Missing Scenario Hashes when do testing of Recommendation Scenario in SAP Marketing Cloud

[valentinamed]

Hi,

Does somebody know why I get empty ScenarioHashes as a response when I want to test recommendation scenario (using postman) form SAP Marketing Cloud? A got back recommended products and calculations for them but not scenario hashes and because of that recommended products are not displayed on the store front.

What can cause?

[valentinamed]

Hi,

Yes, but then you can't assign Recently_Viewed_Algorithm which is a non-optimized type and therefore you can't assign it to optimized one.

I wanted that recommendation scenario changes every time with a new user signin, but that is only possible only with non-optimized algorithm and scenario but then I don't have scenario hashes for storefront which hybris commerce requires.

[valentinamed]

The solution was to implement scenario hash manually on hybris storefront for non-optimized scenario.That is the only way to allow recommendations based on non-optimized scenario working in SAP Marketing Cloud.

This process for the optimized scenario is not necessary, because the optimized scenario already generates scenario hash by itself.

KR,

Valentina

[SauMaris]

Is your scenario "optimized". Per my understanding hashes are for cached recommendation results and thus not generated, if the scenario you use is not "optimized".

[valentinamed]

Yes, but how to do that? To manually insert scenario hash for the non-optimized scenario on the commerce side? Did you mean that?

KR,

Valentina

[SauMaris]

You create a new scenario, which is optimized, create new recommendation model type and recommendation model with a reference to your optimized scenario. Use same settings that are in the original non-optimized recommendation scenario, model type and model.

I would try that in your place.

[valentinamed]

Hi Maris,

You are right, I'm using non-optimized scenario. How then to represent products on storefront because the algorithm that I'm using is RECENTLY VIEWED PRODUCTS algorithm which is non-optimized one.

Do you know what is alternative to this algorithm in optimized scenario?

KR,

Valentina

[SauMaris]

I don't have a definite answer, but maybe try to create your own by copying existing one and making the scenario optimized?