Why is revenue or number of orders lower in a Visitor Data Mart profile compared to a similarly configured Analytics profile?


Webtrends Visitor Data Mart


Why is revenue or number of orders lower in a Visitor Data Mart profile compared to a similarly configured Analytics profile?


Analytics is much more forgiving with respects to the format for a purchase hit. The most common cause of this issue is the result of one or more parameters containing multiple values and the associated parameters they correlate with do not have a matching number of values. For instance, if there are three values (separated by semi-colon) for WT.pn_sku, there MUST be EXACTLY three values (separate by semi-colon) in both the WT.tx_u and WT.tx_s parameters. Any deviation from this will result in the hit not being collected for the purchase event in VDM. Analytics, however, will record the purchase and associated revenue.
Note that when this occurs in Analytics, bad data will still result. Totals will be correct, assuming the appropriate totals were calculated in the application that recorded the unit and revenue values, except where there are less SKU values than units or revenue. If the appropriate number of SKUs is recorded, but only one revenue value, all of the revenue will be attributed to the first SKU. The additional SKUs will show in the Product SKUs report, but could potentially have zero revenue.

Additionally, differences in session tracking may allow a hit to be tracked in Analytics, but not show up in VDM.

Analytics also utilizes smart trimming, whereas VDM does not. VDM also has very particular event retention settings that may remove data that is still visible in Analytics as well.

Analytics has fallback sessionization (thirdparty cookie -> first party cookie -> IP/user agent), VDM does not. It?s third party cookie -> first party cookie,only. There is no capacity for IP/user agent. As a result, VDM will differ from Analytics in the visits and/or visitors by at least the rate identified in Analytics cookie rejection rate reports (with Analytics being higher).

Analytics de-dupes based on WT.tx_i for only afew days. For example, if a WT.tx_i value comes in on one day, and the same ID comes back two weeks later, Analytics counts it as a new record. In VDM, all invoice IDs are preserved for all time. Once a particular WT.tx_i is loaded,all server calls that include this value in the future are thrown out for all time. As a result, VDM will differ from Analytics in that these?duplicate? invoices are being thrown out, and Analytics may be including them(with Analytics numbers being higher due to the over counts).

Analytics breaks sessions over a time period boundary (midnight), VDM does not. As long as continued activity is still coming in, VDM will keep the session open forever. Result is that for traffic covering a date boundary, Analytics will see multiple sessions when VDM does not. So VDM will differ from Analytics in having better tracking of these sessions (with Analytics numbers being higher due to overcounts).

Caution must be taken in configuration of the profiles. ?Global? Analytics hit filters do not apply to VDM, and there is no facility in VDM for a visit filter of any kind.

In general, since these are two separate products, with different storage and aggregation methods, it is unlikely that the numbers will match across the two platforms unless you are very careful with the configurations.