Consulting · Teardown
Google Ads consultant teardown: 3 anonymised accounts, before and after
Three Google Ads accounts I took over in 2026, anonymised at client request. Reported CPA and ROAS, actual cost per SQL or held call, root cause, exact fixes, and the 60 to 90 day movement. All numbers pulled from Google Ads, GA4 and CRM exports, not cropped screenshots.
Updated 2026-12-18. Accounts anonymised by industry and spend band; permission granted to publish aggregate metrics only.
TL;DR
- Account A (B2B SaaS Search, USD 28k spend): reported CPA rose from USD 82 to USD 165 on non-brand, cost per SQL fell 55 percent to USD 290 in 90 days.
- Account B (DTC PMax, USD 46k spend): reported ROAS fell from 4.8x to 4.1x, blended MER rose from 1.9x to 2.7x and became contribution-positive in 75 days.
- Account C (local services BD, BDT 320k spend): cost per qualified call held fell from BDT 3,900 to BDT 1,450 in 60 days.
- Common thread: brand and non-brand mixed, Enhanced Conversions off, PMax without margin-band asset groups, bid strategy stuck on Maximize Conversions.
- Fee band for a Google Ads consultant retainer covering all four: USD 1,800 to 4,000 monthly.
Why teardowns beat case studies
Case studies are marketing collateral: cropped screenshots, headline metrics, no root cause. Teardowns are working documents: before numbers, diagnosis, exact interventions, after numbers, and the metrics that got worse alongside the ones that got better. If a case study only shows the metrics that improved, treat it as a testimonial, not evidence.
The three accounts below are anonymised by industry and spend band, with client permission to publish aggregate before-and-after numbers. Brand names, creatives, and product SKUs are withheld.
Account A: B2B SaaS, USD 28k monthly spend
B2B SaaS, USD 28k monthly spend, US Search-only, demo-request funnel, ACV USD 12k.
- Reported CPA: USD 82
- Actual cost per SQL: USD 640 (13 percent of conversions qualified)
- Brand vs non-brand split: 72 percent of conversions from brand queries
- Bid strategy: Maximize Conversions, no CPA ceiling
- Enhanced Conversions: Off; conversion counted on thank-you page view
- Reported CPA (non-brand): USD 165 (higher, on purpose)
- Actual cost per SQL: USD 290 (55 percent lower)
- SQL rate on conversions: 37 percent (up from 13 percent)
- Window: Day 0 to 90
Brand queries were being counted as paid conversions and the algorithm was optimising to the cheapest converter, brand search, which would have closed organically. Non-brand Search was starved and Enhanced Conversions was never enabled, so signal quality was poor everywhere.
- Split brand and non-brand into separate campaigns; capped brand budget at USD 1,500 monthly.
- Enabled Enhanced Conversions and wired offline conversion upload from HubSpot for SQL and Closed-Won stages.
- Switched non-brand campaigns from Maximize Conversions to Target CPA with a documented ceiling of USD 180.
- Rebuilt ad copy: swapped generic 'try free demo' variants for problem-first hooks tied to the top 3 job titles in the ICP.
Account B: DTC apparel, USD 46k monthly spend
DTC apparel, USD 46k monthly spend, UK + AU, Search plus Performance Max, gross margin 52 percent.
- Reported ROAS: 4.8x (blended, brand included)
- Actual MER: 1.9x (below breakeven of 1.92x at 52 percent margin)
- PMax asset groups: 1 group covering all SKUs, brand not excluded
- Bid strategy: Max Conversion Value, no Target ROAS floor
- Feed hygiene: 18 percent of SKUs disapproved or with GTIN errors
- Reported ROAS: 4.1x (lower, on purpose)
- Blended MER: 2.7x (contribution positive)
- PMax incremental ROAS (holdout): 2.3x on non-brand basket
- Feed disapproval rate: 2 percent
- Window: Day 0 to 75
PMax was cannibalising brand and organic. Blended MER told the true story: the account was losing money on incremental spend. Asset group structure gave the algorithm no way to protect margin.
- Excluded brand terms account-wide from PMax via account-level negatives (support ticket route).
