Random Sample Less Than 1

[Helen_Abramova1]

I wonder if there is a way to get random sample less than 1%? It's not first time I need it, could n't find any way to figure it out rather than adding  additional filters. Are there any tricks?

UPD: Not ideal, but I added 1% to a list, than reference it in another smart list with random sample whatever I need.

Thanks,

Helen

[Grégoire_Miche2 · Accepted solution]

Yes, 50% of 1% will give you 0.5%. But as Sanford points out, the risk is to get 0 as a result. It really depends on the DB sample.

When I need 5 or 10 leads for a test, I browse the database and pick a few manually.

-Greg

[lsimon]

I am trying to send 1 alert to someone, when an email is cast from an EP. I thought I could use random sample to send an alert to someone via 1 person or .25 %, for persons receiving an email from a nurture campaign.

Otherwise, I need to find an exact email being sent that cast, and add that one person in the flow step to send the alert.

[Jo_Pitts1]

@lsimon ,

can you explain a little more clearly what you're trying to achieve?  I've read and re-read your comment, and can't quite fathom it (maybe I just need more coffee!)

Cheers

Jo

[Katja_Keesom]

That's always a good idea @Jo_Pitts1 !

[SanfordWhiteman]

Not entirely clear to me either, but it seems like the idea is to send an alert when the Campaign has gone out, i.e. the mythical concept of an program/campaign-level (as opposed to person-level) alert.

The closest you can get to that is creating a Resource Person — a person dedicated to a particular program/campaign, who is guaranteed to qualify — and then having a trigger constrained by that person. Certainly that makes more sense then trying to select one random actual person.

[Jo_Pitts1]

Maybe that's the case... but I still don't understand the use case... and I am properly caffeinated currently 🙂

[Darshil_Shah1]

I had to re-read the comment to get some idea around the exact use case. @lsimon talks about using 0.25% or 1 person which if I'm not wrong mean that the total population size of people who can receive the alert email is precisely 400. The random sample % would have to be changed every time there's change in the population size to ensure that only 1 person out of all the designated people receive the alert email. IMO, Sandy's idea  of doing this via a designated resource person per campaign/program is far better than going via the random sample approach!

[SanfordWhiteman]

You should be aware that random samples of 1-2% (let alone less) may not return ANY leads.

[Helen_Abramova1]

Well, it means I am lucky   It worked for me so far.

[SanfordWhiteman]

But you won't know when it stops working unless you're continually monitoring the size of the overall population.

[Helen_Abramova1]

Thank you, will keep it in mind. Usually, I need it for one list pull or batch - when I want my sample to be more or less fair representation of the DB.

[Josh_Hill13]

Why would you want such a small sample?

[Helen_Abramova1]

Hi Josh, I need it for testing purposes, data audit, pilot projects, etc.