Marketing KPIs: The Numbers Worth Reporting

I have been in this industry 13 years, and the most common waste I see is not bad marketing. It is a monthly report nobody reads because it is full of numbers that never change a decision. Owners get a twelve page PDF with impressions and engagement rate, they nod, and they still cannot answer the only question that matters: is this working, and should I keep paying for it.

The fix is fewer, better numbers and a reporting rhythm a small business can actually keep. That is deliberately a different job from working out return on investment, and different again from attribution, both of which assume the basics of measurement are already in place.

The decision this post pushes is about the report itself. Cut your dashboard to five numbers, decide who reads each one, and put a review date in the calendar before you add another metric.

What a Marketing KPI Is, and What It Is Not

The plain definition I give clients

A marketing KPI is a number with a target, an owner and a decision attached to it. If a number has none of those three, it is a statistic, not a key performance indicator.

KPIs versus vanity metrics

Vanity metrics go up and make nobody do anything. Impressions, follower counts, engagement rate and email open rate are the usual suspects. They are not useless, but they are diagnostic, not decisive. A KPI is what you would take to a board meeting or a bank, and what you would use to decide where next month's budget goes.

Why a metric without an owner is decoration

Every number on a marketing report should have one name against it, even if that name is the owner of the business. I have seen dashboards with 20 metrics and no owner for any of them, which is why nothing ever changed after a bad month. One number, one person, one decision they are accountable for making.

The Five Questions Your KPIs Have to Answer

Are we reaching the right people

Not how many people, but how many of the right ones. Sessions from your target market, not total sessions. If you sell to trades in Victoria, traffic from overseas does nothing for you except make the graph look better.

Are they engaging

Engaged means they did something deliberate: read a second page, watched the explainer, returned for a second visit. Time on page is a weak proxy for this, and bounce rate alone tells you very little on a single page site.

Are they converting

Enquiries, calls, bookings, quotes requested, orders. Count the actions you would actually pay for, and count them the same way every month so the trend means something.

What does it cost us

Cost per lead and cost per sale, by channel. This is the question most small business dashboards answer worst, because it requires tracking to be joined to spend, and that join is usually broken.

Is it repeatable

The numbers should tell you whether a good month was luck or a system. If all the enquiries came from one referral and one lucky search result, you do not yet have something repeatable, and you should say so out loud.

Choosing KPIs for a Small Business

Start with the decision, then pick the number

Write down the decision the report is meant to support, for example whether to keep spending on a channel. Then choose only the numbers that could change that decision. Anything that cannot is noise, however interesting it looks.

One number per funnel stage

Reach, qualified traffic, leads, conversion and cost per acquisition. Five numbers, one per stage, is enough to run a small business by. It also makes the weakest stage obvious, which is the point of measuring at all.

How many KPIs is too many

If nobody can remember the list without looking it up, there are too many. I have seen businesses track 30 metrics and act on none. Five is a defensible ceiling for a business under a few million in revenue.

The Marketing KPIs I Actually Track

Reach and qualified traffic

Impressions and sessions are fine as context, but the number I care about is qualified sessions: visitors from the locations, keywords or audiences that match people who buy. Everything else is background.

Lead volume and lead quality

Volume tells you if the machine is running. Quality tells you whether it is worth running. I score leads on a simple three point scale and report the mix, because a month with 40 leads and no buyers is worse than a month with 8 good ones.

Conversion rate by source

Overall conversion rate hides more than it reveals. Channel level conversion rate shows you where to spend and where to stop, and it is usually the single most useful line on the page.

Cost per lead and cost per sale

Pull both. Cost per lead can look excellent while cost per sale is unworkable, because cheap leads are often cheap for a reason. Report them together, always.

Customer lifetime value

For any business with repeat purchase or a retainer, lifetime value changes the answer to almost every other question. It also justifies spending more to acquire a customer than a single sale would suggest, and it is the number owners most often leave out.

Measurement practices are shifting in step with this. HubSpot's 2026 State of Marketing report found that "67.5% know how to measure AI impact, compared with 48% in 2025". More teams claim the ability to measure new channels. Far fewer can show a clean number for the basics.

Why Marketing Reporting Goes Wrong

Reporting the channel instead of the outcome

Channel reports describe activity: posts published, emails sent, budget spent. Outcome reports describe results. Owners want the second one and usually get the first, which is why reporting feels like homework.

Data no one trusts

Once a report has been wrong twice, nobody uses it again. The remedy is boring consistency: same definitions, same source system, same date range, every month, even if the numbers are less flattering that way.

Reports with no review meeting attached

A report that arrives by email and is never discussed changes nothing. Put 30 minutes in the calendar, and treat the meeting as the deliverable with the report as its agenda. The bigger problem is not measurement at all: HubSpot's State of Marketing survey reports that "The top challenges marketers cited in the 2026 State of Marketing survey are measuring the ROI of marketing activities (33%), keeping up with the latest trends and new platforms/features (29.8%), generating leads (29.6%), and sales-marketing alignment (27.6%)". Measuring the return is the first of those for a reason, and it starts with the report being read.

