Direction, Then Diagnosis: The Marketing Metrics That Actually Matter

Every platform wants you to believe you need another dashboard.

Google Analytics has hundreds of metrics. Your social platforms track reach, impressions, watch time, engagement rate. Your CRM has its own reports. Your ad platform has another set of charts again. Pull them all up at once and you’re staring at a wall of numbers, no closer to answering the one question that matters:

Is the business moving in the right direction?

The problem isn’t a lack of information. It’s that it’s easy to find yourself getting sucked into monitoring a sea of reports instead of the outcomes that actually move the needle.

The only funnel that matters

At MV Studio, we tell clients to watch a simple growth funnel:

Marketing Spend → Website Visitors → Leads → Customers → Revenue

Think of it as a machine: put a dollar in one end, get more dollars out the other. Spend and Revenue are the two ends of that machine, what goes in and what comes out. Visitors, Leads, and Customers are what’s happening inside it, the moving parts converting one into the other.

That order isn’t arbitrary. It runs from earliest signal to latest confirmation, Visitors moves first and fastest, Revenue moves last and slowest, often weeks or months after whatever actually caused it. By the time Revenue confirms a problem, you’ve lost whatever lead time you’d have had by catching it at an earlier stage instead.

Alongside it, one efficiency metric: Customer Acquisition Cost (CAC), total marketing and sales spend over a given period, divided by new customers in that period. CAC isn’t a stage in that sequence, it’s the machine’s exchange rate, how much you’re paying at one end for what comes out the other, and whether that rate is one worth running.

That’s it. Six numbers.

They tell you almost everything about the health of your marketing and sales: whether you’re attracting the right audience, turning anonymous interest into someone you can actually identify, an enquiry, a sign-up, a saved item, converting that lead into a customer, and growing profitably while you do it.

Everything else is secondary. Not useless, just secondary.

Where vanity metrics creep in

Followers, likes, reach, impressions, page views, video views: all interesting numbers. None of them tell you whether the business is growing.

A Facebook post reaching 20,000 people that generates zero leads hasn’t created value. It’s created a number.

Website traffic doubling while enquiries stay flat isn’t progress. It’s a wider funnel with the same leak.

Vanity metrics feel like momentum because the numbers get bigger. Revenue doesn’t care how many people liked a post. It cares how many people moved through the funnel and came out the other side as customers.

The dozens of other reports sitting in your analytics tools aren’t worthless, they’re diagnostic. Their job is to explain why something changed, not to tell you whether something changed. It’s the difference between a dashboard warning light and a mechanic’s diagnostic scanner: the light tells you something’s wrong, the scanner tells you which sensor’s at fault. You don’t start with the scanner, you start with the light, and only reach for the scanner once it’s on. Confusing the two is how a business ends up with twelve dashboards and no clear read on how it’s doing.

Reading the funnel

The advantage of this approach is that problems get easier to find.

Visitors falling? You’ve got a traffic problem. SEO rankings have slipped, ad performance has dropped, or referral traffic has dried up.

Visitors up, leads flat? There’s a disconnect between what’s attracting visitors and what they find when they land. Either the traffic isn’t the right audience to begin with, or the messaging, offer, call to action, or landing page experience doesn’t follow through on what brought them there.

Leads up, customers flat? The bottleneck’s in sales. Lead quality, response times, follow-up systems, and the actual sales conversations all become worth investigating.

Customers growing, revenue stagnant? More customers should mean more revenue, so when it doesn’t, something’s eating the difference. Average order value might be sliding, discounting could be doing more of the work than it should, or you’re winning customers who spend less and don’t come back. This is usually a pricing, upsell, or customer value problem, and it’s worth digging into properly, revenue is the number everything else in the funnel exists to produce.

Revenue rising, CAC rising faster? Growth looks healthy on the surface while profitability quietly erodes underneath. Check which channel is driving the increase, cheap channels get expensive fast: more advertisers pile in, bid against each other, and drive the price up. Andrew Chen calls this the Law of Shitty Clickthroughs, it’s why a genuinely new channel, AI search is the current example, is worth watching while it’s still uncrowded. But chasing cheap channels only helps if what they buy sticks, and that depends entirely on which part of the channel you’re using. A TikTok Shop sale might hand you a name and address for shipping, but not permission to use it, or under Fulfilled by TikTok, TikTok’s own warehouse handles delivery and you may never see the customer’s details at all, either way, that’s a rented transaction, not a lead. TikTok’s ad platform is the opposite: it’s built to send clicks to a page you own, with your own tracking installed, which is what turns a cheap click into something that compounds. Same platform, two different products, worth knowing the difference before assuming the channel is either a trap or a win.

You don’t start by digging through a dozen reports hunting for answers. You start with six numbers. Only when one of them moves unexpectedly do you go looking for why.

What diagnosis looks like

Here’s an example: four things that never touch the six headline numbers. Some explain a move that’s already happened, others surface before the top-line number moves at all, starting with this one.

Channel mix. Traffic was steady, the headline number hadn’t moved, but the breakdown underneath it wasn’t: a new referral source, AI assistants sending visitors directly to product pages, had appeared and was growing month on month. Nothing in “Visitors” flags that on its own. You only see it by breaking the number down.

Search visibility. One author page was earning tens of thousands of search impressions, real demand, at a solid position, but clicks stayed low despite that volume. Position was actually fine, it was the page’s own title tag working against it, a plain “Title - Author” pattern gives a searcher no real reason to pick it over anything else on the page. That title tag isn’t the same thing as the page’s actual heading, it can be rewritten to be more enticing without touching what a visitor sees once they land, the actual fix here.

Content trending up or down. Total traffic can hold steady while individual pages move in opposite directions underneath it, one title gaining clicks fast while others quietly lose them. Watching each page’s trend, not just where it sits today, catches that movement early, well before it’s big enough to shift the total, and points to specific content worth reinforcing or worth investigating.

On-site search behaviour. What people search for once they’re already on the site, and what returns zero results, is one of the most direct signals available: visitors telling you exactly what they wanted and didn’t find. That’s a content or catalogue gap, not a traffic problem, and it shows up in the search logs.

None of these are headline numbers. None of them belong on the six-number scoreboard. But they’re the difference between knowing that something moved and knowing what to do about it, which is the entire point of keeping the two layers separate in the first place.

Some of them work as early warning, channel mix and content trending can catch a shift before it’s big enough to move Visitors at all, let alone Revenue. Others confirm what’s already happened. Either way, they’re all earlier than Revenue, which is always the last number to tell you the truth.

That’s how analytics should work: a handful of indicators for direction, and detailed reports for diagnosis, in that order.