashboard
Weekly view for every team
-
Sources
CRM, call intelligence, marketing data
-
Data agents
Clean, link, check. 30-day engagement rule
-
Shared data layer
One repository, like Notion
-
Metrics agent
90-day rates, lever impact
-
ICP agent
Rescores the tiers
-
Insights agent
What drives it, what to do
Closed deals flow back, and the ICP agent rescores the tiers
Data agents. Bring CRM, call intelligence and marketing data into a data repository like Notion, clean it and link every record to the right account and deal. Engagement only counts when it's continuous: a one-off download doesn't, and a gap of more than 30 days restarts the clock. Done when closed-won matches finance.
Metrics agent. Compare targets with actuals on trailing 90-day rates and price each lever as the forecast with that lever back at plan. As history builds, add seasonality. Done when the lever impacts add up to the gap.
ICP agent. Rescore the account tiers from every win, loss and piece of feedback. Done when you can say where rep time is wasted.
Insights agent and dashboard. Explain what's driving performance, recommend what to do next, and publish it weekly. Done when the weekly meeting opens on the dashboard.
Design choices
Why it's built this way
Two choices make the difference between a dashboard and a system.
One definition, held in one place. Guardrails, tone of voice, the ICP, metric definitions and the forecasting model all live as documented pages in one repository. They're controlled centrally and locked, but every team can read them. Each page tells an agent what it governs, so the numbers and rules come out the same whichever team pulls them.
People can still reach different conclusions from the same data. That's the debate a weekly meeting should have, instead of arguing about whose numbers are right.
Built like microservices. You don't need the whole system on day one. Start with the core: your sources, the shared data layer and a simple always-on view. Then add services as the need appears, such as channel views, account scoring or lead alerts in Slack. Each one plugs into the same data and guardrails, so you grow the system instead of rebuilding it.
The first step is always a diagnosis: where the gaps are, and what will move the number for this organisation right now. In a smaller company that core can be in place quickly. The content engine playbook is one of the services that plugs in later.
Build order
Start small, then add services
- Trigger
- Data collection
- Agent
- Dashboard
- Added later
-
Sources
CRM, call intelligence, marketing data
-
Shared data layer
One set of definitions and guardrails
Core view
The revenue plan against actuals, always on
-
Add services
Channel views, account scoring, lead alerts, content engine
A working sample
See it working
You can see all of this working in one place. The demo is modelled on real data, with names replaced and company anonymised. It has four views, Overview, Marketing, Accounts and Detail, and shows the plan, the drift, and where time goes against where revenue comes from.
FAQ
Questions
Why 90 days?
It's one quarter of closes. That's long enough that a few deals don't swing the rates, and short enough to catch drift while there's still time to act. The view can switch to 30 days or year to date, but the lever impacts always use 90.
What if our CRM is messy?
Most are. The build starts with the cleanup, and the weekly checks block a forecast built on bad data. It also flags any deal whose close date has moved twice.
How is this different from a BI dashboard?
A BI dashboard reports what happened. This one recomputes what's now required, prices each lever, dates the window to act and says what to do next.
Does it work below £10M ARR?
Yes. With fewer deals, one month can swing a win rate, so it uses a longer window and shows ranges instead of single numbers.
Put the system to work
A plan is a hypothesis. Test yours every week.
Let's look at your plan, your CRM and your forecast meeting, then decide where to start.
Book a free growth review