Great video! You mention that MMM is not as sensitive to regulatory changes, such as GDPR, as attribution model. I do not understand, however, how this is a pro. I believe that GDPR can have a strong influence on sales, for example, so shouldn't the model take those effects into consideration instead of dismissing them?
Exactly! I was quite confused by that too seeing as economic data (eg gdpr) is very useful in most models I’ve built. Sensitivity always depends on the depend variable in question and how it relates to gdpr for example. Maybe I’ve completely missed her point
@@Hitsquaduk they probably meant that it is an effect that despite being generally important it cannot be affected. But if the logic is still that of dependent/independent variables they should still explain how to deal with it
At first, she said they estimated an increase in SALES by 15% and at the end she said they increased STORE VISITORS by 17% and that it exceeded the estimated impact of 15%. That doesn't sound right.
it's 2 different KPIs, having an increase of visitors doesn't necessarly means +15% in sales or more, actually there's a good chance that it's less if you consider the conversion rate @@WeAreFunnel
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
Would it be possible to turn down or eliminate the background music?
Thanks for the feedback! We'll make sure to lower the sound in future videos 🙂
You should do a tool deep dive with a demo account or give users a dashboard with demo accounts
Can someone elaborate if MMM is a statistical model or just a marketing approach model? Do we use Machine learning models in MMM?
Great video! You mention that MMM is not as sensitive to regulatory changes, such as GDPR, as attribution model.
I do not understand, however, how this is a pro. I believe that GDPR can have a strong influence on sales, for example, so shouldn't the model take those effects into consideration instead of dismissing them?
Exactly! I was quite confused by that too seeing as economic data (eg gdpr) is very useful in most models I’ve built. Sensitivity always depends on the depend variable in question and how it relates to gdpr for example. Maybe I’ve completely missed her point
@@Hitsquaduk they probably meant that it is an effect that despite being generally important it cannot be affected. But if the logic is still that of dependent/independent variables they should still explain how to deal with it
@@frrrrrraagree. Confusing statement but useful content nonetheless
At first, she said they estimated an increase in SALES by 15% and at the end she said they increased STORE VISITORS by 17% and that it exceeded the estimated impact of 15%. That doesn't sound right.
How so?🙂
it's 2 different KPIs, having an increase of visitors doesn't necessarly means +15% in sales or more, actually there's a good chance that it's less if you consider the conversion rate
@@WeAreFunnel
This is SOLID 🎉
Glad you like it Zaryn! Thanks for the feedback 💥
Get your data right..4% increase from 41% to 41% channel spend?
honestly, this video is useless from a data scientist view...
why
?
Please remove the background music, its very disturbing.
11 mins advertising. Useless