- Rebuilt PMax into 3 asset groups split by product margin band (30-45 percent, 45-60 percent, 60 percent plus).
- Set Target ROAS floor at 3.2x on the low-margin group, 4.0x on mid, 5.5x on high.
- Fixed feed disapprovals; added custom labels for margin band and season for PMax targeting hints.
- Ran a 21-day geo holdout to validate incremental ROAS on PMax alone.
Account C: Local services, BDT 320k monthly spend (~USD 2
Local services, BDT 320k monthly spend (~USD 2,700), Dhaka + Chattogram, call-first funnel, ticket BDT 8k to 45k.
- Reported cost per conversion: BDT 480 (form-fills + phone impressions)
- Cost per qualified call held: BDT 3,900
- Geo strategy: Country-wide, no bid modifiers
- Ad schedule: 24/7 default; office hours were Sun to Thu 10am to 7pm
- Call tracking: Google forwarding numbers unused; no CallRail or manual tag
- Cost per qualified call held: BDT 1,450 (63 percent lower)
- Qualified calls per month: 112 (vs 42 in the prior 30 days)
- Spend efficiency: 12 percent of budget freed from off-hours + out-of-geo waste
- Window: Day 0 to 60
Conversions were counting form impressions and phone-icon clicks, not held calls. Ads ran through weekends and nights with no one to answer. Bids were flat across a country where 78 percent of qualified leads came from two districts.
- Enabled Google forwarding numbers plus a manual tag on the website form for verified held calls only.
- Set ad schedule to Sun to Thu 09:30 to 20:00 with bid modifiers plus 20 percent during peak call hours.
- Applied +30 percent geo modifier for Dhaka and Chattogram, -50 percent outside.
- Switched from Maximize Clicks to Maximize Conversions with a hard daily budget cap.
- Rewrote 4 ads with the actual price band (BDT 8k starting) to filter out unqualified enquiries.
The pattern across all three
- Reported CPA or ROAS almost always moved in the 'worse' direction when the account was fixed. Cost per SQL, blended MER, or cost per held call moved the other way. Optimising to the platform-reported number is the single biggest failure mode.
- Brand and non-brand were mixed in two of three accounts. Splitting them is a 60-minute change that reveals whether paid Search is actually incremental or just intercepting organic demand.
- Enhanced Conversions and offline conversion upload were missing or partial in all three. This is a week-one setup that changes what Smart Bidding learns from.
- PMax without asset-group discipline or brand exclusion loses money at any gross margin below 55 percent. It works, but only after signal density and margin-band structure are in place.
- Bid strategy defaults (Maximize Conversions, Max Conversion Value) are wrong for accounts with a documented CPA ceiling or breakeven ROAS. Target CPA or Target ROAS with a written floor forces the discipline.
- Local Bangladesh accounts leak 15 to 30 percent of spend to off-hours delivery and out-of-geo bidding by default. Ad-schedule and geo modifiers pay for themselves inside two weeks.
None of these fixes are proprietary. What is scarce is the discipline to run them in the right order, tie reporting to the CRM or the storefront, and defend a higher reported CPA to a founder who has been trained to celebrate the vanity number.
How to vet a Google Ads consultant
Seven questions, tuned for Google Ads specifically. If any answer is vague or defensive, walk.
For the full non-channel-specific version, see the performance marketing consultant buyer's guide.
Google Ads consultant FAQ
What does a Google Ads consultant actually change in the first 30 days?
Five things, in order. Conversion tracking rebuilt (GA4 + Google Ads Enhanced Conversions + offline conversion upload tied to CRM stages, not thank-you page views). Search Terms audit and negatives (kill the 15 to 40 percent of spend leaking to junk queries). Campaign structure consolidation (fewer, higher-signal campaigns; retire single-keyword ad groups older than 2022). Bid strategy shift, usually from Maximize Conversions to Target CPA or Target ROAS with a documented ceiling. PMax asset-group split by product margin band, not by category.
What is a realistic Google Ads CPA in 2026?
For B2B SaaS demo requests: USD 90 to 240 CPA on Search, USD 60 to 150 on PMax when CRM feedback is wired. For DTC ecommerce: 1.8x to 3.5x breakeven ROAS on non-brand Search, 3.0x to 5.0x on PMax. For local services in Bangladesh: BDT 400 to 1,500 per qualified call on Search. Anything reported below those ranges is usually counting brand-search conversions or thank-you page views.
