Fix close dates that keep being pushed
A close date that moves to the end of next month every time it arrives is not a forecast, and the whole quarter is planned against it. MarqIQ measures the habit twice: how many open deals have been pushed three times or more, and how late your won deals really closed against the date they carried a month out. Read Fix stages that move backwards if the stage ladder is the part that looks wrong.
Before you start
Plan required
Any plan except Starter. Both screens sit under Analysis, which Starter does not include.
Permissions required
Neither screen writes anything to your tool. Every change below is made by whoever administers your CRM.
- A close-date field on deals, with change history MarqIQ can read. On Salesforce this is always available, because Salesforce records stage, amount and close date on every opportunity whether or not Field History Tracking is on.
- For the push count: at least ten open deals whose close date has a dated history.
- For the slip measurement: at least 30 won deals, closed in the last two years, that were open for 30 days or more.
Count the pushes on open deals
This is the habit as it is happening, on deals that are still open, so it is the number to take into a pipeline review.
Open Insights in the left-hand menu, then Analysis.
Pick your CRM along the top.
Open the Field discipline tab.
The first read takes up to a minute. The reading is kept for a day and the line above the cards gives the date it was taken.
Find the card headed Close dates are being pushed, not forecast, under Fields being used against their purpose.
Read the two evidence lines.
The first counts the open deals checked whose close date has been moved later three or more times, and gives that as a share. The second gives the total slip on those deals, as a typical figure rather than an average pulled by one extreme deal.
What the card counts, exactly
- A push is a close date more than a day later than the one it replaced. The comparison is between the two dates, not between the times they were typed, so nudging a date forward by an afternoon is not a push and rewriting the same date is not counted at all.
- Deals recorded as won or lost are left out. A closed deal had its close date overwritten with the real one, so counting it would flatter every forecast ever made. A deal whose state MarqIQ does not hold is counted with the open ones.
- The card appears once there are at least ten open deals with a dated history, at least three of them pushed three times or more, and those are 10% or more of the open deals checked. It is tagged Used against its purpose at 25% or more, and Early signs below that.
- The share is of the open deals whose history was read, not of your pipeline. That sample is up to 600 deals on HubSpot and Salesforce, and up to 60 on Pipedrive, Zoho CRM and Microsoft Dynamics, spread evenly across the store.
The card does not name who pushed them
It lists the deals, not the people. MarqIQ names the writer behind a change on its other Field discipline findings, and on this one it does not. To see who moved a date, click Show the N deals, open one in your tool and read that record’s own history. On Salesforce and Microsoft Dynamics the history names the account a change was made under, which may be an integration rather than a person.
Measure how far the dates actually slipped
The push count says the habit exists. This says what it cost you, on deals that really closed.
Open the Forecast accuracy tab for the same tool.
It is offered for HubSpot, Salesforce and Pipedrive. The first read takes up to a minute.
Read the four tiles at the top.
They give the closed deals judged, how many were won, how many were lost, and how many were logged after the fact.
Find the finding whose title gives the share of won deals that closed more than a month late.
Read its evidence line for the three figures you can quote.
How many won deals it was measured on, how many days later than the forecast date the typical one closed, and the share that closed within a week of it.
How the slip is worked out
- For each won deal, MarqIQ reads the close date as it stood 30 days before the deal was decided, from the deal’s own history, and compares it with the day it really closed.
- A deal that was open for less than 30 days is left out of the slip figure: it never carried a close date a month before it closed, so there is nothing to compare. The Logged after the fact tile counts a wider group — every deal open for less than 14 days, which is the bar the rest of the tab uses.
- The figure needs at least 30 won deals with that history. Below that no percentage is claimed at all.
- The finding is raised when 30% or more of those wins closed more than a month late, and is tagged Misleading the forecast at 50% or more.
- Only deals won or lost in the last two years are read at all, and at most 2,000 of them, as an even sample rather than the newest. How this was worked out, at the foot of the tab, prints both numbers.
Lost deals are not in this figure
The slip is measured on won deals only. A deal that slipped for six months and was then lost contributes to the push count on the other tab and to nothing here.
Put a money figure on it, honestly
MarqIQ does not attach an amount to the slip itself, because a deal that closes a month late is late, not lost, and pricing the delay would mean inventing a discount rate. What it does give you is the value sitting past its date right now.
Open Executive Read in the row of views along the top, for the same tool.
Read the Open past close date tile.
