The Marketing KPIs That Actually Matter
Follower counts and raw traffic feel like progress. They rarely predict revenue — here's the honest list of what does, and what to stop leading with.
Most marketing reports are full of numbers that went up. Impressions up 40%. Followers up 2,000. Website sessions up 18%. All true, all reported with a straight face, and none of them tell you whether the business is actually better off.
That gap — between "the number moved" and "the business improved" — is where marketing loses credibility with finance and leadership. This guide sorts the two apart: the metrics that earn a slide in a monthly report because they show reach and activity, and the metrics that earn a seat at the budget table because they predict revenue.
It sits under our broader marketing funnel guide, which maps how a stranger becomes a customer stage by stage. Read that first if you need the funnel model itself; this page is about which numbers to put on the dashboard once you know which stage you're measuring.
What makes a metric a KPI, not just a number
A KPI (key performance indicator) is a metric a business has deliberately chosen to track because it's close enough to an outcome that matters — usually revenue, sometimes retention or margin — that acting on it changes a decision. Every other number is just data. It might be useful diagnostically, but it isn't a KPI unless someone would actually change budget, creative, or targeting because it moved.
That test — "would we act differently if this number changed?" — is the fastest way to sort a real KPI from a vanity metric. Follower count going from 40,000 to 42,000 rarely changes anyone's next move. Cost per lead going from ₹800 to ₹1,400 changes several people's next move immediately.
Why vanity metrics survive anyway
Vanity metrics aren't fake, and they aren't always worthless — the problem is proportion, not existence. They survive in reporting because they almost always go up (which feels good to show), they're easy to pull from a native platform dashboard with zero setup, and they're easy for a non-marketer to understand at a glance. None of those are reasons they predict revenue.
The honest fix isn't to delete every vanity metric from a report. It's to demote them: keep them as small supporting context underneath the KPIs that actually earn the top line, instead of leading with them.
The vanity metrics that get more attention than they deserve
Each of these is genuinely useful in the right context — as a diagnostic, or as one input among several. None of them should be the headline number in a report to leadership.
Follower / subscriber count. A large following can indicate brand reach, but it says nothing about whether those followers ever buy. Accounts buy followers, algorithms show content to non-followers anyway, and a follower gained during a giveaway behaves nothing like one who found you through a genuine search for a solution. Track it as a slow-moving brand signal, not a growth KPI.
Raw impressions and reach. Impressions measure exposure, not interest. A campaign can rack up a million impressions by running on cheap, low-quality inventory nobody actually looks at. Impressions matter only paired with what happened after someone saw the ad — click-through, and further downstream, conversion. On their own, they measure spend, not results.
Raw website traffic (without conversion context). "Traffic is up" is one of the most common false-progress claims in marketing. A spike in sessions from an unrelated viral post, a bot crawl, or a low-intent keyword that happens to rank well can inflate the number while leads or sales stay flat — sometimes traffic rises specifically because the new visitors are the wrong audience. Traffic only becomes meaningful next to a conversion rate.
Engagement rate in isolation. Likes, comments and shares can signal that content resonates, but engagement-optimised content and revenue-generating content are frequently two different things — a meme gets shared, a case study gets a demo request. Report engagement as a content-quality signal, not as proof the marketing spend is working.
Email list size. A list of 100,000 addresses is worth less than a list of 8,000 people who open and click, if the larger list is stale, unengaged, or was never permission-based to begin with. Deliverability and open/click rates on an active segment tell you far more about list health than the raw subscriber count ever will.
The KPIs that actually predict revenue
These are the numbers worth leading a report with, because each one sits close enough to a business outcome that a change in it should change a decision — budget, targeting, creative, or which channel gets more of next quarter's spend.
Cost per lead (CPL). What it costs, on average, to generate one lead — total spend divided by leads generated. It's a genuinely useful efficiency number, but only alongside lead quality; a channel that halves your CPL while also halving lead quality hasn't actually improved anything. We cover this trade-off in more depth in our CPL guide. What counts as a "good" CPL varies enormously by industry, ticket size and sales cycle, so we won't invent a single benchmark figure here — judge it against your own historical CPL and against what a customer is actually worth to you, not a number from a blog post.
Lead-to-customer rate. The percentage of leads that actually become paying customers. This is the number that catches the failure mode CPL alone misses: a channel can generate cheap leads that convert to customers at a fraction of the rate a more expensive channel does, making the "cheaper" channel more expensive in the end. Tracking this rate by channel and by campaign is usually the single fastest way to find where marketing budget is quietly being wasted.
Marketing-sourced (and marketing-influenced) revenue. Revenue from deals that marketing generated, or meaningfully touched, tracked back to the campaigns and channels responsible. This is the metric that finally puts marketing in the same currency as sales and finance — not "we generated 400 leads" but "we generated ₹X in closed revenue." It requires a working CRM-to-marketing data connection to calculate honestly; without one, treat any number claiming to be marketing-sourced revenue with real skepticism.
CAC vs. LTV ratio. Customer acquisition cost (what it costs to win one customer, fully loaded) compared against customer lifetime value (what that customer is worth over the time they stay). A channel with a low CAC but customers who churn in two months can be worse than a channel with a higher CAC and customers who stay for years. Our CAC guide breaks down how to calculate it properly, including the costs that get left out of a naive version of the formula.
