Good MMM AI Prompts Come From Good MMM Analysts
Can AI build MMM models in mins?
Yes.
But would it be accurate and reliable?
Not all.
AI maturity still has a long way to get to that level of accuracy.
At Aryma Labs, the core nucleus ‘The MMM model’ is still built by us (humans). But we leverage AI innovatively to generate insights, disseminate insights, perform deterministic tasks like budget optimization.
Recently we demo-ed our products MMM Synapse, Nebula, Singularity and the yet to be publicly released ‘PPT add-in’ (this is gonna be real game changer) to a group of CMOs, Brand and Media managers. Stay tuned for this release.
One of the CMO was so impressed that he wanted to explore strategic investments in us !!
While we were flattered at the offer, he asked us what makes your AI solutions so accurate, so on point and so rich in insights?
Well the answer is – Our MMM Knowledge.
Yes, our Moat is not the LLM. Soon all LLMs will become a commodity.
In terms of AI products/solutions, what will separate good MMM vendors from mediocre ones is not which advance LLM they use, but on what data the intelligence layer was trained on and plus the prompt engineering skills.
LLM is only one part, the other part is ‘How well do you prompt?’
Consider the following scenarios:
• Sales and media spends trend together
• There are multiple overlapping promotions, strong seasonality
• Saturation curves demonstrating diminishing returns.
AI cannot magically know how to connect the dots and form a coherent accurate narrative.
These overarching rules must be explicitly told to AI in the form of Prompt templates.
For example:
“Check the budget optimization scenario and verify whether the spends are taken out of channel that was demonstrating diminishing returns”
As you can notice, the above is a MMM skill. There is a link between budget optimization and saturation curves (link in comments).
📌 Good MMM Prompts Come From Good MMM Analysts
At Aryma Labs, both Ridhima and I have taken a new role “Chief Prompt Supervisor”. Our excellent MMM analysts provide prompts grounded in MMM knowledge. We then supervise them and edit it to account for edge cases and additional logics.
The quality of an MMM AI system is not determined only by the LLM.
It is determined by the quality of the domain expertise embedded into the prompts, workflows, diagnostics and reasoning chains.
The better the MMM analyst behind the system, the better the AI performs.
📌 Good MMM Analysts – The force multiplier of AI
A lot of people think AI itself is the force multiplier. That is only partially true.
In MMM, AI becomes powerful only when guided by strong MMM analysts.
Because good MMM analysts define:
• The guardrails
• The edge cases
• The statistical checks
• The business constraints
• The causal assumptions
• The operational logic behind the model
In many ways, the first force multiplier is not AI.
It is the MMM analyst who shapes the intelligence behind the AI.
