How to Measure Marketing ROI: The Framework We Use.

Here's why most marketing ROI conversations fail: they measure the wrong things.

A business pays £2,000 a month for paid ads and gets 50 clicks. They call it a success because "50 people saw us." Six months later, they've spent £12,000, gotten 300 clicks, and made zero sales. They blame the platform. The real problem is they were measuring activity instead of outcome.

ROI isn't complicated. It's dead simple: money in, money out, and the gap between them. The hard part isn't the maths. It's isolating what actually caused the outcome, resisting the urge to take credit for things you didn't do, and being honest when something isn't working.

This is the framework we use to keep clients honest about their marketing spend. It works because it forces you to define what "working" actually means before you spend a penny. Most businesses never do.

The quick version: the ROI framework at a glance

Why most businesses measure ROI wrong (and how they know)

There's a pattern. A business hires an agency. Three months in, they ask "where's my ROI?" The agency shows them a dashboard with traffic, clicks, impressions, and engagement. Everything's up. The business leaves the call confused because they still don't know if it's working.

This happens because most measurements confuse activity with outcome.

Here's the tell: if your metric doesn't directly connect to money in or money out, you're measuring something that feels productive but might be pointless.

Activity metrics (everyone measures these, most don't matter):

• Website traffic
• Page views
• Clicks
• Impressions
• Social media followers
• Email open rates

These can all go up while your revenue stays flat. A brand awareness campaign might double your traffic and still not make you a penny - and that might be fine if brand awareness was the actual goal. But most businesses conflate "more traffic" with "more business," and they're not the same thing.

Outcome metrics (these are what matter):

• Leads generated
• Cost per lead
• Conversion rate (leads to customers)
• Revenue per customer
• Profit per marketing pound spent

This is where the confusion sets in, because the gap between activity and outcome is where most marketing budgets disappear.

We ran a paid ads campaign for an e-commerce client that drove 5,000 clicks. Traffic was up 400%. Owner was thrilled. Then we looked at what happened next: 40 people made it to the checkout. 8 people completed a purchase. Total revenue: £1,200 on £2,500 spent in ads. Not ROI. Negative ROI.

The ads were working. The traffic was real. The problem was the landing page converted at 0.2% instead of the industry standard 2%. The ads revealed a problem with the offer, not with the ads themselves. Once the offer was fixed, same traffic volume, 10x better results. See our SuperGro Lasers case study for a real example of e-commerce ROI done right.

That's the difference between measuring activity and measuring outcome.

The BlackFire ROI framework: what we actually track

We don't publish our methodology because every business is different. But here's how we look at whether marketing spend is working:

Step 1: Define the outcome you're paying for

Before a single pound is spent, you need to know: what am I paying for? A lead? A customer? A specific type of customer?

Most businesses skip this step and wonder why they can't measure success.

The answer changes everything:

• Lead-gen businesses need cost per lead and lead quality (what % convert to customers?).
• E-commerce needs cost per transaction and average order value.
• SaaS needs cost per free trial signup and free-to-paid conversion rate.
• B2B services need cost per qualified opportunity and sales cycle length.

Pick the wrong outcome and you optimise for the wrong thing. Optimise for volume of leads and ignore conversion rate, and you end up with cheap leads that don't close.

Step 2: Establish your baseline math

Before you spend anything on marketing, answer these questions:

What's a lead worth to you?

• Take your annual revenue.
• Divide by number of customers you closed last year.
• That's your revenue per customer.
• Now multiply by your average profit margin.
• That's what a new customer is actually worth in profit.

Example: £500,000 revenue, 50 customers closed, 30% margin = £10,000 revenue per customer = £3,000 profit per customer.

How many leads does it take to close one customer?

• Count how many leads you got last year.
• Count how many became customers.
• Divide the first by the second.
• That's your conversion rate.

Example: 200 leads, 50 customers = 25% conversion rate (or 1 in 4 leads closes).

So, what can you afford to spend per lead?

• Take your profit per customer (£3,000 in the example).
• Multiply by your conversion rate (25%, or 0.25).
• That's your maximum sustainable cost per lead.
• Formula: £3,000 × 0.25 = £750 per lead maximum.

