Read the Data Audit for a connected tool

Open the Data Audit

1. 1

   Open Insights in the left-hand menu, then Data Audit.

   Executive Read is the narrative version of the deal store; Data Audit is the measured version of everything.

2. 2

   Pick the tool you want to read along the top.

   The line under the pills gives the tool, its contact count, its opportunity count and the date of the last sync.

Set the created range first

The Created filter at the top scopes the whole page. Set it before you read a single number, because every figure below it changes when you do.

1. 1

   Pick a period: All time, Last 30 days, Last 90 days, This quarter, Last quarter, This year, Last 12 months, or Custom.

   Custom opens two date boxes with to between them. Nothing on the page changes until you click Apply.

2. 2

   Read the sentence beneath the pills before you go on.

   It names exactly what is covered — for example “People created 1 Jul 2026 – 30 Sep 2026 · 176 of 715; deals created in the same period · 42 of 310”. That is the denominator for the page.

3. 3

   Click Clear to go back to all time.

• The range is the date your own tool says the record was created, not the date MarqIQ synced it.
• It scopes people and deals together, so a contact created in the period and a deal created in the period are counted, and nothing else is.
• A record carrying no created date is outside every range. The count of those records is named in Data caveats rather than quietly dropped.
• The range rides in the page address, so a link you send opens on the same view you were reading.
• The field-by-field profiles on the Field Intelligence page always describe the whole store, whatever range is on. On this tab you see that in Data caveats: every finding the engine raised on your contacts is prefixed “Contacts, whole store” while a range is set. The Executive Read page prints the same disclosure in its scope line.

The headline tiles

A grid of tiles, four across, in a fixed order. One list in the code builds both the screen and the Summary sheet of the Excel export, so the two cannot disagree. Money appears only when at least one deal in scope carries an amount — where none does, the counts are shown alone, because “$0” would read as “we sold nothing” rather than “we were not told the price”.

The currency the money is in

MarqIQ reads the currency your tool records on each deal and does not convert between currencies. What you see depends on the deals in scope that carry an amount.

Outcome mix

One bar in three parts — won, open, lost — with the counts and the money on the line beneath it. It is the same deals as the tiles above, drawn as proportions so you can see the shape of the book in one glance.

• The segment widths are each outcome’s share of the deals in scope.
• The money behind each outcome is summed over the deals in that outcome that carry an amount.
• Where no deal in scope carries an amount, the line ends “no amount column mapped, so these cannot be valued” in place of the money, rather than printing zeros.
• Where your deals are in more than one currency, the line ends “valued per currency below, because the deals are in more than one”. The Deal value, by currency table under it has one row per currency: the deals with an amount, the won, open and lost value with the number of deals behind each, the value across all outcomes, the average deal and the largest deal.

Lifecycle funnel

Where every synced contact sits today — not where they have ever been. Each bar is the number of people currently at that stage, drawn relative to the largest stage rather than as a cumulative drop.

• Known stages come in order: Subscriber, Lead, Marketing Qualified Lead, Sales Qualified Lead, Opportunity, Customer, Evangelist, Other.
• Any stage your tool has that is not on that list is appended after them, largest first.
• A custom stage that syncs as an opaque number is shown under your tool’s own picklist label where MarqIQ can read it, and as “Custom stage 12” where it cannot. A bare number never reaches the chart.
• The count is contacts in scope, so the funnel moves when you change the Created range.

Sourced vs. influenced

This is the card that settles the marketing-versus-sales credit argument, because it prints the rule it used and the field it used it on. Every value is classified into Marketing-sourced, Sales-sourced, Partner-sourced, Other, or No source recorded. The line above the bar names the field — “Based on …” — and the last line of the card names the field again with the keyword rule that classified it.

The rule

• Partner, affiliate, reseller and alliance name themselves, and are matched first.
• An explicit Mktg or Marketing token beats a sales token, so “Outbound - Mktg - SDR” counts as a marketing-run motion.
• Then bare sales tokens — Sales, AE, SDR, BDR, Outbound, cold call, prospecting.
• Then marketing channel names — Inbound, Organic, Paid, PPC, SEM, SEO, Social, Email, Webinar, Event, Content, Referral, Direct, and the rest.
• A value matching none of those is Other. A blank value is No source recorded.

See what is in a group, and move a value

Click any group to open the source values grouped under it, largest first, each with its count, its wins and its pipeline. If a value sits in the wrong group — your “Trade show” is run by sales, say — pick the right group in the Move to control beside it. MarqIQ remembers the choice for that tool, recomputes the split on the spot, marks the value chosen, and says how many values you have placed by hand in the line under the card. Every value you have not placed still follows the rule above; Back to the keyword rule undoes a choice. Choices are per tool, need edit access, and land in the Excel export under each group.

Which split you are looking at

There are two, and only one is shown. Read the “Based on …” sub-line before you quote the percentage: it names the field the split rests on.

The influenced row

The purple Marketing-influenced row counts contacts with at least one recorded form conversion — a first or most recent conversion event name, or a conversion count above zero. It appears on the contact split only, and the card states in its own words that influenced overlaps sourced: one contact can be both.

• Where deal sourcing has taken the slot, there is no influenced row and no Marketing-influenced tile. Nothing is wrong; the deal split does not compute one.
• Where no conversion data is synced, Data caveats says the influenced figure cannot be computed and will appear once conversion properties sync.

