MER, Measurement & Attribution

Expert synthesis: Taylor Holiday (Common Thread Collective) · Cody Plofker · Andrew Faris (4× CFO) · Zachary Murray (Triple Whale) · Northbeam team · Rockerbox team · John Cavendish · Ryan McKenzie (Tru Earth) · Jordan West · Richard Gaffin · Stephanie Liu · Michael Pelpola

---

#THE FUNDAMENTAL PROBLEM WITH E-COMMERCE MEASUREMENT

Here is what most e-commerce operators track:

• Platform-reported ROAS (Meta says 3.2×, TikTok says 2.8×, Google says 4.1×)
• Revenue in Shopify

Here is what that actually tells you: Nothing.

Here is why:

1. Multi-touch journeys: Customer sees TikTok ad → sees Instagram Story → Googles your brand → clicks Google ad → buys. Google claims 100% credit. Meta claims 100% credit. TikTok claims 100% credit. Total claimed conversions = 300% of actual.

2. iOS14+ tracking collapse: Apple's App Tracking Transparency removed identifiers. Meta's Pixel now sees approximately 40–60% of actual conversions. The 3.2× ROAS Meta shows you is calculated on incomplete data.

3. View-through attribution: Meta default attributes a sale if a user saw your ad (even without clicking) within 1 day. This creates massive false attribution for retargeting campaigns.

4. Cross-device journeys: Customer browses on phone, buys on laptop. Most platforms can't track this stitching.

The result: Every platform's dashboard shows ROAS that is meaningfully higher than reality. Operators who optimise for platform ROAS are making budget decisions on fiction.

---

#SECTION 1: THE MER FRAMEWORK (MARKETING EFFICIENCY RATIO)

#1.1 What MER Is and Why It's the Master Metric

MER = Total Revenue ÷ Total Ad Spend (all channels combined)

Also called "blended ROAS" or "true ROAS."

Taylor Holiday (CEO of Common Thread Collective, managing $100M+ in ad spend): "MER is the most important number in your business. Everything else is a supporting metric."

The difference between platform ROAS and MER:

Important: MER of 3.4× does NOT mean you're profitable.

MER tells you efficiency of spend. Profitability depends on your margins.

Breakeven MER = 1 ÷ Contribution Margin (before ads)

If your contribution margin before ads is 45%:
Breakeven MER = 1 ÷ 0.45 = 2.22×

Target MER = Breakeven MER × 1.5 = 3.33× (for 50% cushion)
At MER 3.4× with 45% contribution margins: You are profitable.

---

#1.2 Building Your MER Dashboard

The daily MER tracker (Google Sheets):

DATE | TOTAL REVENUE | META SPEND | TIKTOK SPEND | GOOGLE SPEND | EMAIL SEND | TOTAL SPEND | MER
-----|-------------|-----------|-------------|-------------|-----------|------------|----
6/01 | $8,240      | $1,500     | $500        | $300        | $0        | $2,300     | 3.58×
6/02 | $6,180      | $1,500     | $500        | $300        | $800*     | $3,100     | 1.99×
6/03 | $10,450     | $1,500     | $500        | $300        | $0        | $2,300     | 4.54×

*On day 2 you sent an email campaign — that's why MER looks terrible. Email spend inflated total spend, email-driven revenue attributed to organic. Always note send days.

Rules:

• Calculate MER daily, review weekly average, compare to 30-day rolling average
• Email send days distort MER — flag them separately
• Sale days distort MER positively — flag them
• Slow shipping periods distort revenue timing — track order date, not ship date

---

#1.3 New Customer MER (NC-MER) — The Growth Indicator

Regular MER includes repeat purchasers — people who were already going to buy from you regardless of ad spend.

NC-MER (New Customer MER) = New Customer Revenue ÷ Total Ad Spend

This is the metric that tells you if your brand is actually growing.

