Why I Built Saddl: Bringing Marketing Science to the Mid-Market PPC Gap

I spent 15 years building incrementality measurement systems for enterprise advertisers.

I worked on TV attribution, consulted with ad platforms on measurement studies, and led analytics teams where Marketing Mix Modeling was table stakes — not a premium add-on.

Then I became a 7-figure Amazon seller.

And I expected the same measurement rigor I'd spent my career building.

I found a gap.

Not because anyone did anything wrong. But because the economics of tool-building and the reality of operator needs had diverged.

Here's what I discovered: The channel that represented 40% of my total revenue — my single biggest growth driver — had fundamentally different measurement infrastructure than the traditional advertising channels I'd spent my career working on.

Saddl was born from understanding why that gap exists — and who it hurts most.

---

1. The Standards I Took for Granted

In my previous world, these weren't controversial statements — they were simply how measurement worked:

• Correlation ≠ Causation If sales went up after a TV campaign, we didn't just assume the campaign caused it. We ran incrementality studies with geo-based holdout groups to measure true lift.
• Counterfactual Analysis is Standard Practice Before claiming 'our optimization worked,' you had to answer: 'What would have happened if we did nothing?' That's not philosophy — that's basic scientific method.
• Multi-Touch Attribution Requires Multi-Horizon Windows We didn't just look at last-click. We measured 14-day, 30-day, 60-day windows to understand how different touchpoints contributed over time.
• Marketing Mix Modeling Separates Signal from Noise MMM decomposed results into: baseline trends, seasonality, competitive effects, external shocks, and YOUR actual impact. You couldn't claim credit for market forces.

These were the basics. The price of admission for serious advertising measurement.

Clients spending $5M on TV got this level of rigor automatically. Clients spending $500K on display got this.

So when I started running Amazon PPC at scale — spending $250K+ annually, driving over $1M in revenue, representing 40% of my business — I naturally looked for the same standards.

What I found was something different. And once I understood why, it made perfect sense.

---

2. The Gap I Found — And Why It Existed

Here's what I found in the Amazon advertising tool landscape:

The tools weren't designed to answer causal questions. Honestly, that makes sense given the economics:

• Investments: Tool builders prioritize operational features over measurement infrastructure.
• Market Split: Enterprise ($5M+) gets custom attribution; Small sellers (<$50K) get simple automation.
• The Mid-Market Gap: Operators in the $250K - $1M range are left without professional measurement.

Here's where it gets personal:

For a tool builder, a seller spending $250K annually on PPC might not justify custom measurement infrastructure.

But for an operator spending $250K on PPC?

That typically represents:

• $1M - $1.25M in PPC-attributed revenue (based on 4-5 ROAS, common for established sellers)
• Often 40-50% of total business revenue for mid-market operators
• Their single largest controllable growth lever and expense

When I was at the agency, we'd deploy full MMM for brands spending $2M on TV — representing maybe 5-10% of their revenue.

As an Amazon operator, I was spending $250K on PPC — representing 40% of my business — with significantly less measurement rigor.

Not because of ignorance. Because of market structure.

Here's what that means in practice:

You make an optimization: restructure campaigns, adjust 200 keywords, reallocate budget.

Two months later:

• Revenue up 8%
• ACoS down 3 points
• Organic rank improved

What you actually need to know:

• How much was your optimization vs. external factors?
• Did a competitor stock out? Was there seasonal lift?
• Did Amazon's algorithm change?
• Would revenue have grown 6% anyway?

Why it matters:

Without answers, you might scale spending on false positives, abandon strategies that actually worked, or make million-dollar decisions based on correlation instead of causation.

---

3. Why This Matters Now More Than Ever

The competitive reality has fundamentally shifted.

