How to Measure Marketing ROI When You Are Not a Data Team: A Small-Business Method

Quick answer

Most small businesses either measure nothing (“we have always done the parish magazine ad”) or drown in dashboards measuring everything except money. The workable middle is one question asked of every channel (what did a customer from this source cost me, and what are they worth?), answered with one spreadsheet, one habit (“how did you hear about us?”, recorded every time), and a handful of per-channel tracking tricks. Perfect attribution is a myth even for corporates; a small business that knows its rough cost-per-customer by channel beats 90% of competitors’ decision-making.

Fixed-price services make the cost side of this sum trivial: every Luxbranding price is published, so the spend figure is never a guess.

The only two numbers that matter

• Cost per customer (not per click, not per like): everything a channel costs (spend plus your time at an honest hourly value) divided by customers it produced. A “free” channel eating ten hours a week is not free.
• Customer value (first job vs lifetime): a £90 first job from a customer who returns twice a year for a decade is a £2,000 customer. Channels differ wildly here: price-led ad clickers often stay price-led; referred customers refer. Judge channels on the customers they bring, not just the count.

ROI is the second number divided by the first. Anything above ~3:1 on lifetime value is usually worth scaling; persistent sub-1:1 channels are hobbies. The point of the exercise is not accounting precision. It is knowing which of your channels deserves next month’s marginal pound.

Tracking each channel without a data team

The “how did you hear about us?” habit is the free infrastructure underneath all of it: asked on every enquiry call and form, written in the same spreadsheet column every time. Imperfect (customers say “Google” meaning three different things) but directionally reliable within a quarter.

The attribution traps to consciously ignore

• Last-click worship. The customer saw your van, read two blog posts, then clicked an ad, the ad gets the credit. Channels assist each other; treat your numbers as a portfolio, not a league table, and be slow to kill “assist” channels that never close but always appear in the story.
• Vanity metrics as proxies. Reach, impressions and likes are inputs, not outcomes. They only matter if they eventually appear in the customer column, the same discipline as judging social management on enquiries, not followers.
• Judging slow channels on fast clocks. SEO and brand-building compound over quarters; ads work in days. Comparing them month one is comparing a sapling to a bouquet. The fair comparison in SEO vs Google Ads is cost-per-customer over a year.
• Precision theatre. Multi-touch attribution models are a corporate sport. Your version (one spreadsheet, honest costs, quarterly review) produces better decisions than a dashboard nobody reads.

The quarterly review that closes the loop

Once a quarter, thirty minutes: customers per channel, honest cost per channel, divide, compare, decide one change: scale the winner, fix or kill the loser, keep the rest steady. Write the decision down; next quarter, check it worked. That loop (measure, decide, verify) is the entire discipline, and it is why the businesses that do it end up with marketing budgets that compound instead of recycle. The prerequisite is a website that converts the traffic every channel sends it; if enquiries leak there, fix that first, here is where they leak.

Build the spreadsheet once: seven columns, thirty minutes

The whole method fits in one sheet. Add a row per channel per month and fill it in from the platform reports plus your enquiry log.

The seventh column is the one small businesses skip and the one that changes decisions. A channel producing £10,000 of revenue at 15% gross margin contributes £1,500; one producing £4,000 at 60% contributes £2,400. Judged on revenue you scale the wrong channel; judged on gross profit you scale the right one.

A worked example

Take a month for a small trade or service business. Google Ads: £600 spend, 3 hours of your time at £45, 22 enquiries, 5 customers, £4,400 revenue, 45% gross margin. That is £735 of cost against £1,980 of gross profit: about 2.7 to 1, and each customer cost £147 to acquire. Referrals the same month: £0 spend, 2 hours of asking, 6 enquiries, 5 customers, £5,900 revenue at 52% margin. £90 of cost against £3,068: 34 to 1, at £18 a customer.

The naive conclusion is “stop the ads”. The better one is that referrals are drastically underfunded: if two hours of asking produces five customers, what would a structured programme produce at four hours and a £50 thank-you? The exercise is less about cutting losers than about noticing which cheap channel you have been treating as an accident. Sizing the total pot those decisions come out of is a separate question, worked through in how much a small business should spend on marketing.

The numbers that make the ROI calculation fair

• Payback period, not just ratio. A 4:1 return that takes fourteen months to arrive can still bankrupt a business with a 60-day cash cycle. Ask when the money comes back, not only how much.
• Repeat rate by channel. What proportion of each channel’s customers buy again within a year? This is where price-led and trust-led channels diverge most sharply.
• Contribution per lead, not cost per lead. Two channels at £30 a lead are not comparable if one converts at 10% and the other at 40%.
• Capacity ceiling. A channel returning 8:1 that can only ever produce three customers a month is not scalable, however good the ratio. Note the ceiling next to the ratio.
• Seasonality. Compare like months year on year before declaring a channel dead in January.

Common measurement errors that produce confident wrong answers

1. Counting spend but not time. The “free” channel that eats six hours a week costs more than the ad account.
2. Counting revenue but not delivery cost. High-revenue, low-margin work flatters a channel that is quietly making you poorer.
3. Attributing to whatever the customer says last. “Found you on Google” often means they saw the van, searched your name, and clicked. Ask a second question (“and had you come across us before?”) and log both answers.
4. Judging on too few data points. Five customers is a story, not a sample. With small volumes, look at a rolling quarter and expect noise.
5. Ignoring the leads you never answered. A missed-call rate of 20% is a marketing problem wearing an operations costume, and no channel measurement will surface it unless you count enquiries received against enquiries responded to.

Setting up the tracking properly, once

Three pieces of plumbing remove most of the guesswork and take an afternoon. First, UTM tags on every link you control (email campaigns, social bios, QR codes, directory profiles) so analytics stops lumping them into “direct”. Second, a separate phone number or a call tracking setup for anything printed, because print’s whole measurement problem is that a phone call carries no source. Third, defined conversion events: form submissions, click-to-call taps and completed bookings recorded as goals rather than inferred from traffic. The five numbers worth watching once that exists (and the many more worth ignoring) are set out in website analytics for beginners.

One caution: platforms only see visitors who accept tracking, so measured conversions understate reality. Treat the enquiry log as the source of truth and the dashboards as supporting evidence.

Turning the numbers into a decision you can defend

At the quarterly review, write three sentences: this is the channel we are increasing and by how much, this is the channel we are changing or stopping, and this is what we expect to see by the next review. Put a number on the expectation; next quarter, check it. Businesses that keep that record accumulate a written history of what works in their specific market, which no agency pitch or industry benchmark substitutes for.

Two channels deserve extra patience in that process. Referrals, because the returns look absurdly good precisely because nobody funds them properly: a deliberate programme is usually the fastest ROI improvement available, as covered in referral marketing. And organic search, because it is measured on a clock of quarters rather than weeks, accrues to an asset you own rather than rent, and reliably looks like a failure at month three, which is why judging an SEO programme on a single month’s cost per customer is the most expensive measurement mistake on this list.

Sources & Further Reading

• Conversion Tracking, Google
• Search Console, Google
• Marketing Measurement Research, Data & Marketing Association