If you can get leads for £200, you've got room to scale. If your cost per lead is £800, you're underwater.

This is the single number most businesses never calculate. It's also the only number that matters.

Step 3: Set up tracking that connects marketing to money

Here's where most DIY marketing falls apart. You're running ads in one tool, tracking leads in another, and sales data lives in a spreadsheet. Nothing connects.

You need one truth source. It doesn't have to be complicated:

For simple setups (one paid channel, straightforward sales):

• UTM parameters on every ad so you know which campaign each lead came from.
• A CRM that captures lead source.
• A monthly reconciliation: leads from Campaign X → customers from Campaign X → revenue.

For complex setups (multiple channels, long sales cycles):

• Full attribution tool (Google Analytics 4 handles basic attribution; Mixpanel, Amplitude handle advanced).
• Proper CRM setup so every lead is tagged with source.
• Sales team discipline: every opportunity must record where it came from.

A premium glazing business we worked with was running paid ads, SEO, and referrals simultaneously. They had no idea which channel was actually profitable. By tagging every lead with source and tracking them through to closed revenue, we discovered:

• Paid ads: £7.34 cost per lead, 18% conversion to customer, £18,000 annual revenue from £6,000 spend = 300% ROI
• SEO: £120 effective cost per lead (over time), 35% conversion rate, £35,000 annual revenue from £0 annual spend = ∞ ROI
• Referrals: £4 cost per lead (referral incentive), 40% conversion rate, £20,000 annual revenue from £500 spend = 4,000% ROI

They'd been thinking of referrals as "free leads." In reality, they were the most profitable channel. Once they knew that, they invested in referral incentives instead of more paid ads.

None of that was visible before they tracked it properly. It's the same £7.34 cost-per-lead result you'll see in our Mazuli case study.

Step 4: Measure against your cohort, not your fantasy

Every business wants 10:1 ROI. "For every pound I spend, I make £10 back."

Statistically, most marketing falls between 2:1 and 5:1 ROI. Some channels run hotter (referrals, proven SEO) and some colder (brand awareness, prospecting). If you're comparing your paid ads to someone's SEO, you're comparing apples to revenue-per-season-old-apples.

What matters is: are you profitable in your category?

• E-commerce: 2:1 to 4:1 is good, 5:1+ is great, under 2:1 is trouble.
• SaaS: 3:1 to 5:1 is acceptable, 6:1+ is great, under 3:1 needs fixing.
• B2B services: 4:1 to 10:1 is normal (longer sales cycle), under 4:1 is weak.
• Lead gen (selling leads): 8:1 to 15:1 is target, under 5:1 means you're not scaling.

The category matters because different businesses have different unit economics.

The ROI conversation that actually works

Here's what we ask clients when we're deciding if a channel is working:

Question 1: "What did you spend?" Often they don't actually know. "About £2,000 a month." Does that include management fees? Ad spend only? Design time? Get the real number.

Question 2: "How many leads came in?" Again, definition matters. Are we counting form fills or conversations? SQLs (sales qualified leads) or all leads? Get specific.

Question 3: "How many became customers?" This is where most conversations stall because they've never tracked it. But it's essential. You can't measure ROI without it.

Question 4: "What was the total revenue from those customers?" Not just first purchase - total revenue. If you're building customer relationships, a £200 customer today might be £2,000 by year three.

Question 5: "How much did the channel cost, start to finish?" Ad spend, yes. But also design, management, landing page building. Every pound.

Then the math: (Revenue from customers - total channel cost) ÷ total channel cost = ROI

That's it. If it's positive, it's working. If it's negative, it isn't.

What changes the math (and what doesn't)

Things that make ROI better:

• Better targeting — Fewer but higher-intent leads. Costs go down, conversion goes up.
• Improved offer — "Buy now" converts 1%. "Free trial, no credit card" converts 5%. Same traffic, 5x better ROI.
• Faster sales process — Taking 6 months to close instead of 2 months means less lead decay and higher conversion rates.
• Higher customer lifetime value — A customer worth £500 instead of £100 changes what you can afford to spend to acquire them.
• Volume — At small scale, paid ads are expensive per customer because you're competing with small CPCs. At £10k/month in spend, you get better unit economics.

