So, What’s the Real ROI of Performance Marketing Campaigns? A Complete Guide

When a marketing leader asks about the “real” ROI of a campaign, let’s be honest, they’re not just curious. They’re skeptical. They’ve seen inflated numbers and metrics that feel… misleading. To get to the truth of marketing return on investment, you have to dig way deeper than surface-level attribution and measure the actual incremental impact. This guide is all about that. It’s about how to calculate genuine ROI, dodge the common pitfalls that make your numbers look like a fantasy, and build a measurement framework that actually tells you if you’re making money. Whether you’re fighting for your budget in a boardroom or just trying to figure out where the hell to put the next dollar, mastering this stuff changes everything. It turns marketing from a line item expense into a proven revenue engine.

What Even Is Marketing ROI in Performance Marketing?

Marketing ROI is a simple metric on the surface. It’s about quantifying how profitable your campaigns are by weighing the revenue you brought in against what you spent. But in performance marketing, it gets specific. It measures the financial payback from campaigns built to make people do something *right now*—buy, sign up, hand over a qualified lead.

This isn’t like brand marketing, which is all fuzzy feelings about awareness and sentiment. No. Performance marketing ROI is tied directly to cold, hard business outcomes. This accountability is why it’s the number one justification for marketing budgets. It’s the foundation for making decisions with data instead of gut feelings. The metric answers the one question every CFO on the planet loves to ask: “For every dollar we give you, how many are you bringing back?”

Performance marketing lives in channels like paid search, paid social, display ads, affiliate marketing, and email. You know, the places where you can track every click, every conversion, every single customer you acquire. This incredible measurability is what lets you calculate ROI so precisely. But it’s a double-edged sword. It also opens the door to misleading metrics when people confuse correlation with causation, or forget to account for the sales that would’ve happened anyway, even if marketing took the day off.

And that’s where the distinction between reported ROI and real ROI becomes absolutely critical. Reported ROI loves to take credit for revenue that was already on its way. Real ROI? It isolates the *lift*. The sales that only happened because of what you did. Understanding that difference is what keeps you from pouring money down the drain and ensures it flows to the tactics that are actually working.

How to Calculate Marketing ROI: The Basic Formula

Okay, here’s the formula everyone knows: (Revenue from Marketing – Marketing Investment) ÷ Marketing Investment × 100. It spits out a percentage, telling you the return for every dollar you put in.

Let’s break it down. Simple stuff.

• Revenue from Marketing: Total cash you can directly attribute to your campaigns over a set time.
• Marketing Investment: The *complete* cost. Ad spend, yes, but also creative work, tech costs, and the portion of salaries for the people who ran it.
• Result: A percentage that shows your profit multiple.

Here’s a real-world example. A B2B software company drops $50,000 on a LinkedIn campaign aimed at enterprise clients. It pulls in 150 marketing qualified leads. From those, 15 become customers, and their average contract is worth $12,000. So, total revenue hits $180,000.

Plug it into the formula: ($180,000 – $50,000) ÷ $50,000 × 100 = 260% ROI

What this means is for every dollar they put in, they got $2.60 back. A 3.6:1 ratio. But—and this is a big but—this is just the starting point. This isn’t the *real* ROI yet. To get there, you have to tighten up the formula, include *all* the costs, and subtract the revenue that was going to happen anyway.

There’s another way to look at it, which focuses on net profit: (Gross Profit from Marketing – Marketing Investment) ÷ Marketing Investment × 100. This one is better, frankly. It accounts for the cost of goods sold (COGS), giving you a much clearer picture of actual profitability. Super important for eCommerce businesses where margins are all over the place.

And for subscription businesses? Forget first-purchase value. You have to think in customer lifetime value (CLV). A SaaS company might be deep in the red on their first-month ROI, but once you look at the ROI over a 12-month customer lifespan, it’s a completely different—and much happier—story.

The Essential Metrics That Actually Drive ROI

You can’t get to an accurate ROI without understanding the building blocks. There are six critical metrics, and each one tells a different part of the story about how efficient and effective your campaigns are.

Customer Acquisition Cost (CAC) is just what it sounds like: how much you spend, on average, to get one new customer. You calculate it by dividing your total marketing and sales costs by the number of customers you won. Spend $100,000 to get 500 customers? Your CAC is $200. This is the “investment” side of your ROI equation, plain and simple.

