It says $2.4m influenced — out of how many deals?
Every money figure MarqIQ prints carries the count it was measured on, because a sum, a median and a headline count on one screen rarely rest on the same records. This article shows you how to read that count, which records each analysis reads, and what to check before you quote a figure to anybody outside your team. If two screens disagree about the same figure, read Two screens show different numbers for the same thing.
Why a money figure needs a count beside it
$2.4m across 12 deals and $2.4m across 900 deals describe two different businesses. An average that leaves out the deals with no amount on them is not an average of your pipeline.
MarqIQ prints the count beside the figure, every time. Read the line underneath it as well.
Reading a tile on Pipeline health
Pipeline health is the clearest example. To find it, open Intelligence Studio, then Data Sources, then your tool, and stay on the Executive Read tab. It shows up to seven tiles, and no two of them have to rest on the same records. Which tiles appear depends on what your deals carry.
| Tile | The line underneath | What it rests on |
|---|---|---|
| Win rate | 43 won of 118 closed | Closed deals only. Open deals are neither a win nor a loss yet, so they are left out of both sides. |
| Typical win | median of the 31 of 43 won deals with an amount · average $18,400 | Won deals that carry an amount. The median is quoted first and the average beside it, so a book skewed by one large deal is visible. |
| Days to win | median, created → close — or "median · lost deals take 74" where lost deals can be timed too | Won deals that carry both a created date and a close date. |
| Open past close date | 164 of 902 | Open deals only, counted against every open deal. Nothing closed is on either side. |
| Open older than 180 days | a count of open deals, with their value where deals are priced | Open deals only, by created date. This tile takes the place of Days to win where too few deals carry the dates needed to time them, and the day count is written into the label. |
| Top 5 accounts | of won revenue · largest deal 18% | Won deals that carry an amount, grouped by account. A share of revenue, not a share of deals. |
| Amount coverage | deals carrying an amount · only 72% of won deals do | Every deal in range. This is the tile that tells you how much to trust the others. |
Note
The Data Audit tab goes one step further. Where only some of a value's wins carry an amount, it prints the fraction inside the figure — "(12/19 priced)" — rather than quietly under-reporting the total.
When money is not shown at all
MarqIQ values nothing in money until at least three deals carry an amount above zero and those deals are at least 5% of the deals in range. Below that it says "Amounts are missing on most deals, so nothing here is valued in money" and shows counts instead.
The threshold is deliberately low. A few priced deals are still worth showing as long as you can see the coverage, so Amount coverage sits beside them as its own tile.
The populations MarqIQ counts in
These are the populations you will meet. A figure always names the one it used, and two figures on one screen can legitimately use two.
| Population | Where you meet it |
|---|---|
| People | Contacts synced from a tool. Lead scoring, Audience health and every conversion rate that counts people. |
| Deals | Opportunities synced from a tool. Pipeline, win rate, forecast. Pipedrive Leads are excluded — see the article on that. |
| Closed deals | Won plus lost. The denominator of every win rate. |
| Deals with an amount | The denominator of every sum, median and average of money. |
| Captured form fills | What the MarqIQ script recorded. Split into repeats, junk and real fills, each counted separately. |
| Matched contacts | People MarqIQ joined from its own captures to a record in your tool by email address. |
The Influence report, under Intelligence Studio, shows the last of those plainly. Its Pipeline touched tile prints the total over two counts — how many contacts were matched, and how many of those carry a deal. It reads the distinct email addresses behind influence-touched fills, up to 5,000 of them, and where somebody exists in more than one system it counts them once, at their highest known deal value.
Whether the figure read everything, or a sample
Every analysis states its population in one sentence at the top. There are two forms, and the difference matters when you quote a count.
- "Read all 77,662 contacts." Everything in range was read. Counts and shares are both yours to quote.
- "Read 6,000 of 41,220 contacts created in 2026 (of 77,662 in all): an even sample spread across the whole period." Shares are reliable. Counts are the sample's, so a count taken from a sampled read is smaller than your real total.
