Modules · Meeton Ads

Results & proof

Every number on the Ads tab belongs to a week, credits one journey, and counts a person once. This page explains what each figure means, how the week window works, and how the holdout answers the one question the other numbers cannot: did the ads cause the leads?

The week window

Showing	What the numbers cover
This week · so far	The current ISO week from Monday 00:00 UTC to now. Grey figures beside each number are all of last week, so a part-week is not compared like-for-like with a whole one.
Last week · complete	The last full week. Changes are signed deltas against the week before, also complete. Counts move by an amount; rates move by percentage points.
All time · since launch	Totals since each journey went live, with no week-on-week comparison. A placement added later has been counting for less time than the rest.

Per placement

Expand a step on the path to see each placement's numbers for the window.

Figure	Meaning
Reached	Unique visitors who saw an ad on this page, counted once each, the first time a placement here showed them something.
Impressions	Times an ad was shown, including repeat views by the same visitor.
Clicked	Visitors who clicked the ad's button.
Closed in under a few seconds	Visitors who dismissed the ad almost immediately: the quick-close signal.
Leads · Meetings	Visitors who became a Meeton-assisted lead, or booked a meeting, after this placement showed them an ad.

Rates start once 20 visitors have been reached; below that there is too little to divide by. Variants inside a placement report separately, which is what makes them an A/B test.

Per journey

Leads are Meeton-assisted only: someone who chatted, booked or downloaded after the journey showed them an ad. A visitor who only filled in one of your own site forms is not counted here. Click the count to see who they are, and the assists toggle shows what Meeton did for each: chat, calendar, on-site ads, the Library.
Meetings are bookings made while on the journey, after an ad.
A conversion credits only the journey the visitor was on, so a booking made elsewhere does not land here, and a journey's leads can never exceed its reach.
Leads are counted from the moment a journey first shows an ad, so a path armed today starts at zero, and a journey older than the counter reads low. A muted +n? beside a count marks leads credited before Meeton recorded who converted.

Desktop vs mobile

Visitors and journey leads side by side by device, so a mobile experience that loses people shows up as a gap. The one number to compare across the rows is leads per 1,000 visitors: the same device is on both sides of it. Device is read from the browser of each person's latest visit, and phones and tablets count as mobile. The card is a comparison only; it cannot say why a device converts less, but it says where to look first.

Proof: ads against no ads

Term

A holdout group and lift are standard practice in growth, marketing science and incrementality testing: a slice of people is deliberately not exposed so the effect of exposure can be measured against them. Ad platforms run the same kind of test under names like conversion lift or geo holdout.

The holdout compares the lead rate of visitors who saw ads with those who saw nothing.

Everything above counts leads that followed an ad. It cannot say whether the ad caused them. The holdout can: a slice of visitors sees no journey, no popup and no chat ad, and the lead rate with ads is measured against the rate without, counting every lead in both groups, Meeton-assisted or not.

The dial

How many visitors see no ads at all: none, or a share such as 10%. Every option states its cost in your own weekly numbers, roughly how many visitors and leads will not be asked. A smaller share takes about twice as long to a confident answer.

Week boundary

A change starts on the next Monday, never mid-week. Re-splitting visitors part-way through would mix two settings into one week's numbers and neither half would mean anything. The running week keeps its current share until then.

Time to an answer

A confident answer usually takes four to six weeks of completed weeks. The week in progress is counted once it closes on Sunday.

Status	What it means
No holdout	Every visitor sees ads, so there is no group to compare with. Nothing here can say whether the ads caused the leads.
No completed weeks	A holdout is running, but no week has finished yet.
Still collecting	Not conclusive yet: the two rates are close enough that this could still be chance.
Confident	Roughly n more leads over these weeks than the held-out visitors' rate predicts. Only a result that survived the significance test gets this sentence.
Confident, and bad	Roughly n fewer leads than showing nothing would have produced. The ads are losing to no ads; look at quick closes and the loudest placements first.

The card sits at the foot of the Ads page and starts collapsed, with the lift and its confidence in the summary line so you never have to open it to know whether it has something to say. Expand it for the week-by-week table: visitors, leads and rate for both arms, and the lift.

Where else the numbers appear

Contacts and Visitors show each ad impression and click as a journey event on the timeline, with the journey, step and creative named.
Performance → Marketing puts ad-sourced leads beside the other channels.
Conversions under Data & tracking can fire an ad-platform conversion event when an ad produces a lead. See CRM, notifications & team.