Prepared by Slice of the Pie Marketing | For: Owen Finnerty, The Golf Sock | Source: Google Ads UI, customer 133-906-8799, live session Sep 21, 2026, plus Semrush (US) and Merchant Center visibility check. Framework: SOTP Campaign Architecture Playbook, Sep 2026 edition, Chapter 3.
What the account told us
The Google account was built by a previous buyer's offshore hire and nobody has been in it since. It is small, it is profitable, and it is capped. Those three facts pick the structures.
~$206a day, all in. Two PMax at $180, one Search at $21, Brand at $5 and paused. About $6K a month on the channel with the best returns in the account.
6xon Search, 3x on PMax at $16.66 per conversion. Brand clicks convert at $9.23. Every live campaign is "limited by budget."
20%impression share overall, 24% on Shopping, under 10% on your own brand name. Four out of five searches you could win, you are not in.
0Shopping listings captured by Semrush despite $180/day of PMax. The feed is either not serving on the shelf or serving too thin to register.
| Campaign (as found) | Type | Budget/day | Status | Read |
|---|
| MD | Pmax | GS | Performance Max | $145 | Live, limited by budget | The workhorse. ~3x, $16.66 CPA. No brand exclusion, so some of that 3x is your own name |
| MD | Pmax | FDO | Performance Max (feed only) | $35 | Live, limited by budget | A second PMax competing for the same products with no creative. Overlap with GS |
| MD | Search | | Search | $21 | Live, limited by search volume | ~6x. One campaign, brand and nonbrand mixed, 13% impression share. Starved |
| Brand | Search | Search | $5 | Paused | Your own name, undefended. A competitor can buy "the golf sock" for about a dollar |
| MD | DG | RT | Display retargeting | $8 | Paused | Cart recovery exists as a shell. Display is migrating to Demand Gen, so this gets rebuilt, not resumed |
| Standard Shopping / Display / PMax | Mixed | $35 | Paused | Dormant shells from earlier builds. Archived on day one |
The diagnosis in one line: there is no structure to scale. One PMax does everything with no brand exclusion, a second PMax overlaps it, Search mixes brand with nonbrand in one budget, and the brand campaign is off. The playbook's control for a one-product catalog with a working feed is 3.01 Consolidated retail PMax, and its capture layer is 3.02 Brand and category Search. Both are half-built in the account already. We finish them, separate the jobs, and then the budget has somewhere to go.
Why "way more" is the right instinct
Impression share is the tell. At 20% overall and 24% on Shopping, the account is leaving roughly four times its current volume on the table before CPCs even move. Search is the highest-intent traffic there is: someone typing "golf socks" is already shopping. Semrush puts the core category at 4,400 searches a month for "golf socks," 1,600 for "golf socks for men," 1,000 for "mens golf socks," and the gifting lane at 12,100 for "golf gifts" alone, with CPCs between $0.47 and $1.13. Meta finds demand. Google catches it. Right now you are catching a fifth of it.
The two structures
A · Consolidated Retail PMax
Playbook 3.01 · sotp_google_ecom_ecore_*
One PMax campaign for the whole catalog, split into asset groups by pack (six-pack hero, singles and three-pack, twelve-pack, gifting) with mutually exclusive listing filters, brand excluded at the campaign level, tROAS with a return target grounded in the account's own history. This is your existing GS PMax, finished.
Best fit per playbook: one product family, a healthy Merchant Center feed, validated purchase values, short purchase cycle. You: one SKU in four packs, feed live but unverified, Shopify pixel plus Triple Whale.
B · Brand and Category Search
Playbook 3.02 · sotp_google_ecom_esearch_*
Two Search campaigns with separate budgets and separate return targets: Brand (exact, near-100% impression share, cheap insurance) and Nonbrand Category (the nine themed ad groups already on the Ads Plan page, phrase and exact, one RSA per group). Navigational sales can no longer hide prospecting cost.