If your MER is 3.5× but your NC-MER is 1.8×, you are:

• Spending heavily to reach existing customers who would have bought anyway
• Running retargeting too aggressively
• Not acquiring new customers efficiently
• Your growth is decelerating

Taylor Holiday's NC-MER benchmark for growing brands: >2× in year 1, >2.5× in year 2+

How to track NC-MER:

• Klaviyo segments: "Customer placed first order > 365 days ago" = existing customer
• Shopify: Reports → Customers → New vs. Returning
• Triple Whale: Tracks this natively in the Customer dashboard

---

#SECTION 2: ATTRIBUTION TOOLS AND HOW TO USE THEM

#2.1 The Attribution Tool Landscape

First-party tools (your own data):

• Google Analytics 4 (free, data-model based attribution)
• Shopify Analytics (basic, last-click only)

Third-party attribution tools:

• Triple Whale — Best for $1k–$50k/month ad spend. Native Shopify integration. Pixel + server-side tracking. Great UI.
• Northbeam — Better for $50k+/month spend. More sophisticated ML attribution models. Higher learning curve.
• Rockerbox — Mid-tier, good for multi-channel brands with Google/Meta/TikTok/TV.
• Elevar — Best tracking layer (sits under the attribution tools, improves data quality)

Which to use:

• Under $10k/month spend: Google Analytics 4 (free) + MER dashboard (spreadsheet)
• $10k–$50k/month: Triple Whale ($129–$500/month)
• $50k+/month: Northbeam or Rockerbox + Elevar

---

#2.2 How to Set Up Triple Whale (The Complete Guide)

Triple Whale is the most widely used attribution tool in e-commerce under $50k/month.

Setup:

1. triplewhale.com → Connect Shopify (1-click OAuth)
2. Connect Meta Ads account
3. Connect TikTok Ads account
4. Connect Google Ads account
5. Install Triple Whale Pixel on Shopify (Settings → Customer Events → Triple Whale)
6. Configure post-purchase survey (critical — see below)

The post-purchase survey: Triple Whale (and Northbeam) include a "How did you hear about us?" survey in the order confirmation or thank you page.

This is your most accurate attribution data. It's self-reported but unfiltered by algorithm bias.

Survey options to include:

• TikTok (ad or organic)
• Instagram (ad or organic)
• Facebook (ad or organic)
• Google search
• A friend/referral
• YouTube
• I'm a returning customer
• Other

What operators do with this data: Compare self-reported attribution to platform attribution. The gap between "Meta claims X% of purchases" and "X% of customers say they came from Meta" is your attribution inflation indicator.

Industry benchmarks: Meta over-claims by 30–60%. Google over-claims by 20–40%.

---

#2.3 The Attribution Frameworks Explained

Last-click attribution: 100% credit to the last ad clicked before purchase. Very inaccurate for brands with multi-touch journeys. Do not make decisions from last-click alone.

Data-driven attribution (DDA): Google Analytics 4's default model. Uses machine learning to distribute credit across touchpoints based on conversion probability. Better than last-click but still limited by tracking gaps.

First-party data model (Triple Whale Pixel): Triple Whale's model attributes based on your own first-party data collected via their Pixel. More accurate than platform pixels post-iOS14.

Linear attribution: Equal credit to every touchpoint. Useful for understanding influence but not for budget allocation.

Position-based (W-shaped): 40% to first touch, 40% to last touch, 20% distributed to middle touches. Good for understanding customer journey.

The operator's approach: Use a combination — no single model is right.

1. MER for budget-level decisions (how much to spend overall)
2. Triple Whale's "Pixel" attribution for channel-level decisions (Meta vs. TikTok vs. Google)
3. Post-purchase survey for reality-checking platform over-attribution
4. Cohort analysis for understanding actual customer LTV

---

#SECTION 3: COHORT ANALYSIS — UNDERSTANDING YOUR BUSINESS

#3.1 What Cohort Analysis Is (And Why Most Operators Ignore It)

A cohort = a group of customers who made their first purchase in the same time period.

Cohort analysis asks: "Of the customers who first bought in January, how much total have they spent by June?"

This is the only way to understand:

• True LTV over time
• Whether LTV is improving or declining
• Whether new cohorts are as good as old cohorts (critical for scaling)
• Payback period for customer acquisition cost

---

#3.2 The Cohort Revenue Table

Build this in Google Sheets. Pull customer order data from Shopify (Reports → Customers).

COHORT TABLE (Monthly Revenue per Cohort)

          Month 1   Month 2   Month 3   Month 6   Month 12
Jan cohort $45,000  $12,000   $8,000    $5,000    $3,000
Feb cohort $51,000  $14,500   $9,200    (pending)
Mar cohort $48,000  $13,000   (pending)
Apr cohort $52,000  (pending)

Jan cohort 12-month LTV = $45,000 + $12,000 + $8,000 + $5,000 + ... 
(divide by number of customers in cohort = revenue per customer)

What to look for:

• Month 2 revenue as % of Month 1: Your 30-day repeat rate. Target: 20–40% depending on category.
• Month 6 total as % of Month 1: Your 180-day LTV multiple. Target: 1.3–2.0×.
• If Month 2 / Month 1 ratio is falling: Your product or post-purchase experience is getting worse.
• If February cohort LTV < January cohort LTV: The customers you're acquiring are lower quality (possibly a targeting or offer issue).