Customer Lifetime Value (CLV) is the total amount of revenue you can expect from a single customer over the entire time they’re with you. The math is: Average Purchase Value × Purchase Frequency × Average Customer Lifespan. So if a customer spends $50, buys 4 times a year, and sticks around for 3 years, their CLV is $600. The golden rule? Your CAC to CLV ratio needs to be 1:3 or better. If it’s not, you’re on a path to going out of business.

Conversion Rate. This one’s obvious—it’s the percentage of people who do what you want them to do. Buy a thing, sign up for a newsletter, request a demo. Higher conversion rates make your ROI sing because you’re squeezing more value out of the same spend. Just improving your conversion rate from 2% to 3% boosts revenue by 50% without spending a single extra dime on ads.

Return on Ad Spend (ROAS) is ROI’s little brother. It specifically measures revenue generated per dollar of ad spend: Revenue from Ads ÷ Ad Spend. A 5:1 ROAS means $5 in revenue for every $1 on ads. It’s useful for comparing ad channels against each other, but it’s not the whole story because it ignores all the other costs like creative and tech fees.

Cost Per Acquisition (CPA) measures the cost for one conversion *action*—which isn’t always a customer. For a lead gen campaign, it tells you how efficiently you’re making leads. A webinar that cost $5,000 and got you 200 leads has a $25 CPA. Is that good? No idea. You have to compare it to your lead-to-customer rate and deal size to know if it’s profitable.

Click-Through Rate (CTR) is your early-warning system. It’s just clicks divided by impressions. On its own, CTR doesn’t mean squat for ROI. But it’s a good gauge of whether your message is hitting the mark. Low CTR means you’re wasting money on impressions; high CTR suggests you’ve got a good audience-message fit, which *usually* leads to better ROI down the line.

The “Real” ROI: Let’s Talk About True Incremental Impact

Okay, this is it. The most important part of this whole discussion. Real marketing ROI isolates the incremental revenue specifically *caused* by your marketing, not just correlated with it. This distinction is everything. Why? Because standard attribution models are liars. They give marketing 100% credit for conversions that were probably going to happen anyway thanks to brand recognition, word-of-mouth, or just existing customer loyalty.

Think about a retailer running paid search ads on their own brand name. Someone clicks the ad, buys something. Boom, the attribution platform chalks up a win for the campaign. But wait. Many of those people were already looking for the brand *by name*. They would have found the website organically if the ad wasn’t there. The marketing spend didn’t *create* new revenue; it just intercepted demand that already existed. What a waste.

This mess is called attribution bias, and it systematically inflates your reported ROI. Marketing teams pop the champagne over amazing metrics while the C-suite scratches their heads, wondering why the company’s overall growth doesn’t seem to match the marketing PowerPoints. That gap between attributed revenue and real, incremental revenue? That’s almost always the culprit.

The only way to find the truth is incrementality testing. It’s the gold standard. You compare results between a group that saw your marketing and a control group that didn’t. For instance, a streaming service shows ads to people in Chicago but holds back ads from a demographically similar group in Houston. Then they compare sign-up rates. That’s how you find the true lift.

Here’s how it plays out: If Chicago (with ads) gets a 5% conversion rate and Houston (no ads) gets a 3.5% conversion rate, the *real*, incremental lift is 1.5 percentage points. Not 5%. Your ROI calculation should only use the revenue from that 1.5% lift ($150,000), not the total attributed revenue ($500,000). You can see how that dramatically changes the entire profitability equation.

How do you do this in the real world?

• Geographic holdout tests: Just like the Chicago/Houston example. Run campaigns in some places, hold back others.
• Audience split tests: Randomly divide your audience into a test group (sees ads) and a control group (doesn’t).
• Time-based experiments: Turn campaigns on and off, and measure performance during the “on” periods against the “off” baseline.
• PSA (Public Service Announcement) tests: Show your control group a non-commercial ad to make sure you’re accounting for the effect of simply seeing *an* ad.

Without this kind of testing, you’re just flying blind, optimizing for vanity metrics that make you look good but don’t grow the business. I’ve seen it a dozen times. Incrementality measurement is what shows you which investments are actually moving the needle and which are just taking credit for a parade they didn’t organize.

And the difference is massive for your budget. A campaign might scream 400% ROI in your dashboard, but an incrementality test could reveal the real ROI is only 150%. It might still be worth running… but you sure as hell aren’t going to pour more money into it based on that inflated number.

