Back to home
// Full teardown & rebuild
Illustrative · real client, name changed

The whole path, rebuilt: how we’d grow Trailhead Sports

Not one page. Not one ad. Every lever from the first impression to the closed sale. Meta and Google ads, custom funnels, the website, the offers, the copy and creative experiments, follow-up, retargeting, and attribution, all rebuilt so the full system sells together.

// The whole path, end to end

From ad impression to attributed profit.

LeakBroad Meta & Google adsOne audience, generic creative, brand + non-brand blended.
LeakGeneric homepageNo match to why they clicked.
LeakGeneric product pageSpecs first, proof buried, unclear offer.
CheckoutCartMany arrive, fewer convert.
LeakOne-size follow-upSame email & retargeting for everyone.
LeakBlended ROASBrand counted as growth; can’t tell what worked.

Money leaks at every stage, and the reporting hides which one.

RebuiltSegmented Meta & GoogleAngle-tested creative, brand vs non-brand split, intent-matched.
RebuiltBuyer-specific funnelThe page keeps the ad’s promise.
RebuiltProof-first product pageReviews & real-wear proof up top.
RebuiltTested offer + checkoutA reason to buy now that protects margin.
RebuiltMatched follow-upEmail, SMS & retargeting by segment.
RebuiltClean attributionNew vs captured demand, to gross profit.

Every lever matched to the buyer, and measured against real profit.

// The method behind every phase

How our focus groups narrow the field.

Every ad angle, page, offer, and email above starts here. Before we spend a dollar or build a thing, we pressure-test the obvious options against synthetic buyer panels, so we start from the strongest few instead of a blank page or the loudest opinion in the room.

Everything on the table

Every angle worth considering for Trailhead Sports, from the whole team, the reviews, and the market.

12-hour shiftsNurse-approvedAll-day supportMarathon recoveryInjury preventionLightweight speedGift they’ll wearEco materialsLowest priceDurabilityArch supportStyle-led
Focus-group pre-read

Synthetic panels for each buyer type (comfort, runner, gift) score every angle on believability and pull.

Comfort buyersRunnersGift buyers
“12-hour shifts shouldn’t hurt”92
Marathon recovery78
Gift they’ll actually wear71
Lowest price28
The shortlist we test live

Only the strongest few move into real ads, pages, and offers, the rest get parked or dropped.

“12-hour shifts” · comfortBuild now
Marathon recovery · runnerTest
Gift they’ll wearBuild next
Lowest priceDrop
Style-ledPark
What the pre-read actually decides
Which segment to prioritize first
Which message is most believable
Which offer is most compelling
Which objection to handle up front
Which page angle should lead
Which creative deserves budget first

Focus groups don’t pick the winner, live performance does. They just make sure we start from the strongest few, not a random guess, so budget goes to bets that already cleared a bar.

// Phase 01

Paid acquisition, Meta & Google

Traffic isn’t one lever. We rebuild how each channel finds buyers, what it says to them, and how it’s measured, so spend creates demand instead of just harvesting it.

Meta ads (Facebook & Instagram)
Prospecting that creates new demand
Before
One broad audience, one budget, boosted best-sellers.
Generic lifestyle creative shown to nurses, runners, and gift buyers alike.
No angle testing, whatever the algorithm happens to spend on wins by default.
Ad account · today
“Shop the collection” · broadOnly ad
After
Segment-specific angles built from the focus-group pre-read, each with its own hook.
Structured testing: angle × hook × format (static / video / UGC), clean naming and cost caps.
Winners scale, losers cut on a schedule, judged on gross profit, not platform ROAS.
Creative angle test matrix
“12-hour shifts shouldn’t hurt” · comfortScaling
“Recovery starts before the run” · performanceTesting
“The gift they’ll actually wear” · giftingBuild next
Generic lifestyle montageCut
Google ads (Search, Shopping & PMax)
Capturing intent without paying for demand you already had
Before
Brand and non-brand lumped into one “Search” number; brand clicks counted as growth.
Broad match burning budget on irrelevant terms; every click lands on the homepage.
Shopping feed and Performance Max left on defaults.
Search terms · today
“trailhead” (brand)Counted as growth
“shoes” (too broad)Wasting spend
After
Brand vs non-brand split into separate campaigns and reported separately.
High-intent terms routed to matching landing pages (comfort shoes → comfort page).
Negatives tighten non-brand; Shopping titles & attributes tuned; PMax fed segment assets and audience signals.
Restructured · intent-routed
Non-brand: “shoes for nurses”→ comfort LP
Non-brand: “marathon training shoes”→ runner LP
Brand: “trailhead”Tracked apart
// Phase 02

Custom funnels per buyer

Different buyers decide differently. Instead of one path for everyone, we build a dedicated funnel for each high-value segment, page, offer, and follow-up matched end to end.

Segment funnels, end to end
One complete path per buyer, not one path for all
Today All ads Homepage Collection grid Same checkout & email for everyone One path, so every buyer but the obvious one falls out.
Pick a buyer to see the funnel we’d build for them: the real ad, the matching landing page, and what we track. Same catalog, a different path for each.
the ad
Trailhead Sports
Sponsored
Nurses / on feet ad
12-hour shifts shouldn’t wreck your feet.
matching landing page
Nurses / on feet landing page
All-day comfort shoes for people on their feet.
lead + what we track
Purchase
Add to cartSize selectedRepeat buyer
Readout: Comfort buyers convert on fit and return policy, not price.

Real ad creative and landing pages. Every step, from the ad to the tracked lead, is built for how that buyer decides, and the same system runs whether you sell shoes, cleanings, or trucks.

// Phase 03

On-site conversion

Once the click lands, the page has to keep the promise the ad made, entry, landing page, proof, and offer, rebuilt in the order they matter.

