In recent years, a consensus has emerged in TV advertising: outcomes measurement is essential. Today, everyone acknowledges that the point isn’t merely to buy cheap reach — you need to change consumer behavior.
But where in the funnel should we measure? And which metrics should we use to evaluate and optimize campaigns? These questions are still debated, even among practitioners on the forefront of advertising research.
Many advertisers try to measure at the bottom of the funnel, using sales attribution models that attempt to tie each consumer purchase to the specific ads responsible. Sounds good, but advertising can’t drive sales all on its own — it can only get a brand into the consideration set.
In the space between consideration and conversion, other variables take over as the dominant drivers of consumer behavior — including price, product features, online reviews, or plain old physical availability.
That’s why at EDO, we focus on measuring and optimizing the consumer behavior metrics that advertising directly influences and can actually be held accountable for. These are mid- and lower-funnel metrics like brand searches, website visits, store visits, and app downloads.
How sales attribution can lead TV marketers astray
While it’d be nice to tie every sale to an ad, our preference for mid- and low-funnel metrics has been borne out across many case studies looking for the influence of TV advertising at progressively deeper points in the funnel. At the very bottom of the funnel, advertising data is no longer decision-worthy for two reasons:
- The data gets thin. As you move from consideration to purchase, you’re able to measure far fewer outcomes. This creates a less reliable signal.
- The signal gets swamped by other factors. The causal impact of advertising just isn’t there in most places — we’ve looked. Conversion rates are typically more influenced by drivers like price, features, physical availability, and so on.
In most cases, advertising has no discernible impact on conversion rates, but there are a couple of notable exceptions. For instance, first-time TV advertisers often see a one-time boost in conversion rates after they go on air for the first time, since people perceive them as being more credible. Or when they place strong economic incentives to convert inside the ad, such as a rebate or discount.
But for the most part, measuring the impact of TV creative or media by bottom-of-funnel sales obscures the actual impact of an ad. For instance, retailers could be running ads that are highly effective — that is, at driving consumers into the consideration set. But at the moment of truth, consumers might not buy due to recent tariff-driven price increases. You wouldn’t want to abandon what’s working for you because you misattributed the impact of tariffs to advertising.
By analogy, ads that drive consumers into the consideration set are like quarterbacks that deliver accurate passes — setting the team up for touchdowns, even if they don't always cross the goal line themselves.
Mid- and lower-funnel measurement provides immediate, reliable, and actionable feedback on creative and media decisions
In contrast to bottom-of-funnel sales metrics, mid-funnel metrics like ad-driven searches deliver fast, immediate signals that tell us whether an ad worked. Measuring here provides faster, more reliable reads on campaign impact — without confounding factors that muddy the waters at the very bottom of the funnel.
These mid-funnel metrics are the perfect tools for assessing the impact of new ads, identifying high-performing creatives, and driving rotations for maximum impact. These metrics are fast, responsive, behavioral, and underpinned by massive amounts of search data that enable faster and more reliable decisions. In the emerging world of AI-accelerated creative development, optimization cycles will only become shorter.
And for more granular decisions around frequency capping, micro-segmenting, and DMA targeting, we use lower-funnel metrics — but again, not bottom-of-funnel. These lower-funnel metrics include website visits, app downloads, store visits, and others.
It’s all about having the right tool, at the right point of the funnel, for every decision. And it’s rarely the very bottom of the funnel.
But we know sales matter — here’s how we use them to calculate ROI
Of course, we know sales are the reason you run TV advertising — and essential for calculating your advertising ROI.
That’s why we use data-driven assumptions like conversion rate by channel and average annual customer revenue to compute the dollar value of moving a customer into the consideration set.
Take this example from an insurance company:
Let’s say we know the brand’s conversion rate from a quote to a new policy is 5%. And by reviewing the company’s sales data, we can learn that an average customer stays for around three years, with an average annual premium of $1,200. We can then multiply these numbers to get the expected dollar value of moving the customer to consideration.
5% conversation rate x $1,200 average annual premium x 3 year customer lifespan = $180 value for each incremental, ad-driven quote
You don’t need me to tell you how to calculate your ROI from there.
Stop optimizing for what advertising doesn’t control
As someone who’s worked in this business for over 30 years, I empathize with the need to compute return on investment and provide accountability for ad spend. In fact, I spent the early part of my career chasing the idea that I could connect advertising directly to individual sales.
But what I eventually learned, through many trials combining EDO and customer data, is that there’s a better way. When we use mid- and lower-funnel metrics — but not bottom! — we can understand the true, direct impact of TV advertising. By combining these signals with simple, data-driven economic assumptions, you’ll be able to connect advertising to sales and optimize for best results.