My 5 part measurement framework

Most measurement conversations happen too late, after the budget is gone and everyone is reverse-engineering what happened. The best measurement plans are made before a dollar is spent. Here is the broad, 5 step framework I use when thinking about measurement.

1. First, marketing teams need to clearly understand the top-line goals of the business.

This is usually a total revenue number. But if you're a smaller business or startup, it could also be a total number of accounts, or customers in a new vertical.

I know saying "marketing needs to understand the business goals" sounds...obvious. This is not revolutionary. But I've been in rooms where marketing is so far removed from the goals that they think their north star is optimizing a conversion rate, bounce rate, or "engagement" rate (and engagement can mean SO many different things). Yes, those numbers are important. But if a marketing team doesn't have a top-line goal, they'll never understand how to work back to reach it.

2. Define marketing's contribution to the top-line goal.

Backing up from that top-line goal, marketing should be able to build a forecasted plan of the influenced pipeline or revenue they'll contribute. The best way to do this is to use historical data from past channel performance (example: based on your site conversion and lead generation rate, how many people need to come to your site, how many people need to be advertised to or find you organically) to build a bottoms-up media plan. If there is no historical data, then you must rely on benchmarks from your industry or from companies similar in size.

A note on budget: sometimes the forecasted plan will be dictated entirely by budget. If you're operating on a smaller budget, or you're the underdog in a large, competitive industry, then there has to be some acknowledgement that you're not going to acquire customers for some insanely low price, especially if you're just starting out. This is also why I like to build media plans across two or three budget scenarios, so teams can understand what they "get" at different levels of investment.

3. Establish what "good" looks like at each step of the customer journey, channel by channel. These are your primary metrics.

No matter what kind of business you are, your customers almost certainly come to you through some kind of funnel, dropping off along the way as they lose interest, find another solution, or make no decision at all. Mapping that journey can get messy, especially in enterprise software or big-deal-size businesses where a prospect might visit the site seven times, catch an out-of-home ad, save a LinkedIn post, and then answer an inbound sales email. But modeling that experience is still the best way to know what your goals should be and what must happen at each step to hit them.

Once you've mapped it, you can find the metric that actually ties to the goal. For example, if based on the work from step 2, you know you need a 2% form-fill rate to hit revenue given how leads convert once they enter the sales process, then that 2% is the number you watch most closely and judge yourself on. That becomes a primary metric, and I anchor to that as a main measure of performance.

To set what "good" looks like for each, I use a mix of industry benchmarks, historical data, and my own read on the strategy (do I believe the creative and messaging are strong enough to perform above average? Almost always.). "Good" is different for every channel and every business model, and it's different for B2B versus B2C, which is the whole point. Primary metrics I've reported on include attach rate, conversion rate, LTV, CAC, and reach and frequency for awareness work, plus ecommerce numbers like AOV and ROAS. For a B2B parts manufacturer, the leading conversion was a sample request with a hard ROAS floor set monthly. For a healthcare client running programmatic awareness, "good" was measured against CPM benchmarks by service line with seasonal context built in. And for companies where I was judged purely on web performance, it was PDP conversion, bounce rate, scroll depth, and time on page.

4. But you still need to look at a lot of numbers beyond your primary metrics to understand what's going on. These can be secondary metrics.

If you only review a handful of primary metrics, you will miss the story.

For example, on its own – CTR can be a bit of a lame metric to report on, and one I'd hate being presented to me without context if I were a CMO. Clicks are not necessarily interest (I can't even begin to guess how many times I've accidentally clicked on something) and are not sales.

But, if you're seeing a high CTR to a (CTA Call-To-Action) button and no conversions, then that should tell you that something is broken on your web experience. People are getting confused once they get there, or you're targeting the wrong audience entirely. You will miss that story and won't make changes if you're only looking at primary metrics.

I once worked with a software company investing six figures in 6sense who were told their advertising was doing great. According to 6sense, their advertising was receiving millions of impressions and account engagement scores were climbing. But when I reviewed the funnel end-to-end, I saw an extremely low conversion rate, where very few of those ad impressions were leading to any site visits. I helped them reorganize 6sense to be used as a true ABM engine, and not as a DSP with a heavy skew toward banner and mobile ad placements. I also helped them cross reference intent data against Salesforce to understand which of these opportunities needed attention from sales.

I share this story as a reminder that metrics can look great on the surface. Engagements! Eyeballs! But you're always going to need to dig in further. Especially when things aren't working.

5. Watch out for measurement red flags.

I could write an entire post on this one. But here are a few to watch for.

Attribution: Get your attribution set up early, then actually pressure-test it. Pixels can be set up wrong so easily.

Beware of retargeting campaigns: TBH, I think retargeting campaigns can be misleading. Retargeting ads take credit for demand you already created, so a program can look wildly efficient while doing almost nothing net-new. If retargeting is carrying your numbers, make sure your reporting separates the audiences you are re-touching from the ones you are bringing in for the first time.

Anchoring on one metric too hard: I have watched companies get so locked on ROAS that they can never scale. Here is the trap: paid platforms will stop spending your budget the moment they cannot hit the goal you set, even when the ROI is clearly still in your favor. Sometimes the right move is to loosen the target on purpose so the platform can go find new customers.

Pivoting too quickly: This is the one I see most, especially on paid search. Algorithms need data before their numbers mean anything and reacting to early noise or making too many changes across too many campaigns/platforms usually makes things worse. Be patient, and stick to your strategy.