The entry point
Where the click lands first
Before
Every ad points at one homepage with broad lifestyle language.
A nurse and a marathoner see the exact same thing.
Homepage · today
Shop allSale
After
The ad’s promise carries straight into a matching entry.
Comfort buyers and runners get routed to the right path from click one.
Segment-aware entry
For long shiftsFor training
The landing page
Where paid traffic actually goes
Before
Paid traffic is dumped on a generic collection grid.
Nothing speaks to the reason they clicked.
Collection page
After
A comfort-focused landing page for long-shift workers.
Headline matches the ad, one clear promise, one CTA.
“12-hour shifts shouldn’t hurt”
Shop comfort fit →
Product proof
Answering the real objection first
Before
Specs and gallery first; the best reviews are buried below the fold.
The all-day-comfort question goes unanswered.
Product page
↓ reviews way down here
After
Real-wear proof and reviews pulled above the fold.
The comfort + durability objection is handled before the price.
Proof-first
“Wore them a 13-hour shift”
The offer
A reason to buy that protects margin
Before
The offer is unclear, usually a sitewide discount that eats margin.
No hierarchy, so every buyer is trained to wait for a sale.
Offer
20% off everything
After
A clear first-purchase offer with a tested hierarchy.
Bundle and free-shipping threshold vs. a blanket discount, measured on gross profit.
Offer hierarchy · ranked
Comfort 2-pack bundleFree ship $75+Sitewide % off
// Phase 04

The experiment engine

Nothing here is decided by opinion. Copy, creative, and offers all run through the same loop, focus-group pre-read to narrow the field, live A/B to decide the winner, then a readout that feeds the next round.

Copy & message experiments
Headlines, hooks, and angles, tested, not guessed
Before
Headlines chosen by whoever’s most senior in the room.
One version ships; no read on why it did or didn’t work.
Copy · today
“Premium performance footwear”, shipped on a hunch
After
Focus-group pre-read ranks the angles before a dollar is spent.
Top angles go live head-to-head; the winner and the reason it won feed the next round.
Headline experiment · live
“12-hour shifts shouldn’t hurt”Winning
“Nurse-approved all-day comfort”Testing
“Engineered for people on their feet”Refine
“Premium performance footwear”Cut
Creative & offer experiments
A rolling calendar of bets, ranked before spend
Before
Creative made on gut feel; offers changed reactively when sales dip.
No system for what to test next, just whatever’s top of mind.
Testing plan
None, ad-hoc
After
Creative variants tied to validated angles; offer tests (bundle vs threshold vs BOGO) judged on gross profit.
A ranked Bet Map decides what to build, test, refine, or drop each week.
Bet Map · this sprint
UGC: nurse unboxing + wear testBuild now
Offer: comfort 2-pack bundleTest
Static: side-by-side sole techBuild next
Blanket 25% offDrop
// Phase 05

Follow-up & retargeting

Most buyers don’t convert on the first visit. We rebuild the after-click system (email, SMS, and paid retargeting) so the message reflects what each buyer actually came for.

Email & SMS flows
Cart recovery, nurture, post-purchase, win-back
Before
One generic cart-recovery email for everyone who left.
No browse-abandon, no post-purchase cross-sell, no win-back.
Email · to everyone
“You left something behind”
After
Segment-matched flows: comfort buyers get comfort proof, runners get the performance angle.
Cart + browse abandon, post-purchase cross-sell (insoles, socks), win-back; SMS for time-sensitive.
Flows · by intent
Cart recovery · comfort segmentLive
Post-purchase cross-sellLive
Win-back · 60-day lapsedTesting
Paid retargeting
Sequenced, segmented, and measured for incrementality
Before
One catalog ad chasing everyone who visited, including people about to buy anyway.
Overlaps with brand search and inflates reported ROAS.
Retargeting
All visitors · same catalog ad
After
Audiences by funnel stage and segment; recent buyers excluded.
Sequenced messaging (proof → offer → urgency), tested against a holdout for real lift.
Sequence · comfort segment
Viewed LP, no add · proof adStep 1
Added to cart · offer adStep 2
48h no purchase · urgencyStep 3
// Phase 06

Attribution & measurement

None of the above matters if you can’t tell what actually grew the business. We rebuild reporting around created demand and gross profit, so every decision points at the right lever.

The reporting
Created demand vs captured demand, to profit
Before
One blended platform ROAS number; brand search counted as growth.
No read on new vs returning, per-segment profit, or true incrementality.
Dashboard · today
Blended ROAS: one number
After
Brand vs non-brand separated; new vs returning customer revenue split out.
Incrementality checks (geo holdouts / spend-down tests) on the channels that matter.
Per-segment revenue, gross profit, and experiment-level lift on one dashboard.
Dashboard · rebuilt
New demandBrand / capturedNew vs returning
Gross profitIncremental liftSegment revenue
// What we’d expect to move

The whole system points at profit, not platform metrics.

New-customer acquisition
Segmented prospecting that creates demand, not just harvests brand search.
Conversion per segment
Matched funnel + proof + offer instead of a generic grid.
Average order value
Bundle-led offer hierarchy over a blanket discount.
Repeat purchase rate
Segmented follow-up that reflects what they came for.
Gross profit & contribution margin
The number every experiment is actually judged on.
Incremental vs blended ROAS
What the spend truly created, separated from demand you already had.

Illustrative example. Real outcomes depend on your traffic, margins, and market, we commit to what we’d build and test, not to numbers we haven’t earned yet.

That was one hypothetical path. Want yours?

Send your site, product, or launch brief and we’ll run the same teardown across ads, funnels, site, experiments, follow-up, and attribution, and show the first 3-5 fixes we’d make.