Customer Loyalty

[darrynk23768843]

Does anyone use the "Customer Loyalty" reports, and are the numbers believable? I have downloaded some data around this, and found that the categories for new, return and loyal bear very little resemblance to the actual customers that we have.(when comparing number of transactions by individual customer vs categorisation in Adobe). Although the reports would be great to use, there is very little use in having them if they are vastly inaccurate. Also, what is the use of setting the "Unique Visitor Variable" if it does not track customer loyalty by that variable?

Has anyone found that this is better when using the new visitor id service?

Darryn

[TanmayMathur]

Hi Darryn,

Thanks for reaching out. 

I assume that you have gone through how the customer loyalty reports are populated, New, return and loyal customers are defined in the following way

             

New Customer	1 visit and 1 purchase
Return Customer	More than 1 visit and 2 purchases
Loyal Customer	More than 1 visit and 3+ purchases

This might be different from your business definition of customers. You may want to go through this documentation - https://marketing.adobe.com/resources/help/en_US/reference/reports_customer_loyalty.html

Feel free to ask if you are encountering any discrepancy.

Regards,

TM

[darrynk23768843]

Hi TM,

Yes, I have gone through those definitions, and that is what I am basing my comparisons against. For instance, as an example, I have one customer that was categorised as "new" on the 25th June, however, according to our records, they have made 40 purchases, evenly spread between December 2014 and June 25th 2015. The overall accuracy for the 25th June for "new" customers is 30%.

At the same time, I have "loyal" customers that haven't purchased yet, as well as customers that only have one or two purchases. This is at a much lower % discrepancy.(6% - still not acceptable), but overall, when combined with the inaccuracy of the "new" customer categorisation, it makes the reporting worthless.

Thanks,

Darryn

[TanmayMathur · Accepted solution]

Hi Darryn,

Customer loyalty reports are based on a unique visitor ID + Visit_num + Purchase event. 

So the problem may lie in any of those three determiners. We would have to confirm the 'purchase event' for the report is the same as the actual 'purchase' to make a comparison.

Regards,

TM

[darrynk23768843]

Ok, it looks like even though I have set up the "Unique Visitor Variable" within the Admin Tools, it is not using that variable. The attached customer is shown as new for 5 purchases on 5 different visits, so it is clearly not using the variable I set as the unique variable. Will the new visitor service improve this variance?

[img]CustClassifications.JPG[/img]

[darrynk23768843]

Actually, this doesn't seem to be working at all. If you look at the previous attachment, it shows that the visitor is classified as "New" even though they have more than 1 purchase.

Visit number 1 with 1 purchase = New -> correct

Visit number 2 with 1 purchase = New -> correct

Visit number 3 with 1 purchase = New -> incorrect, should be Return.

Visit number 5 with 1 purchase = New -> incorrect, should be Loyal

They then suddenly skip to Loyal in Visit number 6, and 7, and then suddenly for their next purchase in visit 13, they are New again. 

[TanmayMathur]

Hi Darryn,

According to the Loyalty reporting rules

Visit number 3 with 1 purchase = Not return as <2 purchases

Visit number 5 with 1 purchase = Not loyal as <3 purchases

Discrepancy should not happen ideally. I would request you to reach out to ClientCare right away so that they could look deeper into the issue and solve it.

Regards,

TM

[darrynk23768843]

Hi, those were individual visits and orders, not cumulative orders.

Visit number 3 with 1 purchase (3rd purchase in total) = Not return as <2 purchases --> (it is return, as they had made 2 purchases in the previous 2 visits).

Visit number 5 with 1 purchase (4th purchase in total) = Not loyal as <3 purchases --> (it is loyal, as they had made 3 purchases in the previous 4 visits).

I guess I will have to raise with ClientCare.

[TanmayMathur]

I see. In that case you should surely reach out to the ClientCare.

Regards,

TM