Best fit per playbook: shoppers compare categories before choosing a seller, purchase values validated against store orders. You: 6x on the existing Search campaign says the intent is there.
A is the control and gets 60% of the budget. B is the capture layer and gets 40%, released once the query reports show brand and category are cleanly separated. On the bench: 3.08 Feed-based cart recovery (rebuilt on Demand Gen once Google's Display migration settles), 3.04 Margin-separated PMax (only if the six-pack and singles need different return targets after 30 settled orders), and YouTube competitor-intent placements as a December test. The playbook keeps Demand Gen off the launch stack for ecommerce and so do we.
Structure A, in full
Naming: sotp_[platform]_[vertical]_[structure]_[objective]_[segment] for campaigns, asg_[theme]_[signal] for asset groups, lg_[attribute]_[value] for listing groups, ag_[theme]_[match] for ad groups.
sotp_google_ecom_ecore_purchase_catalog MD | Pmax | GS, renamed and finishedPerformance Max · Maximize conversion value · tROAS 300% at launch · $180/day at launch (GS + FDO folded in) · brand exclusion list applied · final URL expansion off · US
asg_sixpack_heroThe pack driving AOV. Listing group: six-pack only. Audience signal: purchasers 180d + Meta website visitors 30d. Gets the best creative and the most spend by design.
- lg_pack_six · six-pack variants only, all colors
- ad_img_sixpack_v01-v06 · 1.91:1, 1:1, 4:5 from the Meta static winners (F3, 124C, 128C families)
- ad_vid_sixpack_v01-v02 · Navy Launch video + 112B cut, horizontal and vertical
asg_core_buyersSingles and three-pack. Listing group: single pair and three-pack. Audience signal: Meta cart/IC 14d + in-market golf apparel.
- lg_pack_single · lg_pack_three
- ad_img_core_v01-v06 · product stills, grip macro, pro proof
- ad_vid_core_v01 · grip tech explainer
asg_twelve_valueTwelve-pack. Listing group: twelve-pack. Smaller group, exists so the highest-ticket SKU is never starved by the six-pack.
- lg_pack_twelve
- ad_img_twelve_v01-v04 · "stock the drawer" flat lay, per-pair math
asg_gift_intent Nov 1 to Dec 18Same catalog, gifting creative and gifting audience signal (Google's gift-buyer in-market segments + Meta gifting concept engagers). Paused outside the window.
- lg_pack_three · lg_pack_six (gift wrap eligible)
- ad_img_gift_v01-v06 · "for the golfer who has everything," gift note callout, shipping cutoff versions
- ad_vid_gift_v01 · unboxing / reaction cut
Campaign-level controlsBrand exclusion list ("the golf sock," "golf sock," "thegolfsock," misspellings) so this campaign only buys nonbrand. Final URL expansion off, PDP and collection URLs only. Merchant Center promotions feed for the BFCM window. Conversion action: Shopify Purchase, deduplicated against GA4, with transaction IDs.
Structure A rules, agreed before launch
Feed first. Nothing scales on a feed that isn't serving. Day 1 to 3: Merchant Center diagnostics, product titles rewritten to lead with "Golf Socks" plus pack and color, GTINs verified, all four packs approved, price and availability matching the site. Semrush says 0 captured listings; we find out why before we add a dollar.
Return target: tROAS 300% at launch, which is where the campaign already sits. Lowered to 250% only inside the Dec 2 to 18 gifting window and raised back on Dec 19. Nonbrand new-customer CPA ceiling is $40, hard stop at $59.
Kill rule (playbook 3.01): after 28 days and 30 settled purchases, restructure if contribution return stays below the floor for two weekly checks. Pause a listing group after three target CPAs ($120) with no sale once its conversion lag has passed.
Scale rule: when the campaign is flagged "limited by budget" and trailing 7-day ROAS is at or above target, raise the daily budget 20% every 72 hours. Never more than 20% at a time on PMax; it resets learning.
