Days since last visit metric

[keithyap]

There isn't any documentation on this out of the box metric, so thought I'd post here to see if anyone has had experience using it.

Basically, we have a custom evar that captures the unique ID of visitors when they login to our mobile app.

Days since last visit ("DSLV") against the unqiue id evar however, seems to work only in some instances.

Client care advised that we need to set our id evar to 'never expire' (currently set to expire on visit) and that this was the issue.

However, further investigation reveals this information to be incorrect.

In production (with id evar expires on visit), DSLV against the id evar seems to work in some cases, even if the visitor has not returned since the last visit.

In development (with id evar never expires), DSLV against the id evar does not work if the visitor has not returned since the last visit, which seems counterintuitive. It also only works in some cases otherwise.

Curious to know if anyone has used this metric with good accuracy against a unique visitor ID. Thanks!

[Gigazelle]

Days Since Last Visit is a dimension, not a metric. If a hit starts a new visit, it checks the timestamp of the last hit (if there is one), and buckets the visit based on the number of days in the past the previous hit was.

Are there any specific instances where you're seeing it work incorrectly?

[keithyap]

Hi Gigazelle,

You're right, it is a dimension. My mistake.

Interesting that you mention it will only check on a new visit. As mentioned above, in production at the moment, we have multiple customer ids that last visited say 9 days ago, and have not visited since, yet Days Since Last Visit returns 9.

I don't see any specific instances where the dimension is not working correctly. Some customer ids we check have it recorded correctly, and others don't. I can't paste screenshots on the forum, but have sent them to client care and am waiting for a reply.

[Gigazelle]

Yeah, for anything that requires looking at personal data, customer care is definitely the best way to go here. Let us know if anything develops from the customer care side of things!

[bombardino3]

Hi Gigazelle!

Could you help me with my question of the previous comment?

Thanks in advance for your help.

[mubarakd5726232]

Did you get a reply from customer care ?

The following suggests the value in the dimension is based on the most recent hit and the relative difference between that visit and the previous.

I would however like to report on the difference between current date (or last date in reporting timeframe) and the last hit/visit.  Is this possible ?

[bombardino3]

Hi!

I'm having some problems to understand what the Days Since Last Visit means. If you check the table below, why Unique Visitor sum of all rows exceeds the Unique Visitor total (143,561)?

Thanks in advance for your help.

Best regards.

[PeterWoodroffe]

Hi bombardino3,

Believe it's because a unique visitor can return to site multiple times and therefore will appear in mulitple days since last visit buckets

e.g. Visitor comes to site Mon, Wed, Thurs. Their visit on Weds (last visit Mon) means the visitor is in the 2 day bucket, their visit on Thurs (last visit Weds) means they are also in the 1 day bucket.

If you switch your metric to visits it will total 100%.

Thanks