Understanding Attribution Models (So You Know How You’re Being Fooled)

Attribution models are the rules that decide who gets credit for a conversion. They directly shape the ROI you see for every channel and campaign. Pick the wrong one, and you’ll get a completely warped view of what’s effective, leading to some truly terrible budget decisions.

The modern customer journey is a mess. It can take days, weeks, months. A B2B buyer might see a LinkedIn ad, read a few blog posts, go to a webinar, download a whitepaper, and finally convert from a sales email. So… which one gets the credit? The answer changes completely depending on the model.

First-Touch Attribution gives 100% of the glory to the very first interaction. It makes top-of-funnel channels like display ads and social media look like heroes. It’s good for seeing what starts a relationship, but it completely ignores all the work that goes into actually closing the deal.

Last-Touch Attribution (the default for most platforms, and usually the most misleading) gives all the credit to the final touchpoint. This model makes bottom-funnel stuff like branded search and remarketing look amazing, while ignoring everything that made the customer ready to convert in the first place. It encourages you to just capture existing demand, not create it.

Linear Attribution is the “everyone gets a trophy” model. It splits credit equally among all touchpoints. Five interactions? Each gets 20%. It at least acknowledges the whole journey, but it’s naive to think a quick website visit is as valuable as a 60-minute product demo.

Time-Decay Attribution gives more credit to touchpoints that happen closer to the conversion. This makes a lot of sense, as recent interactions probably have more influence. It still gives some credit to earlier activities, but weights it less. Pretty useful for campaigns with a clear consideration phase.

Position-Based (U-Shaped) Attribution is a nice compromise. It gives 40% of the credit to the first touch, 40% to the last touch, and splits the remaining 20% among all the interactions in the middle. It values both starting the conversation and closing the deal.

Data-Driven Attribution is the holy grail. It uses machine learning to figure out which touchpoints actually contributed the most by comparing converting paths to non-converting paths. It’s the most accurate, but you need a *ton* of data for the algorithms to work properly.

There’s no single perfect model. The right choice depends on your sales cycle, how many touchpoints people typically have, and what your business goals are. B2B companies with long cycles probably get the best view from position-based or data-driven models. A direct-to-consumer eCommerce store might be fine with last-touch for its simplicity.

But here’s the critical thing to remember: every single one of these models, except for incrementality testing, measures *correlation*, not *causation*. Just because a touchpoint happened before a sale doesn’t mean it *caused* the sale. The smartest marketers use attribution to understand the journey, then layer on incrementality testing to understand the real impact.

Common ROI Calculation Mistakes That Will Wreck Your Results

Even people who know the formulas screw this up. These are the mistakes I see all the time, the ones that inflate numbers and lead to really bad strategic calls. Avoiding these will instantly give you more credibility with your finance team.

Mistake #1: Ignoring Soft Costs and Overhead

This one drives me crazy. So many ROI reports only count ad spend. They just conveniently forget about the cost of making the creative, the marketing tech subscriptions, agency retainers, and employee salaries. A campaign might look like a 500% ROI winner when you only count the $20,000 in ad spend. But when you add the $10,000 for creative, $5,000 in tech fees, and $15,000 in allocated labor (total spend: $50,000), that ROI suddenly drops to 200%. Big difference.

The fix? Build a framework that captures everything. All of it. Platform fees, data costs, A/B testing tools, your attribution software, and yes, a percentage of your team’s salaries.

Mistake #2: Using Dumb Timeframes

Measuring ROI too soon is a classic way to kill a perfectly good campaign. It’s especially bad for products with long sales cycles. That B2B software campaign might look like a total failure after 30 days, but if you measure it after 90 days (which matches their sales cycle), it’s suddenly a huge success.

The fix is simple: match your measurement window to your average sales cycle. For lead gen, you have to track cohorts of leads all the way to closed deals. For awareness plays, you need to be looking at 6-12 month windows to see the delayed impact.

Mistake #3: Crediting All Revenue Instead of Incremental Revenue

This is the big one. The most common and most expensive mistake of them all. You can’t give marketing credit for all attributed revenue without thinking about your baseline. A retailer might brag that their holiday email campaign drove $500,000. But if they did $400,000 in holiday sales last year *without* that campaign, the real incremental impact is only $100,000. Not $500,000.

The fix: Always establish a baseline. Use historical data, control groups, whatever you have. Calculate your ROI on the *incremental* revenue only. That’s the only number that reflects marketing’s true contribution.