Pivot rule: after 30 settled orders, if six-pack and singles contribution margins differ by 20 points and pooled bidding keeps favoring the weaker one, split into margin-separated PMax (3.04). We expect this not to trigger; one SKU, one margin.
Structure B, in full
sotp_google_ecom_esearch_purchase_brand Brand | Search, un-paused and rebuiltSearch · Maximize conversion value · tROAS 800% · $30/day · exact match only · impression share target 95%+
ag_brand_exactYour name, your name misspelled, "golf sock reviews," "golf sock discount code." Landing page: homepage. Cheapest conversions in the account at $9.23. This campaign is insurance, not growth.
- kw_[the golf sock] · kw_[golf sock] · kw_[thegolfsock] · kw_[the golf sock reviews] · kw_[golf sock promo code]
- ad_rsa_brand_v01 · 15 headlines, 4 descriptions, all sitelinks (packs, reviews, guarantee, gifting)
sotp_google_ecom_esearch_purchase_nonbrand MD | Search |, renamed, brand stripped outSearch · Maximize conversion value · tROAS 350% · $120/day at launch · phrase + exact · brand terms as campaign negatives · the nine ad groups from the Ads Plan
ag_core_phraseAd Group 1. "golf socks" (4,400), "golf socks for men" (1,600), "mens golf socks" (1,000), "best golf socks." Landing page: collection. Highest priority, highest budget share.
- ad_rsa_core_v01 · pinned H1 "Golf Socks With Grip Technology," pro proof, no-blister guarantee, free shipping over $70
ag_noshow_phraseAd Group 2. No-show, ankle, low-cut golf socks. Landing page: collection filtered.
ag_performance_phraseAd Group 3. Compression, performance, moisture-wicking golf socks. Landing page: PDP with the tech section anchored.
ag_gifting_phrase Nov 1 to Dec 20Ad Group 5. "golf gifts" (12,100), "golf gifts for men" (8,100), "golf gifts for dad" (2,400), "golf stocking stuffers." Own budget share, own gift landing page. This ad group alone is bigger than the whole core category.
- ad_rsa_gift_v01 · gift-note, ships by Dec 18, "the one golf gift he'll actually use"
ag_womens_phraseAd Group 7. Women's golf socks. Small, always on.
ag_conquest_exactAd Group 8. Competitor names (Del Campo, Swiftwick, FootJoy socks, Bombas golf). Exact only, own daily cap of $20, differentiation copy, never the brand campaign's landing page.
- ad_rsa_conquest_v01 · "Made for golf, not borrowed from running"
ag_custom_phraseAd Group 9. Custom golf socks, team golf socks, logo golf socks, bulk. High AOV, feeds the wholesale lane. Landing page: custom inquiry.
- ad_rsa_custom_v01 · "200 pairs, your logo, drop-shipped"
Campaign-level controlsNegatives: brand list, "free," "pattern," "knitting," "how to," job and wholesale-supplier terms, the launch negative list from the Ads Plan. Ad Groups 4 (novelty) and 6 (adjacent apparel) stay on the bench until core CPA is proven under $40 for 28 days.
Structure B rules
Separate return targets. Brand at 800% because it converts at $9. Nonbrand at 350% because it has to pay for discovery. Judging both from one blended account number is the failure mode the playbook names first.
Kill rule (playbook 3.02): pause a nonbrand theme after three target CPAs ($120) with zero settled purchases. After 28 days and 30 orders, cut a theme's budget if new-customer cost exceeds $40 by 25%, even when brand returns keep the account looking healthy.
Query hygiene, weekly. Search terms report every Monday. Anything brand that leaks into nonbrand becomes a negative. Anything nonbrand converting inside PMax that Search isn't bidding on becomes a keyword.
Pivot rule: if after 28 days profitable product pages have no keyword coverage while category Search is inside its ceiling, add a page-feed DSA companion (3.07). With four products this probably never fires; it is here so the decision is written down.