Assisted conversions. The conversions a channel contributed to without getting final credit — the paid social ad someone saw two weeks before they searched your brand name and converted through organic search. Judging a channel purely on last-click credit routinely undervalues awareness-stage and research-stage channels, and over-values whichever channel happens to sit closest to the purchase. This guide only flags that the problem exists; which specific model should get the credit — first-click, linear, data-driven — is a deeper topic on its own that deserves a dedicated breakdown rather than a paragraph here.
| Vanity metric | What it actually shows | What it doesn't show |
|---|---|---|
| Follower count | Slow-moving brand reach | Whether followers ever buy |
| Raw impressions | How much inventory you bought | Whether anyone cared |
| Raw traffic | That something drew visits | Whether the visits were the right audience |
| Engagement rate | Content resonance | Revenue impact |
| Email list size | Historical reach of the list | Whether the list is still active |
How SEO reporting fits into this
Everything above is a cross-channel framework — it applies whether the spend is in paid media, organic, social, or all three together. If you specifically need the SEO-side version of this argument — rankings and raw organic sessions aren't KPIs either, and what to report on instead for an SEO programme — that's covered in depth in our SEO reporting guide. Read this page for the whole-business KPI set; read that one when the question is specifically "is our SEO retainer working."
Building a KPI set that doesn't drown the business in dashboards
The goal isn't to track every metric on this page for every channel every week. It's to pick a small set deliberately, tied to the stage of the funnel each channel actually serves.
- Pick one north-star metric. Usually marketing-sourced revenue, or CAC vs. LTV for a subscription business. This is the number that goes in front of leadership.
- Map supporting metrics to funnel stage. Awareness-stage channels get reach and assisted-conversion credit; consideration-stage channels get CPL and engagement; bottom-of-funnel channels get lead-to-customer rate and CAC. Trying to hold a top-of-funnel channel to a bottom-of-funnel KPI is a common, avoidable mistake.
- Fix tracking before you fix the number. A CAC or CPL calculated on broken conversion tracking is a confident wrong answer, not a useful one. Instrumentation comes before optimisation.
- Report monthly, decide quarterly. Most of these KPIs are too noisy week to week to act on weekly. Watch them monthly for early warning; make actual budget and channel decisions on a slower, quarterly cadence once the noise has settled.
- Keep one vanity metric per report, clearly labelled. Reach or follower growth is fine as a small line item for brand-health context — it just never gets to be the headline.
What "good" looks like, and why we won't give you a single number
CPL, CAC, and conversion rate all vary too much by industry, ticket size, and sales cycle length for a single benchmark figure to be honest. A ₹500 CPL might be excellent for a low-ticket D2C product and a loss-making number for an enterprise B2B sale. Anyone quoting you one universal "good CAC" number is simplifying past the point of usefulness. If you want the honest, India-context treatment of conversion-rate benchmarks specifically, see our conversion rate guide — and for ROAS specifically, what a good ROAS actually looks like walks through why a single ROAS target misleads without a margin figure next to it.
Where to go from here
Start by auditing your own last monthly report: for each number on it, ask whether anyone would actually act differently if it moved. Anything that fails that test gets demoted to supporting context, not deleted outright. If you want a KPI framework and reporting cadence built for your specific funnel and industry, that's exactly what we scope in every performance marketing engagement — management fees typically run ₹25,000 to ₹1,00,000+ a month depending on scope, always separate from ad spend.
Frequently asked questions
What's the single most important marketing KPI?
There isn't one universal answer — it depends on the business model. For most companies selling directly, marketing-sourced revenue or the CAC vs. LTV ratio is the closest thing to a north star, because both connect marketing activity directly to money. For an early-stage business still validating channels, cost per lead paired with lead-to-customer rate is usually more actionable day to day.
Are vanity metrics ever worth tracking at all?
Yes, as supporting context rather than headline numbers. Follower growth and reach can be useful for gauging brand health over a long horizon, and raw traffic is a fine diagnostic for whether content is being discovered at all. The mistake is leading a report with them or treating their growth as proof marketing is working.
How is this different from your SEO reporting guide?
This page is the cross-channel framework — the KPIs that matter across paid, organic and social together, and which ones are vanity metrics regardless of channel. Our SEO reporting guide applies the same underlying logic specifically to an SEO programme — what to actually report on instead of rankings and raw sessions. Read this page for the whole-business view; read that one for the SEO-specific version.
How often should we review these KPIs?
Monthly for visibility, quarterly for actual decisions. Most of these numbers — CPL, CAC, lead-to-customer rate — carry enough natural noise week to week that reacting to a single bad week usually does more harm than good. Watch monthly, decide quarterly, and only move faster than that if a number moves sharply enough to be an obvious outlier.
Does tracking marketing-sourced revenue require attribution modeling?
It requires connecting marketing activity to closed revenue in your CRM, which is a data and tracking problem more than a modeling one. Deciding exactly how much credit each individual touchpoint gets on the way to that revenue — which is what attribution modeling answers — is a separate, deeper question. This guide only covers which top-level numbers to track; a dedicated breakdown of attribution models is a topic on its own.
Want a KPI set built for your actual funnel?
We set up tracking, define the KPI set, and report on revenue and CAC — not impressions and follower counts.