diff --git a/docs/source/notebooks/mmm/mmm_counterfactuals.ipynb b/docs/source/notebooks/mmm/mmm_counterfactuals.ipynb
index db0bafb4..798d47b6 100644
--- a/docs/source/notebooks/mmm/mmm_counterfactuals.ipynb
+++ b/docs/source/notebooks/mmm/mmm_counterfactuals.ipynb
@@ -795,7 +795,11 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Because we are simulating this world, we know the true causal impact. The next step will be to train an MMM on the actual data and then use it to estimate the causal impact of the marketing campaign."
+    "Because we are simulating this world, we know the true causal impact. The next step will be to train an MMM on the actual data and then use it to estimate the causal impact of the marketing campaign.\n",
+    "\n",
+    ":::{note}\n",
+    "Notice in the bottom plot that the causal impact of the campaign continues on for a short duration after the campaign as ended. This is a direct result of the adstock function in the MMM which models the lagged effect of media spend on sales. Similarly, there is a slight lag effect at the start of the campaign. There is also noise in the causal impact - this makes sense because the way how we similated this data was based on changes in media spend, not on a specified level of sales difference.\n",
+    ":::"
    ]
   },
   {