Introduction to Brand Impact Measurement

Introducing Brand Impact Measurement:

You know brand builds future growth. But without a number to prove it, you're defending spend in budget meetings instead of scaling the channels that are building it.

Brand Impact is Fospha's brand measurement product. It measures the causal relationship between your awareness spend and long-term business value, so you can show your CFO a channel is working before the revenue arrives.

The problem Brand Impact solves

1. Brand spend is judged by rules that were never designed for it

Performance marketing targets people ready to buy now. Brand marketing builds demand in everyone else, the people who aren't ready yet but will be in the future. The problem is that by the time brand spend turns into a sale, it's the lower funnel channel that gets the credit. Brand ends up being judged on short-term numbers it was never designed to move.

2. Brand channels get cut because you cannot make the connection between spend and revenue

Revenue often rises during a brand campaign. But without a way to prove that link, the credit goes to lower funnel channels instead, and brand budgets get squeezed. This means channels get cut even when they're working, simply because nobody could point to the evidence.

3. Without brand investment, you end up marketing to a shrinking pool

Performance marketing converts people who are already interested. If nothing is bringing new people into that group, you're targeting the same audience over and over. That audience doesn't grow, and every new customer becomes more expensive to acquire.

Where Brand Impact sits alongside Fospha Core

How the model works

Brand Impact uses a Bayesian causal graph model, trained on roughly two years of a client's historical data from Google Analytics (GA4) and their ad platforms. The key word is causal, not correlational: it estimates the probabilistic causal relationship between brand spend and leading indicators, accounting for everything else moving at the same time.

The inputs: two years of your own data

Trained on your history, not a generic industry curve, so it reflects your real seasonality, mix and trends.

Why two years?

• Learns your pattern. Enough history to learn your brand's specific seasonal pattern and longer-term trend, not a generic assumption.
• Stable, not noisy. Because Brand Impact works weekly, two years gives a broad view of upper-funnel investment, closer to a media mix model than a daily read, so the picture doesn't swing on week-to-week noise.

The leading indicators: the signals we chose

Indicators with a statistically validated causal link to downstream conversions, moving six to ten or more weeks ahead of revenue, enough lead time to act while it still matters.

That lead time is what makes them useful: they tell you whether a campaign is working while there's still time to act on it, not months later when the budget review has already happened.

The link: how channels connect to outcomes

Brand Impact shows what your spend actually caused, not just what happened around it. Impressions build in a curve, not a straight line: your first one works harder than your ten-thousandth, so returns naturally level off. That's expected, not a sign a channel has stopped working. The model also separates your spend's true effect from other things happening at the same time, like seasonality or market trends.

Adstock: marketing that takes time to land

Brand marketing rarely pays off the moment an ad runs. Brand Impact learns two things about every channel: when its effect peaks, and how spread out that effect is. A narrow spread spikes fast and fades fast. A wide spread builds slowly and fades slowly too.

An impression this week can keep driving visits and conversions for weeks afterward, and importantly, the effect usually isn't strongest the moment the ad runs. Brand Impact captures this with delayed adstock, a carry-over curve shaped like a gentle hill over time: the impact builds to a peak some weeks after the impression, then fades. The model learns two things per channel from the data:

• When the effect peaks: the delay to its strongest impact (anywhere within a roughly 6-month, 26-week window)
• How spread out it is: how many weeks the effect lingers around that peak, a wide gentle hill vs. a sharp spike

Seasonality: isolating the baseline

Demand moves for reasons that have nothing to do with marketing: holidays and seasonal trends. Brand Impact accounts for this directly with monthly seasonal effects learned from each client's own historical data, credited to a baseline rate of sales. The baseline reflects what a typical month, like December, looks like, so a seasonal spike isn't automatically credited to your ad spend. Credit only goes to a channel when the model estimates that channel actually contributed to the result.

The output: The Brand Impact Ratio (BIR) 

The BIR answers a single question: is this channel building brand more or less efficiently than its share of spend would suggest?

Across 138 clients, 50 markets, and 36 channel segments, awareness channels consistently achieve a median BIR above 1.5x, while conversion channels typically score below 1.0x despite receiving the majority of budget at most brands.

It's a number a brand manager can read weekly, a CMO can put in a QBR, and a CFO can use to make a budget decision.

Core use cases of Brand Impact Measurement for all teams