The Marketer’s Guide to Balancing Discovery and Intent: How to Build a Media Mix That Actually Works

I’ve watched countless brands make the same mistake by trying to treat media budget based on the “what have you done for me lately” principle. Throwing money at whatever channel hit last week or maybe even the last couple and praying it keeps working. Performance Marketing just doesn’t work that way.

What actually works? A structured approach to media buying and planning that understands the fundamental tension between Discovery channels (finding new audiences who don’t know they need you yet) and Intent channels (capturing people actively searching for what you sell).

These two types of channels serve completely different purposes, operate on different timelines, and need to be measured differently. Get this balance wrong, and you’ll either burn cash on awareness that never converts or slowly starve your pipeline by over-harvesting existing demand.

I’ve seen this first-hand with clients who either bring on an advisor or marketing strategist when there is no full-time marketing lead. The thought is (at least in their eyes) that Meta is way underperforming Google and we should cut the budget entirely or just maintain some level of remarketing. This ends up with a funnel that dries out sooner than they expected.

Understanding Your Channels: Discovery vs. Intent

Before we get into budget splits and metrics, we need to get crystal clear on what each channel actually does. This sounds basic, but I can’t tell you how many times I’ve seen companies try to treat Meta Ads the same way they treat Google Search.

Discovery Channels (Awareness/Demand Generation)

These are your top-of-funnel engines—platforms where people aren’t actively looking for you:

• Meta Ads (Facebook/Instagram)
• TikTok Ads
• YouTube Ads
• Display/Programmatic
• Influencer partnerships
• Podcasts and sponsorships

The goal: Create demand. Plant seeds in minds. Build mental availability so when someone eventually needs what you sell, you’re the first name that pops up.

Intent Channels

These capture existing demand from people already in-market:

• Google Search (Non-Brand)
• Bing Search
• Google Shopping
• Amazon Ads
• Retargeting (technically capture, not creation)

The goal: Harvest the demand you’ve created. Be there when someone raises their hand and says, “I need this.”

Here’s the critical insight most marketers miss: Intent channels look more efficient because they’re capturing demand Discovery channels created. If you cut your Discovery budget to fund more Search spend, you’ll see great numbers for a quarter—then watch your pipeline slowly dry up as you stop feeding the top of the funnel.

Google Search didn’t create that customer’s need. It just happened to be there when they went looking. The YouTube ad they saw three weeks ago? That’s what planted the seed.

The Starter Budget Split

You’ve probably heard of Les Binet and Peter Field’s research recommending a 60/40 split favoring brand over performance (IPA/Binet & Field research). That research is solid, but it’s based on large, established brands with the cash reserves and board patience to wait 12-18 months for brand equity to compound.

In my experience working with companies, you need to start heavier on Intent/Performance to prove the model and generate cash flow, then shift toward Discovery as you scale. This usually works with earlier stage companies because of the need to show value and show a return on ad spend.

Why? Because you need to:

• Demonstrate ROI before you have permission to invest in longer-term plays
• Generate the revenue that funds your Discovery experiments
• Build confidence with leadership that you know how to deploy capital efficiently

My recommended starting framework:

When to shift more toward Discovery:

The signal to push on discovery channels comes from your Intent channels, you can start shifting budget when you see:

• Search impression share maxed out: You’re already showing for 90%+ of relevant queries. There’s no more demand to capture.
• CPAs climbing despite optimization: You’ve squeezed efficiency gains, and now you’re just paying more for the same customers.
• Branded search flatlining or declining: Your pool of people actively looking for you isn’t growing. That’s a demand generation problem, not a capture problem.
• Retargeting list sizes shrinking: You’re not refilling the top of funnel.

When these signals appear, you’ve hit a ceiling and the only way to grow is to create new demand and build new audiences, which means investing in Discovery.

For context, the Fall 2024 CMO Survey found that marketers say their ideal split would be 50/50 between brand-building and short-term performance—but their actual allocation is 31.2% brand and 68.8% performance (The CMO Survey, Fall 2024). Most companies are already over-indexed on Performance. The question is whether that’s strategic or just default behavior because it’s easier to measure.

The Metrics That Actually Matter

Here’s where things get specific. You need both universal metrics across your entire media mix AND channel-specific KPIs that respect what each channel is actually trying to accomplish.

Universal Metrics (Track Across Everything)

1. Blended CAC and ROAS – Your overall efficiency. What’s your total revenue divided by total ad spend? What’s your all-in cost to acquire a customer? This is your north star that keeps individual channel metrics honest.
2. Marketing Efficiency Ratio (MER) – Total revenue divided by total marketing spend. This is your reality check that ignores attribution gaming. When platforms fight over who gets credit, MER tells you the truth.
3. New Customer Acquisition Rate – What percentage of conversions are genuinely new customers vs. repeat buyers? If your “efficient” campaigns are just re-converting existing customers, you’re not growing.
4. Time-to-conversion by channel – How long from first touch to purchase? Discovery channels have longer conversion windows. Judge them accordingly.

Channel-Specific Metrics

Discovery Channels:

Intent Channels:

A note on Meta metrics: CPM is a directional metric, not a success metric. It tells you about auction efficiency and audience targeting, but it doesn’t tell you if the campaign is working. A low CPM with terrible creative is worthless. Focus on cost-per-outcome metrics, with CPM as a secondary metric to help improve performance.

And notice I didn’t list “last-click conversions” as a primary metric anywhere. That’s intentional. Last-click attribution systematically over-credits Intent channels and under-credits Discovery. Use it as directional guidance.

Using Data to Adjust Your Budget (Without Losing Your Mind)

You don’t need a PhD in statistics or a six-figure measurement platform to make smart budget decisions. You need discipline, patience, and a few key practices.

