TTL and Recency, Frequency

[deepay216044]

Hi ,

Can someone help me with how the TTL of the traits and Recency and frequency settings in segments work together.

Say, I have T1 (TTL15 days), T2 (TTL 15 days).

S1 has rule -->  (freq(T1) >=1) <=15D) AND NOT (freq(T2) >=1)<=15D

UUID1 realises the Traits and segment in following fashion:

Trait/Segment	Day 1	Day  2	Day 3	Day 4	Day 5	Day 6	Day 7	Day 8	Day 9	Day 10	Day 11	Day 12	Day 13	Day 14	Day 15	Day 16	Day 17	Day 18	Day 19	Day 20	Day 21
T1	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	N	N	Y	Y	Y	Y
T2	Y	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	Y	Y	Y
S1	N	N	N	N	N	N	N	N	N	N	N	N	N	N	N	Y	Y	Y	N	N	N

T1 was realised on Day 1 abd T2 TTL last day was Day 1. On days 16 and 17 although trait 1 was not realised within last 15 days but since due to TTL UUID 1 was qualified in T1 just two days ago. So does, UUID qualify for segment 1?

On day 18 however, I understand why it would be a part of S1.

Any additional information would also be helpful besides this.

[Tof_Jossic]

deepay216044​ here is my take on your scenario:

If T1 was realised on Day 1 at 12noon, then I would expect the realisation to be valid for 15 days after that so up until Day 16 at 12noon.

In this case your user would be qualified for S1, having not been in T1 population since Day 1.

On Day 17, T1 drops and since T2 is still not realised again, S1 drops

On Day 18, T1 is being realised again and since T2 is still not realised again, S1 still stands

On Day 19, T1 is still in its TTL but T2 gets realised again so S1 drops.

[deepay216044]

Christophe_Jossic​

Thanks a lot for your response. But I am not 100% clear here.

Do you mean to say that UUID1 would be qualified for S1 from Day 2 till day 16 (12 noon). and then again qualify on day 17?

If this is the logic we use, then how did the "recency" value play a role? (because this solution would hold true for this scenario as well T1 OR T2, correct me if my understanding is wrong here)

And how does "frequency" get counted (is there a counter at the backend which increments with every new realization?

Thanks

[Tof_Jossic · Accepted solution]

deepay216044​ this is one of those difficult conversations around the topic and it can be very confusing but essentially your table is good.

- T1 realised every day from day 1 to day 15

- T2 realised on Days 1 (Say 12noon)

- On day 16, even though T1 has not been realised for the day, it was realised the previous day and has a TTL of 15 days so is still valid

- Still on day 16, up until 12noon, T2 is valid having a TTL of 15 days so no S1 qualification.

- Still on day 16 say at 3pm, if T2 has not been realised since day 1, the TTL of 15 days expires and the user drops from that trait population. T1 is still valid from its last realisation so at that point in time there is a qualification for S1

- Same for day 17, T1 is valid but not T2 so S1 qualifies

- On Day 18 you have a new T1 realisation which will only extend the TTL to an extra 15 days. T1 remains valid and T2 is still dropped so S1 qualifies.

Yes we have timestamps in our system for any trait realisation, including last realisation and it is what’s being used to count realisation frequency if that makes sense.

Hope this helps.

[Tof_Jossic]

Quick clarification that came to light on https://experienceleaguecommunities.adobe.com/t5/adobe-audience-manager-questions/recency-on-frequen...; segments will disqualify at the beginning of each day so hours and seconds are ignored. If you have recency of 1 day, and you make the event at 11pm, you'd effectively disqualify about an hour later.

[Tof_Jossic]

You can vote / like on corresponding idea https://experienceleaguecommunities.adobe.com/t5/adobe-audience-manager-ideas/aam-should-treat-recen...