Different tools count different things over different windows using different rules. The discrepancy is usually expected rather than an error, and the resolution is deciding which number you manage by.

Why the numbers differ

Not because one is broken. Because they are measuring different things and were built for different purposes.

Different definitions. One tool counts a form load, another counts a submission, another counts a submission that produced a reply.

Different attribution windows. An ad platform credits a conversion to a click that happened days or weeks earlier; an analytics tool may credit the last source before the visit.

Different attribution rules. One gives full credit to the last source, another to the first, another divides it.

Different counting. One counts every conversion, another counts one per visitor per period.

Different collection. Blocked scripts, consent settings, and privacy protections affect tools unequally.

Which means two tools reporting different figures for the same month is the expected outcome rather than a fault to be diagnosed.

The specific mismatches

The ad platform reports more than analytics

Usually attribution windows and view-based credit. The platform counts a conversion it believes it influenced days earlier; analytics counts the source of the visit in which it happened.

This is the most common discrepancy and it is not an error in either.

Analytics reports more than your own records

Usually definition. A tracked conversion may be counting anything the tag fires on, including duplicate submissions, test entries, and spam.

This is the discrepancy worth investigating, because it usually means the tracking is measuring something other than enquiries.

Your records report more than analytics

Usually collection. Phone calls, walk-ins, and referrals never appear in analytics, and blocked scripts remove some of what would.

Two reports from the same tool disagree

Usually date ranges, time zones, filters, or one report applying a segment the other does not. Worth checking those before assuming anything deeper.

The one number to manage by

The decision that resolves this practically.

Pick your own record of enquiries and booked jobs as the figure that matters, because it is the one with a definition you control and it is the one connected to revenue.

Then treat the tools as directional: useful for comparing periods, comparing channels, and spotting movement, rather than as counts to reconcile.

A business that spends its monthly review reconciling three tools has spent the time available for deciding anything.

Making them comparable

Where reconciliation is genuinely needed.

  1. Match the date range exactly, including the time zone each tool uses.
  2. Match the attribution window where the platform allows it.
  3. Check what each is actually counting, which is a settings question rather than a guess.
  4. Exclude your own activity in every tool, not just one.
  5. Compare trends rather than totals, since two tools moving together is more informative than either figure.

The last point is the practical resolution. If both tools show enquiries rising, they rose, regardless of whether one says forty and the other sixty.

When a difference does indicate a fault

Since not every discrepancy is expected.

A tool reporting zero where another reports activity usually means broken tracking rather than a definitional difference.

A sudden change in the gap between two tools, where it was previously stable, indicates something changed: a tag removed during an update, a consent setting altered, or a form modified.

Figures that are wildly implausible against your own experience, such as conversions several times your actual enquiry count, indicate the tracking is firing on the wrong thing.

Those are worth investigating. A steady difference of a reasonable proportion is not.

What to do with the discrepancy

Record it rather than resolve it.

Noting that the ad platform typically reports a certain proportion above your own count means next month's figures can be read against that expectation.

That converts an unexplained gap into a known relationship, and a change in the relationship becomes the signal rather than the gap itself.

Most businesses never do this, and consequently re-investigate the same discrepancy every quarter.

The point worth holding

Precision here is not available and is not needed.

The decisions these numbers inform are coarse: whether to continue a channel, whether to increase a budget, whether something changed. Those do not require agreement to the unit.

What they require is a consistent measure over time and a definition you understand, which is your own record of enquiries and jobs, kept alongside whatever the tools report, as described in reading a report somebody else prepared.

Frequently asked questions

Why do tools report different numbers?

Different definitions, attribution windows, attribution rules, counting methods, and collection. Two tools disagreeing is the expected outcome rather than a fault.

Why does the ad platform report more than analytics?

Attribution windows and view-based credit. The platform counts a conversion it believes it influenced days earlier; analytics counts the source of that visit.

Which discrepancy is worth investigating?

Analytics reporting more than your own records, since it usually means the tag is firing on something other than genuine enquiries.

Which number should I manage by?

Your own record of enquiries and booked jobs, because you control the definition and it connects to revenue. Treat the tools as directional.

When does a difference indicate a fault?

A tool reporting zero, a sudden change in a previously stable gap, or figures wildly implausible against your own experience.

What should I do with a steady discrepancy?

Record it as a known relationship, so a change in the relationship becomes the signal rather than the gap being re-investigated every quarter.

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.

Three tools and three different numbers?

We work out which is counting what and pick the one worth managing by, which is usually your own.

Start a Conversation