A sample is spread evenly across the whole period rather than taken from the most recent records, so it is not weighted towards this quarter.
What each analysis reads, and where it stops
Each analysis has a cap of its own. Read the population line on the analysis in front of you rather than carrying one number between screens.
| Analysis | What it reads | How much |
|---|---|---|
| Pipeline health, on the Executive Read | Deals created in the page's date range | Every one |
| Source truth | People created in the page's date range | Every one. It also reads up to 20,000 of MarqIQ's own captured fills when it joins your records to what the script recorded. |
| Field discipline and the Field Intelligence tab | Contacts, deals and accounts, profiled field by field | Up to 6,000 contacts, 20,000 deals and 10,000 accounts, per object |
| Lead scoring | People created in the page's date range | Up to 80,000 |
| Funnel | People created in the page's date range | Up to 80,000 |
| Forecast accuracy | Deals won or lost in the last two years, created in the page's date range | Up to 2,000 |
| Form fills | Form fills MarqIQ captured, newest first | Up to 60,000 |
| Audience health | People created in the page's date range | Every one |
A sampled read says so in its own population line. A read of everything says so too.
Take care
Forecast accuracy carries a window of its own on top of the date range you set. It reads only deals won or lost in the last two years, whatever range the page is on, and its population line says so. A forecast older than that says little about today's pipeline.
The date range at the top of the page is your tool's own created date on the record. A record carrying no created date sits outside every range, and MarqIQ says how many there are and counts them under All time.
Currency
MarqIQ stores each deal's currency code as your tool records it, and no analysis converts between currencies.
The Executive Read, the Data Audit tab and the deal card of the Integration Audit Report read that code. Where every deal with an amount is in one currency, they print their figures in it. Where your deals are in several, they add nothing across currencies and give each currency its own totals, with the number of deals behind each. Where your tool records no currency, they print plain numbers with no symbol.
The other screens add amounts up as the numbers they are. Most of them print a dollar sign, the report builder among them. Source truth abbreviates instead, writing 2.4M and 18k with no symbol at all.
If your deals are in one currency, read the figures as that currency and the totals are correct. If you sell in several, a total on those other screens mixes them. Quote by currency from the Executive Read or the Data Audit tab, or from your own tool.
Before you quote a figure outside your team
- Read the count underneath it, and quote the two together. Say "$2.4m across 118 closed deals", not "$2.4m influenced".
- Check Amount coverage. If a third of your deals carry no amount, say so before somebody asks.
- Check whether the population line says all or a sample, and never quote a sample's count as a total.
- Check whether the analysis has a window of its own on top of the page range. Forecast accuracy reads only deals won or lost in the last two years.
- Check the date range the page is set to. It is the record's created date, not the close date.
- Check that your deals are in one currency.
Take care
Do not add two figures from two screens together. A sum over deals and a sum over matched contacts are counted on different populations, and adding them counts some money twice.
If it did not work
A total and an average on the same screen do not reconcile
Why: They rest on different denominators, and both say so. The total covers every deal in range. The average covers only the deals that carry an amount.
Fix: Read the line under each tile. The number of priced deals is printed there, and dividing the total by it reproduces the average.
Your revenue figure is lower than the one in your CRM
Why: Deals with no amount on them contribute nothing. If a quarter of your wins are unpriced, a quarter of your revenue is missing from every sum on the page.
Fix: Read Amount coverage on Pipeline health. It states what share of deals carry an amount, and what share of won deals do, as two separate figures.
A count on one analysis is far smaller than the same count on another
Why: The two analyses have different caps. Lead scoring reads up to 80,000 people and Field discipline profiles up to 6,000 contacts, so the same workspace produces two counts that are both correct.
Fix: Read the population line at the top of each analysis. It states what was read and out of what, so you can see which of the two is a sample.
The count changed since you last looked and the figure with it
Why: A sync brought in new records, or the date range on the page is different from the one you had.
Fix: Compare the population line, not the figure. It states the range and the record count, so you can see which of the two moved.
Next steps
Related articles
Was this article helpful?