---

#3.3 LTV:CAC Ratio — The Exit Multiplier

LTV = Lifetime Value (total revenue per customer across all purchases) CAC = Customer Acquisition Cost (total ad spend ÷ new customers)

LTV:CAC targets:

CAC Payback Period: How many months until you've recouped the cost of acquiring a customer in gross profit?

CAC Payback Period = CAC ÷ (Monthly Revenue per Customer × Contribution Margin %)

Example: CAC = $45, Monthly avg spend after acquisition = $25, CM = 40%
Payback = $45 ÷ ($25 × 0.40) = $45 ÷ $10 = 4.5 months

Target: Under 6 months for sustainable growth. Under 3 months = exceptional.

---

#SECTION 4: INCREMENTALITY TESTING

#4.1 Why Incrementality Testing Changes Everything

The ultimate question in paid advertising: "Would these customers have bought even if I hadn't run this ad?"

This is the question that attribution models cannot answer. They can only tell you which ads were seen or clicked. They cannot tell you if those ads caused the purchase.

Incrementality testing creates holdout groups — control audiences who never see your ads — to measure the causal effect of your advertising.

The result: You discover your "true ROAS" — not what the platform claims, but what revenue actually lifts because of the spend.

For many brands: Reported ROAS is 3.5×, Incremental ROAS is 2.1×. The difference is purchases that would have happened anyway (organic, email, brand search, repeat customers).

---

#4.2 How to Run a Simple Holdout Test

Meta Holdout Test (use Meta's built-in tool):

1. Meta Ads Manager → Experiments → Brand Lift or Conversion Lift
2. Select your campaign
3. Choose holdout size: 10–20% of audience (they see no ads)
4. Run for 2–4 weeks minimum
5. Meta reports: incremental purchases, incremental ROAS

Manual holdout test (simpler, any platform):

1. Define test: Run ads in cities A, B, C. Don't run in city D (control).
2. Compare revenue per capita in test cities vs. control city during the test period.
3. If test cities grew 20% more → your ads are driving incremental growth.
4. If growth is equal → your ads are not incrementally driving purchases.

What to do with incrementality data:

• If incremental ROAS < 1.5×: The channel is not efficiently driving new purchases. Reduce spend, reallocate.
• If incremental ROAS > 2.5×: The channel is genuinely driving incremental growth. Scale.
• Use incrementality results to weight attribution models (reduce platform-reported ROAS by gap %)

---

#SECTION 5: THE WEEKLY OPERATING RHYTHM

#5.1 The Metrics Review Calendar

Daily (5 minutes):

• Yesterday's revenue vs. 7-day average
• MER (calculate in spreadsheet: Shopify revenue ÷ total ad spend)
• Any anomalies: revenue spike or drop >20%?

Weekly (30 minutes):

• MER vs. target
• NC-MER vs. target
• Cost per purchase by channel
• Top 3 creatives by spend and ROAS
• Email open rate + revenue per send
• Returns and chargeback rate

Monthly (90 minutes):

• Cohort analysis update
• LTV:CAC by acquisition cohort
• P&L vs. budget
• CAC payback period
• Channel mix shift (are you more or less reliant on any one channel?)
• New customer cost trend (is it rising, flat, or falling?)

---

#5.2 Red Flags — When to Act Immediately

Stop everything and investigate if:

• MER drops >25% week-over-week (algorithm change, creative fatigue, tracking broken, platform policy issue)
• Return rate spikes above 15% in any 7-day period (product quality or expectation issue)
• Chargeback rate above 0.8% in any month (fraud, customer service failure, or shipping problem)
• Email deliverability rate below 85% (domain reputation damage — stop emailing, fix first)
• Ad account restricted (pre-established backup account ready to activate?)

Review and optimise if:

• NC-MER declining 3 consecutive weeks (customer quality declining)
• CAC rising >20% over 30 days (competition, saturation, or creative fatigue)
• Email click rate below 1% (subject line or content problem)
• Repeat purchase rate declining on Month 2 cohort (product or experience problem)

---

Next module: IDS_07_Brand_Building.md — Positioning, story, and the Hormozi value equation for building brands that command premium.