Setting a Reporting Cadence You Will Keep

Weekly numbers versus monthly numbers

Lead volume and spend are worth watching weekly, because a broken form or a paused campaign shows up quickly. Conversion rates and cost per lead need a month of data before they mean anything. Testing a conversion rate on a week of small numbers produces panic, not insight.

The monthly review that takes 30 minutes

Five numbers against target, the trend, and one decision. If it takes two hours, it will be skipped in a busy month, and the months it gets skipped are exactly the months you needed it.

The quarterly reset

Every quarter, review whether the targets themselves are still right and whether any metric has stopped earning its place. Businesses change, and a KPI set that suited you a year ago may now be measuring something you no longer sell.

How to Keep Your Data Trustworthy

One definition per metric, written down

What counts as a lead. Whether a quote counts as a conversion. Whether you include spam form fills. Write the definitions in one document and use them everywhere. Most disagreements about numbers are disagreements about definitions.

Fixing the tracking before trusting the numbers

Test that your forms, calls and checkout events all record correctly before you build reports on top of them. HubSpot's research found that "Just 65% of marketers say they have high-quality audience data", which means a third are reporting from a shaky foundation. Supermetrics' Marketing Data Report 2026 points to the same thing: "According to the Marketing Data Report 2026, the top marketing data problems are unified customer view (34%), predictive analytics and forecasting (34%), and competitive intelligence (33%)." A dashboard built on data you do not trust is worse than no dashboard, because it still gets read. The same report lists where measurement effort is heading over the next 12 months: the most popular investments are marketing mix modelling (40 per cent), AI creative testing (37 per cent) and A/B testing (36 per cent), which is a fair guide to where a small business gets the most value from tightening its numbers.

What to do when two tools disagree

Pick one system as the source of truth for each metric, write it down, and stop reconciling by hand every month. Platforms will always disagree slightly on sessions and conversions because they count differently. Decide once which number you report, then leave it alone.

Building a One Page Marketing Dashboard

The five tile layout I use

One tile per funnel stage, each showing the current period, the previous period and the target. Five tiles on one screen, readable in a minute. If a tile needs a second page to explain it, the metric is wrong.

Showing trend rather than a single month

A single month is noise. Show the last six months as a line so you can see direction, and mark the month you changed something so you can see whether it did anything.

Making it readable on a phone

Most owners read reports on a phone, standing up, between jobs. If the dashboard needs a laptop and a zoom, it will not be read. Test it on the smallest screen you own before you call it finished.

Getting the plumbing right is usually a data problem before it is a reporting problem, which is where Google Analytics consulting helps, and GA4 migration work fixes the foundations if your setup was never finished. If you want the measurement reviewed as part of the wider picture, a marketing audit looks at what you are measuring, and ROI analysis services take it further when you are ready to judge the return rather than the report.

Marketing KPIs for Australian Small Business: Worked Example

A service business with a long sales cycle

A commercial fitout business might report qualified enquiries, quotes issued, quotes won, average job value and cost per job won. Weekly they watch enquiry volume and spend. Monthly they watch quote to win rate. A falling win rate is a sales problem or a pricing problem, and the dashboard should name which within a month.

An ecommerce business with a short one

An online store might report qualified sessions, add to cart rate, checkout completion rate, average order value and cost per order. Weekly they watch orders against a target. When orders fall, the split between traffic and conversion tells them whether to spend or to fix the site, and that single distinction is worth the whole reporting exercise.

Marketing KPIs FAQ

How many marketing KPIs should a small business track?

Five is a good ceiling, one for each funnel stage: reach, qualified traffic, leads, conversion, and cost per acquisition. If a sixth number never changes a decision, it does not belong on the page.

What is the difference between a KPI and a metric?

Every KPI is a metric, but not every metric is a KPI. A KPI is a number with a target, an owner and a decision attached. Sessions is a metric. Qualified leads per week against a target of 12 is a KPI.

How often should I report on marketing?

Weekly for lead volume and spend, monthly for conversion and cost per lead, quarterly for channel performance and strategy. Daily reporting creates noise and anxiety in small teams.

Is measuring marketing ROI the same as reporting KPIs?

No. They are separate jobs. Reporting a KPI tells you what happened. Working out return on investment tells you whether the money was well spent overall, and that is a different exercise with different inputs. HubSpot's 2026 State of Marketing found measuring the ROI of marketing activities is the top challenge marketers cite, at 33 per cent, which is a good reason to do the reporting job properly first.

What if my different tools give different numbers?

They usually will, because they define a lead or a session differently. Pick one system as the source of truth for each metric, write the definition down, and stop reconciling by hand every month.

Do I need expensive software to track marketing KPIs?

Not to start. A spreadsheet connected to your analytics and your CRM will do for the first year. Supermetrics' 2026 report found the top marketing data problem is a unified customer view, at 34 per cent, and that problem is about definitions and plumbing more than software.