Things that DON'T matter (but everyone obsesses over):

• Click-through rate — A 5% CTR is great. A 0.5% CTR is terrible. But if the 0.5% has a 50% conversion rate and the 5% has a 2% conversion rate, the "bad" channel is more profitable.
• Cost per click — £0.50 clicks look cheap. £5 clicks look expensive. But a £5 click with a 30% conversion to customer beats a £0.50 click with a 1% conversion every time.
• Website traffic — 100,000 visitors a month means nothing if they spend 2 seconds on your site and leave. 1,000 visitors spending 5 minutes each is often more valuable.
• Social media followers — Irrelevant unless they convert. A follower who never buys isn't a business asset; it's vanity.
• Email open rates — 30% open rate on emails that don't convert is worse than 15% open rate on emails that do.

Measure what drives profit. Ignore everything else, no matter how impressive it sounds.

ROI by channel: realistic ranges for 2026

These are real ranges we see for UK businesses, not fantasy numbers. We go deeper on the platform-level tradeoffs in Meta Ads vs Google Ads.

Paid Search (Google Ads)

• Cost per lead: £15 to £150 depending on sector
• Typical conversion rate (leads to customers): 8-20%
• ROI range: 2:1 to 6:1
• Timeline to profitability: Immediate (first month)
• Why it varies: Highly competitive sectors (legal, finance) run expensive. Niche B2B sectors run cheap. Conversion is your offer - good offer, good ROI.

Paid Social (Meta/Instagram)

• Cost per lead: £2 to £40 depending on targeting and creative
• Typical conversion rate: 5-15% (lower intent traffic)
• ROI range: 2:1 to 5:1
• Timeline to profitability: 2-4 weeks (after testing)
• Why it varies: Creative quality is everything. A great ad to a broad audience beats a mediocre ad to a perfect audience. Most businesses run mediocre ads.

SEO (Organic Search)

• Cost per lead: £0 to £20 (once established)
• Typical conversion rate: 15-35% (high intent, warm traffic)
• ROI range: 10:1 to 50:1+ (year 2 and beyond)
• Timeline to profitability: 6-12 months to break even, then exponential
• Why it varies: Competitive market = longer timeline. Low competition = months 4-5 profitability.

Email Marketing

• Cost per lead: £0.20 to £1.00
• Typical conversion rate: 2-8% (depends on list quality)
• ROI range: 30:1 to 100:1+
• Timeline to profitability: Immediate (first campaign)
• Why it varies: Only works if you have an email list. Building the list takes time/cost. Once built, it's the highest-ROI channel by far.

Content Marketing (Blog, Guides, etc.)

• Cost per lead: £10 to £100 (amortized across all content)
• Typical conversion rate: 5-25% (warm, educated leads)
• ROI range: 3:1 to 10:1+
• Timeline to profitability: 6-12 months
• Why it varies: Quality matters. A guide that answers real customer questions becomes an asset for years. Generic content is worthless.

The payback period: when does ROI actually matter?

Here's a number most businesses ignore: payback period. How many months until you recoup your marketing investment?

If you spend £5,000 on a campaign and make £5,000 in profit from it immediately, payback period is zero. Great.

If you spend £5,000 and make £1,000 a month profit for 6 months, payback period is 5 months. Acceptable.

If you spend £5,000 and make £200 a month profit, payback period is 2.5 years. That's only acceptable if the customer stays with you for 2.5+ years. If they leave after 1 year, you lost money.

This is where most small businesses get hurt. They invest in SEO (good idea), but set it up wrong (costly), take 12 months to see results (risky), and don't have enough cash left to compound it. Six months in they're out of money and kill the program right before it works.

If your channel can't hit those benchmarks, either fix it or kill it.

The two mistakes that destroy ROI

Mistake 1: Attributing credit to the wrong channel

A customer sees your social media ad on Monday. Forgets about you. Searches for you on Google on Thursday (Google Ads captures this). Buys on Friday. Which channel deserves credit?

Most analytics say Google, because it's the last click. But Meta created the awareness that made the Google search happen. Both deserve credit.

This is called attribution, and it's the reason most marketing ROI looks worse than it actually is.

Multi-touch attribution tries to split credit: Meta gets 40%, Google gets 60%. But the math is complicated and different tools do it differently.