Launch sequence
Days 1 to 7 · Feed, tracking, rename (Oct 1 to 7)
- Merchant Center: diagnostics, title rewrites, GTIN check, all four packs approved. Confirm the Shopify sales channel is the feed source, not a stale third-party feed.
- Conversion tracking: one Purchase action, Shopify plus GA4 deduplicated, transaction IDs on. Reconcile the last 30 days of Google-reported purchases against Shopify orders.
- Rename GS PMax and the Search campaign to the SOTP standard. Archive the dormant shells.
- Build the brand exclusion list and the nonbrand negative list. Apply to PMax and nonbrand Search.
- Day 5: un-pause Brand Search as sotp_google_ecom_esearch_purchase_brand at $30/day. This is the single fastest win in the account.
- Day 7: fold FDO into the core PMax. Four asset groups live with the Meta static winners resized.
Days 8 to 30 · Read the first cohort (Oct 8 to 31)
- Nonbrand Search restructured into the seven live ad groups by Oct 10. Brand terms stripped, negatives applied, one RSA per group.
- Weekly: search terms, impression share by campaign, budget-limited flags. First 20% budget lifts expected Oct 12 to 15 on whichever campaign clears target first.
- Oct 24 to 31: rivalry drop gets a Merchant Center promotion and a temporary listing group in the six-pack asset group. No new campaign for it on Google.
- Day 30: reconcile settled purchases. Release the Search allocation to 40% of the Google budget if brand and nonbrand query reports are clean.
Days 31 to 60 · Gifting on, budgets up (Nov)
- Nov 1: gifting asset group and gifting ad group go live. Gift landing page live on the site.
- Nov 1 to 20: cheapest CPCs of the quarter. Push budgets hardest here, 20% every 72 hours while limited-by-budget and on target.
- Nov 14: BFCM promotion loaded in Merchant Center, sale sitelinks and callouts drafted and approved. Same lock date as Meta.
- Nov 21 to Dec 1: promotion live in Shopping, sale RSAs swapped in on brand and core. tROAS holds; Google does not get a lower target for a sale.
Days 61 to 90 · Peak, cutoff, defend (Dec)
- Dec 2: PMax tROAS to 250% for the gifting window only. Gifting ad group budget share to 40% of nonbrand.
- Dec 12: shipping cutoff copy swapped in across RSAs and gifting images. Dec 18: gifting asset group paused, tROAS back to 300%.
- Dec 19 to 24: e-gift card sitelink on brand and core. Dec 25 to 31: budgets step down to the November 1 level.
- Jan 2: 90-day review. Settled cohorts against the Oct 1 baseline, Google claims reconciled to Shopify and Triple Whale, then the next test: YouTube competitor intent, cart recovery on Demand Gen, or a margin split if the data says so.
The 90-day scaling plan
Daily Google budget by period, if the guardrails hold. Today: about $206/day. Impression share says the ceiling is four to five times that before CPCs move. Every step up is earned by the 20% rule on a campaign that is both budget-limited and on target.
| Period | PMax (A) | Brand Search (B) | Nonbrand Search (B) | Daily total | Period total |
|---|
| Oct 1-7 | $180 | $30 | $60 | ~$270 | $2K |
| Oct 8-31 | $180 → $260 | $30 | $120 → $180 | ~$400 | $10K |
| Nov 1-20 | $260 → $400 | $40 | $180 → $280 | ~$620 | $12K |
| Nov 21-Dec 1 (BFCM) | $450 | $60 | $300 | ~$810 | $9K |
| Dec 2-18 (gifting) | $500 → $600 | $60 | $350 → $400 | ~$1,000 | $17K |
| Dec 19-24 | $400 | $60 | $200 | ~$660 | $4K |
| Dec 25-31 | $260 | $40 | $150 | ~$450 | $3K |
| Q4 Google total: ~$57K | Oct ~$12K · Nov ~$19K · Dec ~$26K | | ~$57K |
That is roughly three times what the channel would spend if nobody touched it (~$18K for the quarter at $206/day). At the account's current blend of 3x PMax and 6x Search, and allowing for CPC inflation in December, that is $170K to $200K in Google-attributed revenue for Q4 against $57K in spend. Combined with the Meta build, total paid media for Q4 lands around $230K, which is the "close to $100K a month in Q4" number from our call, reached in December.
