How to Measure Marketing ROI Without a Big Analytics Team
DIGITAL MARKETING · 10 MIN READ

How to Measure Marketing ROI Without a Big Analytics Team

Four crude measures, honestly maintained, beat one sophisticated dashboard you do not trust. Precision was never available to a small business anyway.

By Tariq Sallam·September 2026

Attribution is now modelled rather than observed, platforms over-report, analytics under-reports, and no dashboard you build will reconcile them. Small businesses respond to this by either building something elaborate they distrust or by measuring nothing.

There is a third option: four crude measures that are hard to argue with, maintained honestly. That is what I use with clients and it is enough to make every decision a small business needs to make.

Quick Info

Primary measure
Blended cost per acquisition
Most accurate for small businesses
Asking every enquiry where they found you
Best test of causation
Turning a channel off
Stop reporting
Sessions, impressions, follow counts, platform-reported ROAS as fact
Time needed
Thirty minutes a month
What you gain
Directional truth instead of false precision

Measure one: blended cost per acquisition

Total marketing spend divided by total new customers, monthly. Include everything: media, tools, agency fees, and a realistic value for your own time if you do the work.

—It ignores attribution entirely, which is its strength.
—It cannot be inflated by any platform's reporting.
—Compare it against what a customer is worth to you over their lifetime. That ratio is the whole business case for marketing.
—Track it monthly and watch the trend, not the individual month.

If blended cost per acquisition is stable while customer numbers rise, you are scaling successfully, whatever the channel dashboards say. If it climbs, you are reaching the edge of efficient demand.

One crude number nobody can dispute is worth more than five precise numbers that contradict each other.

Measure two: ask them

"How did you hear about us?" asked at the point of enquiry, recorded consistently. Unfashionable, and frequently the most accurate attribution available to a small business.

01Ask on the enquiry form as an optional open field, and again on the call.
02Record it somewhere you will aggregate: a spreadsheet or a CRM field.
03Expect vagueness. "Google" covers organic, paid and maps, and that is still useful.
04Review monthly against your spend. If nobody mentions the channel you spend most on, that is a finding.
05Accept that people misremember. In aggregate, over months, the pattern is reliable enough to act on.

I have seen this reveal that a client's best source was a referral relationship nobody was maintaining, while the budget went entirely to ads.

Measure three: cohorts by month

Group customers by the month they first bought and track what they are worth over time. This is what tells you whether your marketing is bringing in good customers or merely cheap ones.

—Record first purchase month, first purchase value and total value to date.
—Compare cohorts: are customers acquired in March worth more or less at six months than those from January?
—A falling cost per acquisition with falling cohort value means you are buying worse customers, which looks like success on the dashboard.
—This is a spreadsheet, not a system. Twenty minutes a month.

Measure four: turn it off

The only reliable test of causation available to a small business. Reported returns tell you what a platform claims; an off period tells you what actually changes.

01Pick one channel. Retargeting is the best first candidate because its reported return is the most flattering.
02Turn it off entirely for two to four weeks, holding everything else constant.
03Compare total enquiries and total sales, not per-channel attribution.
04If total volume falls, the channel was adding demand. If it holds, it was claiming credit for demand you already had.
05Repeat annually, because the answer changes as your traffic mix changes.

This is uncomfortable and it is the single most valuable measurement exercise I run. In some accounts it justified the spend; in others it freed a budget nobody realised was redundant.

The tracking setup worth having

Modest, and enough to support all four measures.

01Analytics on the site, with conversion events for enquiry, form submission and phone click.
02Search Console, for impressions and queries.
03A source field on your enquiry form, and the question asked on calls.
04Server-side conversion tracking to your ad platforms, sending real outcomes back.
05A single spreadsheet: month, total spend, new customers, source breakdown, blended cost per acquisition.
06That is the whole stack. One spreadsheet and four tools you probably already have.

What to stop reporting

—Sessions as a headline number. With AI answers absorbing informational queries, falling sessions may accompany rising visibility.
—Impressions, unless you are measuring awareness deliberately.
—Follower counts and engagement rate, which have almost no relationship to revenue for a small business.
—Platform-reported return on ad spend presented as fact. It is internally consistent and it over-reports.
—Open rates, which privacy features have made unreliable in both directions.
—Anything you have never once acted on. If a metric has never changed a decision, delete it from the report.

The metrics that lead the revenue

Blended cost per acquisition is a lagging measure. Three things move earlier and are worth watching.

—Qualified enquiries per month, which is upstream of customers.
—Branded search volume, which rises when your marketing is landing even without clicks.
—Response time on enquiries, which determines how many of them convert.
—Conversion rate on your key pages, which multiplies everything upstream.

Those four move before revenue does, and each is directly actionable, which is more than can be said for a modelled attribution figure.

A thirty-minute monthly routine

01Fill in the spreadsheet: spend, new customers, blended cost per acquisition.
02Aggregate the "how did you hear about us" answers.
03Update the cohort sheet with last month's customers.
04Note qualified enquiries and average response time.
05Compare against the previous three months, not against last month alone.
06Write one sentence about what you will change. If there is nothing, change nothing.

Thirty minutes, four numbers, one decision. That is a functioning measurement practice for a business of any size below a marketing department.

When to invest in more

When spend reaches the point where a 10% improvement is worth more than the cost of measuring it properly. For most small businesses that is somewhere above £10,000 a month in media.

Below that, the crude measures are not a compromise. They are the appropriate level of precision for the size of the decisions you are making.

Frequently Asked Questions

How do I measure marketing ROI without an analytics team?

Blended cost per acquisition, asking every enquiry where they found you, monthly cohort values, and occasionally turning a channel off. Thirty minutes a month.

What is blended CAC?

Total marketing spend divided by total new customers in a period. It ignores attribution disputes and is the number the business actually runs on.

Why do my platform and analytics numbers disagree?

Different attribution models and windows, plus modelling for conversions that cannot be observed. Use platform numbers to compare ads, and your own numbers to judge the channel.

Is asking customers how they found us reliable?

Individually, no. In aggregate over several months, it is frequently more accurate than any dashboard available to a small business.

How do I know if a channel actually causes sales?

Turn it off for two to four weeks with everything else constant and compare total sales. It is the only real incrementality test at small scale.

Which metrics should I stop tracking?

Sessions as a headline, impressions, follower counts, and any metric that has never once changed a decision you made.

Before You Go

Four measures, one spreadsheet, thirty minutes a month, and the willingness to turn a channel off occasionally. That is more measurement discipline than most small businesses have and it is enough.

Then use it to set next year's budget from evidence rather than from last year's number. The allocation guide is where that goes.

Four crude numbers. Thirty minutes. One decision.

WRITTEN BY TARIQ SALLAM
Marketing Consultant. Entrepreneur. Content Creator.

I'm a marketing consultant, entrepreneur and content creator. I help businesses grow through practical marketing, websites, SEO, content and AI.

More About Tariq →

Keep reading