Mistake #4: Forgetting About Customer Lifetime Value

If you only look at the first purchase, you’re undervaluing your customer relationships and making acquisition look way less profitable than it is. A subscription business might lose money on the first month. But when you look at the 300% ROI they get over a 24-month customer lifespan, you realize it’s a brilliant investment.

The fix: Use CLV in your ROI calculations, not just the initial sale. This is non-negotiable for subscription models or any business with high repeat purchase rates. You have to look at cohorts to understand real lifetime value.

Mistake #5: Double-Counting Conversions Across Channels

This happens when every channel uses last-touch attribution and they all try to claim credit for the same conversion. If you just add up the ROI from each channel report, you’ll get a total that’s bigger than the actual revenue the company made. It’s a mathematical impossibility that reveals your tracking is broken, but it often slides by until finance tries to reconcile the numbers.

The fix: Use a unified attribution platform that de-duplicates conversions. And stop letting every channel claim 100% of the credit for the same sale. Use a multi-touch model that spreads the credit around more intelligently.

Channel-Specific ROI Measurement Nuances

Different channels do different jobs, so you can’t measure them all the same way. You have to tailor your approach. Understanding their quirks stops you from misjudging their value and screwing up your budget.

Email marketing has an almost comically high ROI. Industry averages of $36-$42 for every dollar spent are actually real. It’s because the cost to send is so low and you’re talking to a warm audience. Measuring email ROI is pretty direct with tracking and pixels, though it gets tricky when you try to figure out how it works with other channels.

Paid search advertising is transparent. You spend money, you see results. Google Ads campaigns going after high-intent keywords can pull in 200-500% ROI for eCommerce and even 300-800% for high-margin services. But—and I’ve said it before—branded search is often just harvesting existing demand. You have to use incrementality testing to see if it’s actually creating new business.

Social media advertising on platforms like Facebook, Instagram, or LinkedIn is all over the map. ROI depends entirely on your targeting and creative. B2C can see 150-400% ROI, and B2B might get 200-600% if they target decision-makers well. Social is great for finding specific people, but you need long attribution windows because it often starts a journey that finishes somewhere else.

Content marketing and SEO are a nightmare to measure. The payback is delayed, and the attribution is fuzzy. Organic search is insanely cost-effective once you get it going, but your ROI math has to account for months or even years of investment before the traffic really shows up. Mature content programs, though? They can generate 500-1000%+ ROI as traffic compounds with very little ongoing cost.

Display advertising and programmatic usually have lower direct ROI (think 100-300%), but that’s not their main job. Their job is awareness and retargeting. You have to measure display on assisted conversions and view-throughs, not last-click. These campaigns are what make your higher-ROI channels like search and email work so well.

When you look at everything together, you realize channels don’t work in a vacuum. They’re a system. The combined ROI is almost always greater than the sum of its parts because awareness channels feed the consideration channels that set up the conversion channels for success.

Overcoming the Hurdles of ROI Measurement

Even the most sophisticated marketing teams run into roadblocks with this stuff. Knowing the challenges—and how to deal with them—is key to getting practical, useful ROI tracking in place, even if it’s not perfect.

Data fragmentation is the monster under the bed for every marketer. Your data is everywhere: ad platforms, web analytics, CRM, marketing automation… and none of them talk to each other properly. Each has its own definitions and tracking. Without a way to connect them, a complete ROI calculation is just a pipe dream.

Solution: Get a customer data platform (CDP) or some other tool to unify your data. And for God’s sake, create consistent UTM standards and naming conventions. You need a single source of truth for your key metrics, not ten conflicting reports.

Long B2B sales cycles are a pain because they separate the marketing action from the revenue event. A deal might close a year after the first touchpoint. How do you connect that campaign to the revenue? It takes serious tracking and a lot of patience.

Solution: Don’t just measure closed deals. Measure marketing’s contribution to pipeline generation and acceleration. Use cohort analysis to follow leads all the way through the sales cycle, even if it crosses into the next fiscal year. Calculate a preliminary ROI on pipeline value and then update it as deals close.

Privacy regulations and tracking limitations are making our jobs harder. It’s the truth. GDPR, CCPA, Apple’s ATT… they are all chipping away at our ability to track individual users across the web.

Solution: We have to adapt. Shift to privacy-friendly methods like aggregated lift studies, marketing mix modeling, and focusing on your first-party data. Double down on owned channels where you can still track things, and use statistical models to estimate the impact of the channels you can’t see as well.