Simple solution: Track the customer journey without obsessing over attribution. Ask yourself: "Would this customer have purchased if that channel didn't exist?" If no, that channel deserves credit.

Mistake 2: Conflating marketing ROI with business ROI

A business spends £10,000 on paid ads, gets 20 customers, and makes £50,000 revenue. That's 5:1 ROI on the ads.

But the business owner is frustrated because the company only made £10,000 net profit that year. The ads "worked," but the business is still barely profitable.

Why? Because revenue isn't profit. The customers cost money to deliver, to support, and to ship. The ad ROI was great. The business ROI was bad.

Marketing ROI and business ROI are not the same thing. Marketing ROI is important, but it's not the full picture. You need to know:

• Not just revenue per customer
• But profit per customer
• After all costs (COGS, support, returns, refunds)

Ask your accountant: what's the average profit margin per customer? That's the real number your marketing should optimise for.

The ROI dashboard that actually works

You don't need a fancy tool. A spreadsheet works fine. Here's what to track monthly:

You're looking for trends. Is cost per customer going down (good efficiency)? Is revenue per customer growing (good offer)? Is ROI trending upward or staying flat?

If it's flat or declining, something broke. Find it and fix it.

When to kill a channel (and when to double down)

Kill a channel when:

• It's been 3+ months and cost per customer is above your break-even point. (You're paying more to acquire than they're worth.)
• Cost per customer is increasing month over month. (Platform saturation or declining quality.)
• Payback period exceeds your cash runway. (You'll run out of money before the profit arrives.)
• A better channel exists. (Your budget is finite. Put it where ROI is highest.)

Double down when:

• Cost per customer is dropping month over month. (You're finding better audiences, better creative, or better targeting.)
• Payback period is under 30 days and shortening. (Efficiency is improving, scale is available.)
• ROI is 4:1 or better. (You're in profit territory with room to spend more.)
• It's your only channel and it's working. (Master one before adding complexity.)

One caveat: don't expect immediate perfection. Most channels need 4-8 weeks of testing before they find their groove. Kill it too early and you never let it work. Let it run too long and you waste money on something that won't.

The answer is: run it for long enough to get statistical significance (usually 100-200 conversions minimum), then decide.

Frequently asked questions

What's a "good" marketing ROI?

Context matters. E-commerce expects 2-4:1. B2B services expects 4-8:1. SaaS expects 3-6:1. If you're in year one with a new channel, 1.5:1 might be fine - you're building the asset. If you're in year three and still at 1.5:1, something's wrong.

How do I compare ROI across channels if they work differently?

Compare them on cost per customer and customer lifetime value, not on the headline metrics. £200 cost per customer from paid ads plus a 1-year average customer lifetime is different from £50 cost per customer from referrals but a 3-year average lifetime. Do the math on the full customer lifecycle.

Why is my marketing ROI positive but my business is losing money?

Your customer acquisition cost might be profitable, but your margins are being crushed by delivery costs, returns, or support. This is a business problem, not a marketing problem. Fix the margin first, then optimise marketing.

Should I measure ROI monthly or quarterly?

Monthly to spot trends and problems early. Quarterly to make big decisions (kill a channel, double down, pivot). Some channels (SEO, content) are too slow for monthly judgment.

What if I can't track leads back to their source?

Use UTM parameters on every link. Tag every lead form with the source. Ask every customer "how did you hear about us" at point of sale. You'll never have perfect data, but 80% accuracy is enough to make good decisions.

Is ROI the only number that matters?

No. Revenue concentration matters. If 80% of your ROI comes from one customer, you're not actually building a business - you're managing a client. Diversity of revenue sources is worth something.

How do I improve ROI without spending more money?

Improve offer quality (better positioning, better customer fit), improve conversion rate (landing page tests), or improve customer lifetime value (sell more to existing customers). All three improve ROI without increasing spend.

The one question that changes everything

Before you spend on marketing, ask: "What does a customer need to be worth for this channel to be profitable?"

Then ask: "Is that realistic for my business?"

If it's not, don't start. Fix the offer first, then run paid ads. Build the product first, then run paid ads. Get the sales process working first, then run paid ads.

Marketing amplifies a good business. It can't fix a bad one.