Ask your CRM what your pipeline looks like right now and it answers instantly. Every deal, every stage, every amount, exactly as it stands this second. Ask it what changed since last month (which deals slipped, which stage started leaking, whether this quarter's wobble is new or has been building since spring) and it goes quiet.
That's not a bug. It's the idea the whole category was built on. A CRM is a system of record for what is: the current state of every deal, kept clean and up to date so a rep can work their day. What was is a second-class citizen. History gets kept on the side, usually only for the big events (stage changes on a deal you happen to open), and never in a shape you can actually count across the whole pipeline. Come Monday, last Monday's board is gone.
The strange part, to an engineer, is that the data isn't missing. HubSpot sees every one of those changes as it happens. It just stores them in a way that's built for one deal at a time, not for the question you actually have, which is always about the whole pipeline moving over weeks. So the value is right there, captured. It's just locked in a shape that takes real work to get back out.
It's worth being precise about which numbers this actually hits. Ask HubSpot your win rate, your open pipeline value, your average deal size, or where your leads came from, and it answers in a click. Those are the easy numbers, the ones every team already has on a dashboard, and they tell you the score. The catch is that the score never tells you why it's moving. For that you need the four below, and those are exactly the ones a live CRM won't hand you, because every one of them is a story about change over time.
The numbers that only history can give you
None of the metrics below can be read off today's board. Every one of them is a comparison between two points in time, which is exactly what a snapshot-in-time makes possible and a live CRM makes hard. Here are four worth the trouble, what each one tells you, and how teams pull it out of HubSpot by hand today.
1. Stage-to-stage conversion
Of the deals that reach a given stage, what share move forward versus die there. The stage with the worst drop-off is where your revenue is leaking out of the funnel. Track it month over month and you learn the thing that matters most: whether a leak is new this month or has been chronic all along, and nobody noticed because you only ever saw the snapshot.
By hand: export every deal with its current stage and HubSpot's "date entered stage" fields. Take a cohort (say everything created last quarter), count how many ever reached each stage, and divide each stage's count by the one before it. Then do the whole thing again next month and line the percentages up side by side.
2. Time in stage
The median number of days a deal sits in each stage before it moves. A stage that's slow is where deals go to quietly die; a rep whose deals sit longer than everyone else's is where your next coaching conversation is. Watch it over time and rising time-in-stage becomes an early warning. You can see the quarter starting to clog weeks before it shows up in closed revenue.
By hand: HubSpot stamps a date each time a deal enters a stage. Export those, subtract each stamp from the next per deal to get days-in-stage, and take the median per stage. Split it by rep to find who's slow. The maths is trivial; keeping the export clean enough to trust it is not.
3. Close-date slippage
How far, and how often, deals push their close date out. This is the number that explains why your forecast lies: a deal that has slipped its close date three times is going to slip it again, and weighted pipeline has no idea. Count slips per deal and per rep and you can see which commits to actually believe.
By hand: export deal id and close date this month, and again next month. Match the two on deal id, flag every deal whose close date moved later, and record by how many days. Tally the slips. The catch: you can only ever compare against snapshots you remembered to take. Miss a month and that comparison is simply gone.
4. Net pipeline movement
New pipeline created in a period versus what left it (won plus lost) for the same period. It tells you whether you're filling the top of the funnel faster than you're draining the bottom. It's the momentum number, and it turns negative long before a bad quarter is obvious from the headline figure.
By hand: export deals by created date and by closed date for the period, sum the amounts on each side, and subtract. Repeat every period and chart the trend. Simple in isolation, but only meaningful as a series, which means doing it on a schedule and never missing one.
The ritual that holds it all together
Every one of those numbers depends on the same underlying habit: taking a full snapshot of the pipeline on a regular cadence and keeping the old ones. So that is what disciplined teams do. Someone technical enough to be trusted with it runs a monthly (sometimes weekly) export of the whole deal list to CSV, date-stamps it, files it next to the last one, and then spends a morning in a spreadsheet lining this month's export up against the previous ones: matching on deal id, diffing the columns, rolling it up into what moved, what slipped, and where it leaked.
It is genuinely worth doing. The teams that keep it up walk into a pipeline review knowing exactly what changed and why, while everyone else is squinting at a live board and guessing. If you're not doing some version of this, you are leaving your most useful reporting on the table.
Why almost nobody keeps it up
The problem isn't the idea, it's the upkeep. It's a technical, disciplined chore that lands on someone's plate every single month, and it breaks in ways that are easy to miss:
- It eats a morning, every month, from exactly the person whose time is most valuable: the one technical enough to be trusted with the export.
- It punishes any gap. Miss a month and the chain has a hole you can never fill, because last month's pipeline is already overwritten. The series is only as good as your consistency.
- It breaks when the CRM changes. Rename a stage, add a property, tidy up a pipeline, and your lookups silently match the wrong rows. The spreadsheet doesn't error. It just quietly starts lying.
- One wrong filter and the whole comparison is off, and there's no test to catch it. You find out when a number looks weird in a board meeting, if you're lucky.
- It's monthly at best. Whatever leaked on the 3rd, you learn about it on the 30th, a full cycle after you could have done anything about it.
So most teams that know they should do this still don't, or do it for a while and let it slide. Not for lack of understanding, but for lack of the time, the discipline, and the appetite to babysit a spreadsheet forever.
Which is the whole reason we built Reven.eu: it takes that snapshot for you every hour instead of once a month, keeps the recording forever, and has all four of these numbers computed and current the moment you open it: no export, no reconciling, no morning lost to a spreadsheet.