Multi-device customer journeys mean our tracking is fragmented. People browse on their phone, research on their laptop, and buy on their desktop. Or they see an ad online and buy in a store. Old-school cookies can’t connect those dots.

Solution: Get users to log in. Authenticated experiences let you track people across devices. You can also use probabilistic device graphing. But honestly? You have to accept a certain amount of uncertainty. Chasing perfect precision is a fool’s errand in today’s world.

Tools and Platforms for ROI Tracking

The modern martech stack has tools built for this. Marketing analytics platforms like Google Analytics 4, Adobe Analytics, and HubSpot can pull together cross-channel data into ROI dashboards. They connect what you’re doing to what happens on your website and—if you integrate them with your CRM—to actual revenue.

Attribution platforms like Bizible, Dreamdata, and HockeyStack are specialists. They live and breathe multi-touch attribution, offering all the models from first-touch to data-driven. They plug into your ad channels and CRM to map out the entire customer journey.

Customer data platforms (CDPs) like Segment and mParticle are the foundation. They unify all your customer data into single profiles, which makes accurate attribution and CLV calculation possible. A good CDP ensures all the other tools in your stack are working with consistent, clean data.

What’s a Good ROI for Performance Marketing?

Okay, the number you’re all here for. A 5:1 marketing ROI (that’s a 400% return) is generally considered really strong in most industries. It means for every $1 you spend, you get $5 back in revenue. But—and you knew there was a but—what counts as “good” is wildly different depending on your business model, margins, and goals.

For direct response eCommerce, benchmarks are usually somewhere between 200-400%. The market is crowded, margins can be thin. A 300% ROI is a solid win. Anything under 150% and you should probably be worried.

B2B SaaS companies play a different game. They’re often aiming for a much higher 500-1000% ROI because their customer lifetime values are so high. That said, they’ll often accept negative ROI in the short term if they’re in a growth phase, prioritizing grabbing market share over immediate profit.

Professional services and consulting firms tend to see 300-700% ROI. They have high margins and strong referral networks that amplify the initial marketing investment.

Honestly, context is everything. A huge, established brand might be thrilled with 200% ROI on an awareness campaign. A cash-strapped startup might need 600% ROI just to survive. A campaign to enter a new country will probably have lower ROI (150-250%) than one in an established market (400-600%+) because you’re starting from scratch.

The sales cycle changes everything. Short-cycle businesses (under 30 days) need to see that ROI fast. Long-cycle businesses (90+ days) have to be patient and measure over longer periods; their short-term ROI might look terrible even when the long-term return is fantastic.

But the most important benchmark isn’t some industry average. It’s your own historical performance. If you’ve been cruising along at 350% ROI and it suddenly jumps to 600%, you should probably check if your tracking broke. If it drops to 200%, you might have a new competitor or a saturated market on your hands.

Industry-Specific ROI Considerations

Let’s get specific. B2B companies with deals over $50,000 are measuring ROI over 6-12 months. They’re okay with high customer acquisition costs (maybe 30-40% of the first year’s contract) because they know a customer will stick around for years. Their ROI math absolutely must include expansion revenue and retention rates.

B2C eCommerce lives and dies on tight margins and short timelines. They need to see that ROI *now*. They’re obsessed with ROAS (Return on Ad Spend) measured over 7-14 day windows. A successful campaign needs to hold a 3:1 to 5:1 ROAS ratio to be viable.

Financial services is a whole other beast. They’ve got compliance costs and long sales cycles. For insurance or banking, you might be looking at 18-24 month cycles but with huge customer lifetime values. Getting CLV right is the only way to make a realistic ROI assessment.

Strategies to Actually Improve Your Marketing ROI

You want better ROI? It’s not magic. It’s systematic optimization. Audience segmentation is your biggest lever—I mean it. Targeting smaller, more defined audiences can boost conversion rates by 30-50% because your message is just more relevant. You stop wasting money on people who were never going to buy anyway.

Conversion rate optimization (CRO) is like getting free money. You improve ROI without spending more on ads by getting more value from the traffic you already have. A/B testing your landing pages, your CTAs, your checkout process… these things can lift conversion rates by 15-40%. If you take a campaign from a 3% to a 4% conversion rate, you’ve just increased your ROI by 33% without touching your ad budget.