What produces customers

Up to nine dimensions, each counted over the contacts that became customers — at the Customer or Evangelist stage, or on a won deal — and ranked by how many customers the value produced. It is the volume answer: what is loudest.

• The nine dimensions are Lead source, Latest source, Owner, Lead status, Campaign, Country, Industry, Company and Job title.
• A dimension appears only when the field behind it is filled on at least 8% of the contacts in scope and holds two or more distinct values.
• Each dimension shows its top five values.
• The rate beside each value is a conversion rate: customers carrying that value ÷ every contact in scope carrying it. It is not a win rate, and it must not be compared with one.
• The whole card appears only once at least three contacts are customers. Below that, Data caveats says how many you have.

What converts best

The same dimensions, re-ranked by rate instead of by count. This is the card that shows the segment converting far above baseline on modest volume — the one a list ranked by count can never reach, because it is never in the top five by volume.

• A value is ranked here only when at least 30 contacts in scope carry it. Below that floor it is left out.
• The floor exists so that “1 of 1 converted — 100%” can never reach the top of a report you hand to a board.
• Each dimension shows its top five values by rate, labelled “Ranked by rate, minimum 30 records.”
• It sits after the volume card on purpose: “what produces the most” is the question you arrive with, and “what converts best” is the one that changes next quarter.

Volume leaders

The same dimensions counted over every contact in scope, not only the ones that converted. This is the shape of the database, and it is the denominator every other card on the page divides by — read it before you argue with a percentage.

• Top five values per dimension, by record count, across every contact in scope.
• Where your deals carry a source field with two or more distinct values, a further table appears named after that field — headed Top deal lead sources where the field is called Lead Source — showing its top five by deal count.
• Picklist values are shown as your tool labels them, never as the stored code. Where two codes share one label, their counts are added together.
• Campaign values stored as ids are resolved to campaign names where a connected tool can name them. Ids nobody can name stay ids, and Data caveats says why.
• An owner field that resolves only to a raw id gets no table at all and is left out of Report dimensions found, because nobody can read “Owner: 27231495”. Its fill rate is still measured, and Data caveats reports it.

Data caveats

The last block on the tab, and the one to read before you send anything. It collects everything that would otherwise make a reader distrust the page, and raises each objection first. Amber items are warnings; grey items are notes.

What you can find in it:

• How many opportunities carry no amount, so you know what the revenue totals rest on.
• How many contacts have no email address, and therefore cannot be matched across systems.
• How many duplicate emails there are, and that counts treat each record separately.
• The percentage of contacts carrying no lead source, with the field name in quotes — worded to distinguish “this field exists and is empty” from “nothing here is a lead source at all”.
• Whether an owner field is present, and what share of contacts have no owner, with the field name in quotes.
• How many people or deals fall outside the range you picked, and how many carry no created date at all.
• On Pipedrive, how many Leads were set aside from the deal count, so the page reconciles against what you see in your tool.
• Campaign ids that could not be resolved to names, and what to connect to resolve them.
• Findings the field-intelligence engine raised on either object — a source field going stale, a stage with variant spellings, a placeholder close date, wins with no close date.
• Whether focus filters are set in Intelligence settings. Reports honour them; this audit does not, and describes the full synced store.

Recompute, and how fresh the figures are

The audit is worked out once per data change and stored, because reading every record of a large store takes tens of seconds, and far longer on a very large one. You are shown the stored audit. When it is not the latest, the page says so.

1. 1

   Check the “last synced” date under the page heading.

   Figures are worked out from what a sync brought in. Nothing changes between syncs unless you recompute.

2. 2

   Look for a grey line under the tabs.

   When the figures are an earlier computation the line says so, with the date: “These figures were computed on 6 Oct 2026, 09:30, before the latest sync.” No line means the figures are current. Export report, the Journey and Funnel tabs, the analyses, the Suggested Plan and Ask carry the same line.

3. 3

   Click Recompute to work the audit out again now.

   It reads the store as it is at that moment and shows the result when that fits in one request, which it does for most workspaces. Use it after a field-role change, or before a number goes into a pack — and afterwards check that the grey line is gone.

What the grey line tells you

On a very large workspace

An audit of hundreds of thousands of people does not fit in one request, so it is worked out a part at a time: at most 100,000 people to a read and 10,000 deals to a page, each part kept as it finishes. The scheduled refresh runs every hour and takes the workspaces in turn, so a large workspace gets a part each time its turn comes, not every hour. Three things follow.

• A first audit says “This audit is still being worked out”, with how many people have been read so far. The page asks again by itself; you can leave it and come back.
• Recompute cannot finish inside one request. The page says “it has been handed to the hourly refresh” and goes on showing the audit computed before, with its date. You have not been given fresh figures until the grey line is gone.
• What has been done is kept for as long as it is being added to. If nothing is added for 48 hours it is dropped and the next attempt starts from the store as it is.

Data Audit or Integration Audit Report — which to send a client

They are different reports on different data and they answer different questions. Sending the wrong one is the commonest mistake on this screen.

• Send the Data Audit export when the conversation is about numbers — pipeline, win rate, sources, conversion, and whether the data behind them holds up.
• Send the Integration Audit Report when the conversation is about the setup — field fill rates, forms, emails sending, automations running, and what a stale token or a missing permission is costing.
• Send both for a first-time discovery on a new client. Between them they cover the condition of the data and the condition of the plumbing.