What can stop the ramp: a feed that still doesn't serve after the day-3 fix (then PMax holds at $180 and Search takes the growth), or nonbrand CPA over $50 for three days (then nonbrand cuts 25% and PMax absorbs). Brand Search never gets cut; at $9 a conversion it is the last thing to turn off.
What Google needs from you (assets)
Google eats less creative than Meta, but it eats it in specific shapes. Almost all of it is resized from the Meta winners, so the marginal production is small. What it needs more than creative is a clean feed and landing pages that match the query.
~40images across four asset groups: 1.91:1, 1:1, 4:5, plus 5 logo variants. Resized from the Meta statics that already win.
6videos, 10 to 30 seconds, horizontal and vertical. Navy Launch, 112B, grip explainer, gifting unboxing. Without these PMax makes its own, badly.
8responsive search ads, 15 headlines and 4 descriptions each, plus sitelinks, callouts, structured snippets, price and promotion extensions.
3landing pages that don't exist yet: gift landing page, custom/team inquiry page, no-show or performance filtered collection. The Ads Plan already scoped them.
| Asset | Count | Source | Due |
|---|
| Merchant Center feed fixed and approved | 4 products, 16 variants | Slice of the Pie, in the retainer | Oct 3 |
| PMax asset groups (images, logos, text) | 4 groups, ~40 images, 15 headlines, 5 long headlines, 5 descriptions each | Resized from Meta winners, in-house | Oct 7 |
| PMax videos | 6 | Meta video winners re-cut, videographer for the grip explainer | Oct 7 (4), Oct 31 (gifting 2) |
| Responsive search ads | 8 (brand, 7 nonbrand groups) | In-house, from the Ads Plan concepts | Oct 10 |
| Extensions | 8 sitelinks, 6 callouts, 2 snippets, price ext, promo ext | In-house | Oct 10 |
| Gift landing page | 1 | Josh builds on the new PDP framework, we supply copy and sections | Oct 28 |
| Custom / team inquiry page | 1 | Same | Oct 28 |
| BFCM and cutoff RSA swaps, gifting images | 4 RSAs, 8 images | In-house | Nov 14 |
| Merchant Center promotions (rivalry, BFCM) | 2 | In-house | Oct 20, Nov 14 |
What we need from you to start Oct 1
- Admin on the Google Ads account (133-906-8799) and Merchant Center
- GA4 admin, so the conversion action gets deduplicated properly
- Confirm which feed Merchant Center is reading (Shopify channel vs. a third-party app)
- Josh's timeline on the two landing pages, or the go-ahead for us to build them
- BFCM offer by Nov 7, same as Meta
- Shipping cutoff date from the 3PL
- A yes by Friday, Sep 25
Pricing for this build is on the pricing page: the ad-buying tier covers both this and the Meta build; the Fractional VP tier adds the landing pages, SEO, and everything else. The paid-media audit this is built on is the STAOS Ads Audit. The keyword research behind the ad groups is on the Ads Plan.
Data Sources
Google Ads UI, customer 133-906-8799, live session Sep 21, 2026: campaign list, budgets, statuses, impression share, cost per conversion. Semrush (US) keyword volumes and PLA capture, July and September 2026. SOTP Campaign Architecture Playbook, September 2026 edition, structures 3.01 to 3.08, pivot map, and first-90-days plan. Business numbers from the Sep 21 call with Owen. No campaign, budget, or setting was changed to produce this page.
Slice of the Pie × The Golf Sock