Track Leading Indicators, Not Just Lagging Results

If you only measure conversions, you’ll always be reacting to what happened, not anticipating what’s coming. Build a dashboard that tracks:

• Branded search volume: Rising means your Discovery channels are working. Falling = trouble ahead.
• Direct traffic trends: Another signal of brand strength
• Email/SMS list growth rate: Future revenue you own
• Retargeting pool size: Is your top-of-funnel refilling?

When these leading indicators are healthy while your conversion metrics hold steady, you’re building durable growth—not just renting it.

Run Incrementality Tests

The most valuable question in marketing isn’t “which channel got the conversion?” It’s “would this conversion have happened anyway?”

Simple tests you can run:

• Geo Holdouts: Turn off a channel in 10-20% of your markets for 4-6 weeks. Measure the difference.
• Platform Lift Studies: Meta, Google, and TikTok all offer built-in measurement tools, use them.
• Spend scaling tests: Increase spend 30% in a channel for a month. Did incremental revenue increase proportionally, or did you hit diminishing returns?

These tests won’t give you perfect answers, but it’s better than staring at attribution dashboards or another set of reports and hoping they’re right. The truth is, they could be right or they could be wrong, depending on who you ask.

Respect the Funnel Math

Here’s a reality check most marketers skip. If your business has a 90-day sales cycle, your Meta ads from Q1 are driving Q2 revenue. Stop judging them on Q1 performance. Map out your typical customer journey:

• First touch to first site visit: __ days
• First site visit to email signup: __ days
• Email signup to purchase: __ days

Now look at your reporting windows and see if they are even close to aligned? If you’re judging YouTube on 7-day conversion windows when your average time-to-purchase is 45 days, you’re measuring the wrong thing.

Don’t Get Fooled by Randomness: Mistakes That Will Burn You

This is where I’ve seen the most careers derailed. Data is only as good as your interpretation of it and humans are terrible at interpreting data objectively.

Mistake #1: Week-Over-Week Optimization

This is such a killer and something I will never understand. Frankly it’s the reason we don’t have weekly standing meetings with clients because early on, I learned that some clients will want to kill keywords, creative, campaigns, etc. because they saw a 10% dip in a week and started hitting the panic button. Next week another one dips. Pause. Next week one does really well. Increase. I’ve seen this a few times and it’s something we just won’t allow anymore.

If you do this, you will optimize yourself into a corner based entirely on noise.

The reality for most digital marketing campaigns, a timeframe of just one or two weeks is insufficient to establish any meaningful pattern. Short timeframes don’t give you enough sample size, and random variation easily masquerades as a real trend (NoGood).

• The fix: Minimum 4-6 week windows for optimization decisions. For smaller accounts with less volume, extend to 8-12 weeks. I know your CMO wants answers now. Tell them that bad answers faster aren’t actually better.

Mistake #2: Ignoring Seasonality

Your February data is not comparable to your November data. Period.

If you’re making year-over-year comparisons without normalizing for seasonality, or worse, comparing this month to last month without context, you are hurting yourself because they data is not 1:1. You are making decisions in la la land.

Build a seasonality curve from 2+ years of data. Know which weeks are naturally strong and weak. Then you can judge performance and adjust budgets against that baseline vs. a lazy month-over-month trend.

Mistake #3: Over-Optimizing to Platform Metrics

Google tells you to maximize conversions. Meta tells you to optimize for purchases. Each platform is optimizing for its own success, not yours.

The platforms don’t know about:

• Your margins by product or segment
• Customer lifetime value differences by channel
• Brand perception and positioning
• Whether they’re cannibalizing organic or other paid channels

Build your measurement stack to answer your questions, not the platform’s questions.

Telling the Story: Reporting to Leadership

Here’s where technically excellent marketers often fail. You’ve done the work. You have the data. But if you can’t communicate it in a way that builds trust and drives decisions, it doesn’t matter.

Structure Reports Around Business Outcomes

People, especially executives care about KPIs that impact the bottom line. Revenue, CPA/ROAS, and MER should be what lead the conversation. CPCs and CTRs can be good secondary metrics but generally should be saved for marketing teams to help them understand their importance or relevance to campaign performance.

Things like:

• We generated $X in revenue at a blended CAC of $Y, compared to our target of $Z.
• Our Meta prospecting campaigns drove an estimated $X in new customer revenue, supported by a 15% increase in branded search volume during the campaign period.
• We ran a geo holdout test on YouTube and saw a 12% lift in conversion rate in test markets vs. control. We’re increasing investment next month.
• Based on our analysis, we’re shifting 15% of Display budget to CTV, with an expected incremental revenue impact of $X. We’ll validate with a 6-week test.

And, to be real, admitting what you don’t know builds more trust than pretending you have all the answers. If you don’t know how a test is going to perform or what your incrementality testing is going to yield, don’t make up a number. Try “Our best estimate for YouTube ROAS is between 2.4x and 3.2x based on our previous testing”

One Tip You Can Apply Today

If you take nothing else from this, here’s your actionable takeaway:

Calculate your Marketing Efficiency Ratio (MER) weekly for the past 90 days and look at the variance.

MER = Total Revenue ÷ Total Ad Spend

If your MER is stable (within 10-15% variance week over week), you can trust shorter-term data more. But even then, I’d recommend monthly optimization cycles at minimum.

The marketers who win long-term are the ones who resist the urge to tinker every week based on noisy data. They build a thesis about how their channels work together, test that thesis with discipline, and make decisions based on evidence rather than anxiety.

Balancing Discovery and Intent isn’t about finding the perfect formula. It’s about building a system that generates demand and captures it, measures both with appropriate patience, and improves quarter over quarter.