How it works
We model how your money actually moves.
Parable fits a response curve to your store's history and estimates how profit may change at different ad budgets. It brings revenue, costs and uncertainty into the same view so you can choose a smaller next step to test.
A reported return does not set your next budget
Attribution reports assign credit for recorded sales. Reports from different platforms can overlap, and an average return at today's budget does not tell you what the next dollar will earn. A spending decision also needs your margins, costs and an estimate of how returns may change.
Read the guide to choosing your next ad budget →Marketing mix modelling (MMM)
Instead of following individual clicks, this family of methods looks at your whole business over time (revenue, ad spend, seasonality) and uses statistics to estimate the relationship between spending and outcomes. Parable applies response modelling to a single Shopify store, using aggregated history rather than tracking visitors on your storefront. A fitted relationship alone does not prove that changing a budget will cause the predicted result.
Why it's genuinely hard
This is a deep field for a reason. Four problems have to be solved at once:
Long buying windows
A sale can follow an ad days or weeks later. Parable's weekly model allows for delayed response, but the history cannot identify every customer's buying journey or prove which ad caused a sale.
Saturation (diminishing returns)
Extra ad spend can return less revenue as you reach more of the same audience. A fitted response curve estimates how returns change with spend, then compares revenue with costs to estimate where profit may peak.
Baseline separation
Some sales may happen without the current ads, through repeat buyers, brand awareness or word of mouth. Separating that baseline from ad response is an estimate, and uncertainty in it affects the curve.
Uncertainty, on purpose
History is noisy and some budgets have never been tested. Parable shows an estimated range and flags limited support, so a peak outside your observed spend is a question to test rather than a confirmed outcome.
What's under the hood
Parable combines your Shopify history with the advertising and analytics sources you connect. It estimates a baseline and seasonal demand, fits an ad-response curve, and applies your costs and margin to compare modelled profit at different spend levels. The available history and source coverage affect how much confidence to place in the estimate. Connect the ad platforms you actually use and check your costs before interpreting the curve.
We'd rather be honest than impressive
Fitting past results is not the same as validating a future budget change. Use the estimate to choose a smaller test, allow for delayed purchases and seasonal changes, and compare the outcome with the prediction. Parable stays read-only. You decide whether to change a campaign or budget.
Explore an example store