Reallocate your budget based on marginal ROI. This means constantly shifting money from your worst performers to your best. Most companies find that 20% of their campaigns drive 80% of their ROI. So, move the money! But be aware of diminishing returns. You can’t just pour infinite money into your top channel; the 100th customer always costs more to acquire than the 10th.

Retargeting and nurture sequences are your secret weapons. They improve ROI by converting warm leads who are cheaper to acquire than cold ones. Someone who visited your site but didn’t buy can be retargeted for 50-70% less than what it cost to get them there the first time. Email nurture sequences can guide a lead to conversion for almost no additional cost.

Creative testing and refresh cycles are non-negotiable. People get sick of seeing the same ad. It’s called ad fatigue, and it can tank your ROI by 20-40% in just a month or two. You need to constantly refresh your creative and test new messaging to see what works.

None of these things work in isolation. It’s like making a good soup—you need all the ingredients. But the compound effect is incredible. A campaign at 200% ROI that gets a targeting boost (1.3x), a CRO lift (1.25x), and better creative (1.15x) doesn’t just add up. It multiplies. You end up with 469% ROI.

Conclusion

So, the real ROI of performance marketing. It’s not in your dashboard. It’s buried under layers of attribution assumptions and incomplete cost accounting. You have to find the true, incremental impact. To do that, you need to track every single cost, pick the right attribution model (while knowing its flaws), run incrementality tests, and stop making the common calculation mistakes. The gap between what your dashboard says and what’s actually happening often determines whether your marketing spend is driving real growth or just taking credit for things that were going to happen anyway. Start here: audit your current ROI calculations for the mistakes in this guide. Then, pick your biggest campaign and run an incrementality test. That’s how you find your baseline, that’s how you find your true impact. And that’s how marketing stops being a cost center and becomes the growth engine it’s supposed to be.

Frequently Asked Questions

What’s the difference between ROI and ROAS?

Think of it this way: ROI is the big picture, ROAS is a close-up. ROI (Return on Investment) measures your total profitability by looking at revenue versus *all* your marketing costs—ad spend, people, tech, creative, everything. ROAS (Return on Ad Spend) just looks at revenue versus what you spent on ads. So ROAS is good for judging if your ads are efficient, but ROI tells you if you’re actually making money. A campaign can have a great 5:1 ROAS but a weak 2:1 ROI once you factor in all the other costs.

How long should I wait before calculating ROI on a campaign?

The simple answer? Your measurement window should match your sales cycle. If you’re an eCommerce store where people buy right away, you can measure ROI in 7-14 days. But if you’re a B2B company with a 90-day sales cycle, you need to wait at least 90-120 days. For stuff like content marketing, you need to look at 6-12 month windows. If you measure too early, you’ll kill good campaigns; measure too late, and you can’t optimize fast enough.

What costs really need to be in my marketing ROI calculation?

All of them. Seriously. Direct costs like ad spend and agency fees are obvious. But you also need creative costs (design, copy, video), technology costs (your automation platform, analytics tools), and a portion of your team’s salaries based on how much time they spent on the campaign. Some companies even add a bit of overhead. If you leave out the “soft” costs like creative and tech, your reported ROI will probably be inflated by 30-50%. It’s a fantasy number.

So what’s a good ROI for performance marketing?

A 5:1 ratio (400% ROI) is a great benchmark, but “good” really depends. B2C eCommerce is often happy with 200-400%. B2B SaaS might shoot for 500-1000%. Professional services, maybe 300-700%. But more important than those averages is how you’re doing against your own history, and whether your customer acquisition cost is way lower than your customer lifetime value. A 3:1 CLV to CAC ratio is a healthy place to be.

How do I measure the *true* incremental impact of my marketing?

Incrementality testing is the only real way. You have to compare a group that sees your marketing to a control group that doesn’t. You can do this with geographic holdouts (run a campaign in one city but not another), audience split tests (randomly divide your target list), or just turn campaigns on and off. The difference in performance between the test group and the control group is your *real* lift. Anything else is just relying on attribution, which can’t tell you if you actually *caused* the sale.

Which attribution model is the most accurate for ROI?

If you have enough data (we’re talking thousands of conversions a month), a data-driven attribution model that uses machine learning will give you the most accurate picture. If you don’t have that kind of volume, a position-based (U-shaped) model is a decent bet for complex B2B sales—it gives credit to the first and last touches, and some to the middle. But I’ll say it again: every attribution model just shows correlation, not causation. Only incrementality testing gets you to the real, causal impact.