The Triangulation Fractals in MMM

Triangulation as a pattern keeps recurring in MMM

There is a lot of talk about Triangulation in the MMM space. Many vendors have some variant of this triangulation feature.

At Aryma Labs, we believe the true marketing effectiveness or ROI can be discerned as a result of MMM, Experimentation and Causal inference.

But there is also a fractal pattern.

If you are wondering what is a fractal ?

A fractal is a geometric pattern that keeps repeating itself. This is a very layman explanation. For more rigorous explanations check out this link

But here is a simple illustration of a fractal

Triangulation in MMM

The triangulation as a method keeps repeating itself in the whole MMM project lifecycle.

When we build a MMM model, at the EDA phase, certain insights are unraveled. For e.g. a particular campaign shows strong effect on the sales.

But this could be just correlational in nature.

To validate this further, we build the MMM model.

The MMM model may further provide evidence of this particular campaign driving sales.

So now we have two sources vindicating us.

But this won't be enough. We would need to see if these finding makes domain/business sense also.

For e.g. One would ideally want to check whether the campaign in question can indeed really affect sales given the duration of its running and spend share it had.

🎯 If all the three (EDA, Model and Domain Experience) validate the hypothesis that 'campaign positively affects sales', then I think we are on a stronger footing than just either one of them vindicating the hypothesis.

This fractal nature of triangulation is also applicable to how one should validate Experimentation and causal experiments (more on this in future posts).

But overall we find it fascinating that triangulation has a fractal nature to it in MMM.

---

Thanks for reading.

For consulting and help with MMM implementation, Click here

Stay tuned for more articles on MMM.