Account based selling for small teams, without the enterprise stack
Account based selling was designed by teams with budget for it. What genuinely transfers to a team of four, what does not, and the rules that bind it.
Read
A dashboard is not a report. A report tells you what happened; a dashboard is supposed to change what you do on Monday. Most fail that test.
Most B2B sales dashboards are built the wrong way round. Somebody opens the CRM reporting tab, sees what it can chart without extra work, and charts it. The result is a wall of activity counts that nobody has ever used to make a decision, refreshed weekly, defended annually.
The test for whether a number belongs on a dashboard is narrow: if it moved by half, would anybody do something different this week? Most of what ends up on sales dashboards fails that question, and a dashboard where nine of twelve tiles fail it trains the team to ignore all twelve.
The most common structural fault has a documented fix that most small teams never turn on. A deal's stage and a deal's likelihood of closing are two different facts, and CRMs model them separately for good reason.
Salesforce's documentation on forecast category mapping sets out the separation explicitly. Stages such as Prospecting, Qualification and Needs Analysis map to a Pipeline category; Proposal or Price Quote maps to Best Case; Negotiation maps to Commit; Closed Won maps to Closed and Closed Lost is Omitted. The documentation also notes that "if the stage of an opportunity changes, the probability and forecast category also change", and that reps can override the mapping on deals they own.
That override is the important part, and it is why the two fields exist. Stage describes what has happened. Forecast category describes what the person closest to the deal believes. A dashboard built only on stage will tell you a deal at Proposal is nearly closed, which is exactly the belief that produces a missed quarter.
These are the ones that reliably change a decision in a small B2B team. Not all of them are hard to produce, but all of them require the CRM to be honest, which is the real project.
| Number | The decision it changes | Cadence |
|---|---|---|
| Qualified pipeline against target | Whether to prospect or to close | Weekly |
| Commit value versus best case | What to tell the board | Weekly |
| Stage-to-stage conversion | Which part of the process to fix | Monthly |
| Average days in current stage | Which deals to intervene on | Weekly |
| Deals with no next step booked | Which deals are actually dead | Weekly |
| Win rate by source | Where to spend next quarter | Quarterly |
Two of those deserve a note. Days in current stage is more useful than total deal age, because total age tells you a deal is old and stage age tells you where it got stuck, which is actionable. And deals with no next step booked is the single most predictive hygiene number we know of, for an unglamorous reason: a buyer who will not put thirty minutes in a calendar has told you something the CRM stage has not caught up with yet. That failure mode is the subject of why B2B deals stall.
Activity counts come first. Calls made, emails sent and meetings booked are trivially easy to chart and almost impossible to act on, because they measure the input rather than whether the input worked. They are worth tracking privately for coaching a new rep and worth keeping off a dashboard that management reads, since the moment activity is a target the team optimises the count.
Total pipeline value is second. Without a qualification standard behind it, total pipeline is a number the team can inflate by leaving dead deals open, and it usually is. Pipeline is only meaningful against a written definition of qualified, which is why we treat a qualification framework that survives an audit as a prerequisite for the dashboard rather than a separate project.
Third is any average that hides a small sample. With eleven closed deals a quarter, average deal size is not a statistic, it is one large deal and a rounding error. Show the list rather than the mean when the count is small, and resist the chart that implies precision you do not have.
If you track one meta-number, track what you predicted at the start of the quarter against what closed. It is the only measure of whether the rest of the dashboard is telling the truth, and it is nearly always the first one a team skips, because it is the one that can embarrass someone.
A team that is consistently forty percent optimistic does not have a forecasting problem it can fix with a better chart. It has a qualification problem, and the dashboard is faithfully reporting it. The fix lives upstream, in how the pipeline is built in the first place.
There is a strong temptation to solve this with a BI platform. For a team under about ten people, that is usually premature, and the useful sequence is the reverse: run the six numbers from CRM reports or a spreadsheet for a quarter, find out which ones anybody actually looks at, then build only those properly.
Our colleagues at Rivl make the same argument from the engineering side in a piece on what "real time" should mean for a sales dashboard, and their conclusion is worth borrowing: most teams asking for live data need yesterday's data to be correct, which is a different and much cheaper problem.
None of this works on a CRM nobody updates. Every number above is a claim about reality that the CRM can only make if a person told it the truth, promptly. Where the data is entered on Friday afternoon from memory, the dashboard is a well-formatted opinion.
This is why the first ninety days of any of this is mostly process rather than analytics, a sequence we set out in the first 90 days of business development. Getting five fields filled in reliably beats getting twenty fields filled in occasionally.
A dashboard does not create pipeline and it does not close deals. At best it shortens the time between something going wrong and somebody noticing, which is genuinely valuable and considerably less than what dashboards are usually sold as.
There is also a real case for not building one yet. Below roughly twenty deals a quarter, most of these numbers are too noisy to read, and a weekly conversation over the actual list of open deals will beat any chart. Build the dashboard when the list has grown past the point where one person can hold it in their head, and not as a way of appearing to run a sales function before there is one to run.
Send us your last two quarters of closed deals, won and lost. We will tell you which of the six numbers you can actually produce today, and which ones your CRM cannot support yet.
Book a Free Consultation