Amazon's seller base grows daily. What was a land grab five years ago is now trench warfare. More sellers means:

• Relentless price compression — someone's always willing to go lower
• Shrinking margins — the race to the bottom is real

Plus, the operational stakes just got higher:

• CPCs rising 15-30% YoY in many categories
• Margins compressing from competition and fees
• Attribution windows lengthening (14-30 days for Amazon's learning algorithms)
• Auction complexity increasing with algorithm updates

The cost of being wrong:

For a seller spending $250K on PPC:

• A single wrong strategic decision (scaling something that didn't work) can cost $50K-$100K
• Confusing market lift with optimization impact leads to over-investment
• Missing what actually works means leaving money on the table

"According to a 2025 Teikametrics report, only 18% of sellers accurately track full-funnel attribution."

The other 82% are making high-stakes decisions based on correlation, hoping it's causation.

---

4. What Bringing Measurement Rigor to Mid-Market PPC Actually Looks Like

Saddl isn't revolutionary. It's applied marketing science — the same principles I spent 15 years building for TV, display, and brand campaigns.

The difference: We're bringing these standards to the mid-market Amazon PPC segment that's been underserved.

Here's what that means:

Counterfactual Analysis

For every optimization, we model 'What would have happened if you did nothing?' using time-series forecasting and statistical controls. This separates your impact from baseline trends.

Multi-Horizon Measurement

We track impact across 14-day, 30-day, and 60-day windows to account for Amazon's auction learning algorithms and delayed attribution.

Market Decomposition

We separate your optimization effects from: baseline trends, seasonality, competitive dynamics, external shocks, and Amazon algorithm changes.

Decision-Level Statistical Significance

We don't claim 'wins' based on random variation. Every significant decision gets significance testing. Sometimes the best insight is "not enough data yet — wait another week."

Revenue Protected

We measure impact even when revenue drops — because preventing a bigger decline is often the real win during market downturns.

This is table-stakes measurement in traditional advertising. It should be accessible to mid-market Amazon operators too.

Let me be concrete about ROI:

Scenario: Spending $20K/month on PPC

⚠️ Without Incrementality

• • 50 optimization decisions per quarter
• • Assume 70% were wins (dashboard math)
• • Reality: Only 40% drove true impact
• Yearly Cost of False Positives: $50K+

✅ With Measurement

• • Same 50 optimization decisions
• • Know exactly which 20 actually worked
• • Scale only on proven incremental lift
• ROI on Measurement: 20-40x

For an operator spending $250K annually on PPC, sophisticated measurement isn't a luxury. It's one of the highest-ROI investments you can make.

---

5. The Mission: Solving for the Underserved Middle

Saddl exists because I lived both worlds.

I spent 15 years building incrementality measurement for enterprise advertisers with massive budgets.

Then I became a mid-market Amazon operator and discovered: The segment that needs this most has access to it least.

Not because of ignorance. Because of market economics.

But market economics are changing:

The mid-market segment is now:

• Large enough to matter (tens of thousands of sellers in the $250K-$1M PPC range)
• Sophisticated enough to value it (operators asking causal questions)
• Competitive enough to need it (margins don't allow for 30% waste)

That's the gap we're closing.

We're not inventing new science. We're making existing measurement standards accessible and affordable for the operators who've been stuck in the middle.

We're bringing the rigor that $5M TV campaigns get to the $250K PPC budgets that often matter more to business outcomes.

Because if you're spending $20K/month on PPC — representing 40% of your business, defending against daily price wars — you deserve better than 'trust the dashboard and hope for the best.'

You deserve decision intelligence, not just automation.

You deserve to know which of your optimizations actually moved the needle.

You deserve to separate your impact from market noise.

You deserve the same measurement standards that enterprise brands take for granted.

That's what Saddl delivers.

---

Closing: An Invitation

If you're a mid-market operator spending $250K+ annually on PPC:

You're in the gap. Your budget is too large for basic automation to be enough, but the market hasn't prioritized building the measurement layer you need.

You're making decisions worth hundreds of thousands of dollars based on correlation, not causation.

You're probably wondering: "Did my optimization actually work, or did something else change?"

You're asking the right question. Now there's finally a tool built to answer it.