Audiences
16 Jun – 13 Sep · vs previous period: 18 Mar – 15 Jun
Who the money is being shown to. Two numbers decide everything on this page: how many people a list holds, which is what makes frequency real, and when it last took anybody new in.
- People on the smallest list
- 3,850
- —
- Retargeting · cart, 7 days, and it earns more per lira than anything else in the account
- Highest seven-day frequency
- 4.64
- —
- On Retargeting · cart, 7 days. The line is 4.
- Spend on lists that have not moved
- ₺130K
- —
- Money going to people chosen from a file nobody has re-uploaded
- Lists past the line
- 3
- —
- A list built from a file is called stale after 60 days
Every list the money is pointed at
Spend and return follow the period above. How big a list is and when it last took anybody new in are true now, not over the period — a list does not have a history the way a campaign does.
| List | People | Last took anybody new | Frequency, 7 days | Spend | Return |
|---|---|---|---|---|---|
| Broad · 25–54, coffeeBroadMetaNo list. Age, country and the platform's own targeting.Subscription · always on · Cold · brewing interest (₺55K) · Back to work · coupon · Broad · 25–54 (₺12K) · Autumn Blend · launch · Broad · 25–54 (₺5,866) · and 2 moreBig enough that frequency is never the problem here. When this one tires it is the work, not the list. | 2,400,000 | Rebuilds itself daily | under 0.1× · room left | ₺77K | not countedMixed targeting: half the money on it was never asked for an order |
| Retargeting · cart, 7 daysRetargetingMetaAdded to a basket on the web store and did not check out.Subscription · always on · Retargeting · cart, 7 days (₺38K)The smallest list in the account and the one that earns most per lira. It is also the one that hits a frequency of four in under a fortnight. | 3,850 | Rebuilds itself daily | 4.6× · at its ceiling | ₺38K | 2.71× |
| Retargeting · 30-day video viewersRetargetingMetaAnyone who watched three seconds of a reel in the last thirty days.Autumn Blend · launch · Retargeting · 30-day viewers (₺3,251) | 41,200 | Rebuilds itself daily | 0.6× · room left | ₺3,251 | 3.97× |
| Lookalike · 2% of purchasersLookalikeMetaHas not movedBuilt by the platform from the customer file below.Autumn Blend · launch · Lookalike · 2% of purchasers (₺2,871)Rebuilt from a file uploaded in June. It is modelling who bought in the first half of the year, which was a different half of the year. | 1,100,000 | 3 months agouploaded 2 Jun | under 0.1× · room left | ₺2,871 | 1.65× |
| Customer list · purchasers to MayCustomer listMetaHas not movedUploaded from the order book, hashed.Nothing runs against it. It is the file the lookalike above was built from.Three and a half months old. Everyone who bought since is missing from it, and everyone on it who has bought twice since is being advertised to as though they had not. | 6,42068% matched | 3 months agouploaded 2 Jun | Nothing runs on it | — | — |
Not lists at all These three carry the most money in the account and none of them is a set of people. Two are sets of words somebody typed, and the third is a file Performance Max was handed and will not discuss. There is no size to divide by, so there is no frequency, and this page prints neither rather than inventing one. | |||||
| Performance Max · signalsLookalikeGoogleHas not movedThe catalogue feed plus the customer list, as signals rather than targeting.Performance Max · catalogue · Asset group · all beans (₺87K) · Performance Max · catalogue · Asset group · equipment (₺41K)The signal list is the same June file. Performance Max will not say what it did with it. | not a list | 3 months agouploaded 2 Jun | Not a list — no size to divide by | ₺127K | 5.58× |
| Non-brand termsSearch intentGooglePeople typing what they want rather than who they want it from.Search · non-brand · Specialty coffee beans (₺36K) · Search · non-brand · Pour-over kit (₺27K) · Search · non-brand · Decaf (₺19K) | not a list | Rebuilds itself daily | Not a list — no size to divide by | ₺81K | 2.65× |
| Brand termsSearch intentGooglePeople typing the shop's name.Search · brand · Brand · exact (₺33K) · Search · brand · Brand · with a product (₺18K)Not a list and not a size. It is a set of words, and the question it raises is on the campaign, not here. | not a list | Rebuilds itself daily | Not a list — no size to divide by | ₺51K | 10.86× |
The two hairlines are 2.5 and 4. Past the first, the same people are starting to see it more than once a week; past the second, no amount of new creative helps, because the list has run out of people rather than the picture having run out of interest.
Frequency is impressions over the people the list holds, across the last seven finished days. Where a list is shared with a campaign running on a traffic objective the return is left blank rather than printed, because half of the money on it was never asked to produce an order.
What Omni would do about it
Written by the list check, from the age of the file rather than from anybody's judgement. Nothing is uploaded anywhere until a person says so.
- audience
Rebuild · Lookalike · 2% of purchasers
Uploaded 2 Jun → Re-uploaded from the order book, and set to refresh monthly
The file it is modelled on is 104 days old, so it is looking for people who resemble the first half of the year rather than this one.
- audience
Rebuild · Performance Max · signals
Uploaded 2 Jun → Re-uploaded from the order book, and set to refresh monthly
The signal file is 104 days old. Performance Max is being told what a customer looks like by everyone who bought before June, and it will not say what it did with that.