It counts open deals whose close date is more than seven days in the past, as “12 of 88”, and the line underneath gives their total amount where your deals carry amounts. Where they do not, it counts the open deals created more than 180 days ago instead.
Quote both halves of that figure together.
The count and its denominator travel as one number. A total with no denominator is the sentence a board remembers and nobody can check.
Take care
Where fewer than three deals, or under 5% of the set, carry an amount above zero, every money figure on the Executive Read is absent rather than wrong. A missing money figure is not a zero.
The three operational changes
These are the fixes MarqIQ prints on the card, in the order it prints them.
Tie the close date to something the buyer has agreed.
A signed-off timeline or a procurement step gives a date that can be defended. The end of the month gives a date that moves every month.
Make a third push a decision point in the pipeline review.
A deal whose date has moved three times is either re-qualified with a new agreed date or closed as lost. It does not get a fourth push.
Count pushes in a field, so the third one is visible without a history check.
MarqIQ looks through your own deal fields first and names an existing number field whose label mentions push, slip, moved, rescheduled or times changed. Where you have none, the card says so. On HubSpot it adds that a workflow triggered when the close date changes can add one to a number property.
Put that field on the pipeline board and in the forecast review.
Where owners are named, and where they are not
MarqIQ names owners on the forecast-category finding, not on the slip. Where your workspace has a forecast category field, Forecast accuracy shows what each call turned into and names the owners whose commits hold and the owners whose do not — and it names an owner only with five or more deals in that category, so a run of three unlucky deals never becomes a person’s reputation. The slip figure itself is reported for the whole set.
Check the stage probabilities while you are there
A pushed date and an overstated stage probability inflate the same weighted number, and Forecast accuracy tests both on one read.
- The stage table gives each stage’s configured probability beside the share of closed deals that reached it and were won, with the count of closed deals behind each row.
- A stage is called says more than it delivers only with at least 30 closed deals behind it and a configured probability more than five points above the top of the 95% range for the real rate. Below that no verdict is claimed.
- Where a pipeline carries probabilities that your tool is not applying, MarqIQ says so once, naming every pipeline it is true of, because every weighted total in that pipeline view is unweighted while the numbers sit there looking maintained.
If it did not work
There is no Forecast accuracy tab for your tool
Why: It is offered for HubSpot, Salesforce and Pipedrive. Those are the tools whose deal history carries stage, forecast category and close date in a form MarqIQ can read.
Fix: Use Field discipline instead, which runs on six tools and counts pushes on open deals. It answers how often dates move, though not how late deals really closed.
Forecast accuracy says too few closed deals have enough history to judge
Why: Fewer than 30 deals both closed in the last two years and were open for at least 14 days. A rate on fewer than that would move with a single deal.
Fix: Widen the Created range at the top of the screen, or come back when more deals have closed. Nothing is claimed rather than a percentage being claimed on a handful of deals.
The Logged after the fact tile is large
Why: Those deals were open for less than 14 days, so they were entered once they were already decided. Nothing on the tab judges them as a forecast.
Fix: This is worth acting on in itself. Deals created at contract stage were never in the pipeline while they were open, so the forecast could not have included them. Create the deal at the first real sales conversation.
The push count looks low against what your reps describe
Why: The card counts a move to a date more than a day later than the one before it, on open deals, in a sample. A deal that is pushed and then closes leaves the open set, and a small nudge inside a day is not a push.
Fix: Read the slip figure on Forecast accuracy beside it. The push count describes the open pipeline; the slip describes what really happened to the deals that closed.
You cannot tell who pushed a date
Why: This card lists the deals, not the writers. The writer is in your tool’s own record history.
Fix: Click Show the N deals, open one in your tool and read its close-date history there. On Salesforce and Dynamics, check whether the account named belongs to a person or to an integration before you raise it with anybody.
The figures have not moved since you changed the review
Why: Both tabs keep their reading of your tool for a day, and both count what has already happened.
Fix: Click Read the latest from your tool to force a fresh read. Then watch the share rather than the count: it falls as deals created since the change replace the ones in the sample.
Pipedrive numbers look thin next to HubSpot’s
Why: Pipedrive gives a record’s change log one record at a time, so MarqIQ reads up to 60 deals rather than 600. The tab says so in its coverage note.
Fix: Read the share, not the count, and quote the sample size with it. For a full count, build the report in Pipedrive using the check printed under See it for yourself.
Next steps
Related articles
Was this article helpful?