PMax or Search: which should a consultant push first?
Search first for any account under USD 15k monthly spend or with fewer than 30 conversions per month. PMax needs signal density to work: below 50 conversions per month it optimises to whatever converts cheapest, which is usually brand + shopping cannibalisation. Above 50 monthly conversions with CRM-tied values fed back, PMax often out-earns Search on incremental revenue, if asset groups are split by margin band and brand terms are excluded via account-level negatives.
How much does a Google Ads consultant cost in 2026?
USD 1,800 to 4,000 per month for a solo consultant retainer covering strategy, conversion setup, weekly optimisation, and CRM-tied reporting. One-off audit and rebuild: USD 2,000 to 5,500. Agencies quote USD 3,500 to 10,000 monthly for the same scope. For accounts below USD 5k monthly spend, a consultant retainer is usually not economical: an audit plus DIY handoff makes more sense.
What is broken in most Google Ads accounts?
Seven patterns recur. Conversions counted per session not per transaction. Enhanced Conversions off. Brand and non-brand mixed in the same PMax asset group. Search Terms report untouched for 60+ days. Bid strategy stuck on Maximize Conversions with no CPA ceiling. Ad-schedule and geo-bid modifiers untouched since account launch. Reporting anchored to Google Ads CPA rather than blended CAC or cost per SQL.
How is a Google Ads consultant different from an agency account manager?
A consultant is accountable to pipeline or contribution margin, not to Google Ads CPA. Deliverables include CRM-tied reporting (Google Ads to HubSpot, Salesforce, or Shopify), a documented target ROAS or target CPA ceiling, and monthly incrementality or holdout review on brand and Performance Max. An agency account manager usually reports platform metrics inside a dashboard and rotates every 8 to 14 months. Both can run ads; only one is accountable to profit.
Do you work with Bangladesh clients on Google Ads, or only international accounts?
Both. BD clients typically run Search plus PMax for local lead-gen and ecommerce with BDT budgets USD 800 to 6,000 monthly. International retainers cover US, UK, EU, GCC, and AU accounts with English creative, CRM-tied offline conversion upload, and monthly incrementality tests on brand and Performance Max.
When should I switch from Maximize Conversions to Target CPA or Target ROAS?
Once the campaign has 30+ conversions in the last 30 days with stable conversion action definitions. Set Target CPA at your current 30-day average, then tighten by 10 to 15 percent every two weeks while watching volume. If volume falls more than 25 percent, hold the target and rebuild ad copy or landing pages before tightening further.
Frequently asked questions
What does a Google Ads consultant actually change in the first 30 days?
Five things, in order. Conversion tracking rebuilt (GA4 + Google Ads Enhanced Conversions + offline conversion upload tied to CRM stages, not thank-you page views). Search Terms audit and negatives (kill the 15 to 40 percent of spend leaking to junk queries). Campaign structure consolidation (fewer, higher-signal campaigns; retire single-keyword ad groups older than 2022). Bid strategy shift, usually from Maximize Conversions to Target CPA or Target ROAS with a documented ceiling. PMax asset-group split by product margin band, not by category.
What is a realistic Google Ads CPA in 2026?
For B2B SaaS demo requests: USD 90 to 240 CPA on Search, USD 60 to 150 on PMax when CRM feedback is wired. For DTC ecommerce: 1.8x to 3.5x breakeven ROAS on non-brand Search, 3.0x to 5.0x on PMax. For local services in Bangladesh: BDT 400 to 1,500 per qualified call on Search. Anything reported below those ranges is usually counting brand-search conversions or thank-you page views.
PMax or Search: which should a consultant push first?
Search first for any account under USD 15k monthly spend or with fewer than 30 conversions per month. PMax needs signal density to work: below 50 conversions per month it optimises to whatever converts cheapest, which is usually brand + shopping cannibalisation. Above 50 monthly conversions with CRM-tied values fed back, PMax often out-earns Search on incremental revenue, if asset groups are split by margin band and brand terms are excluded via account-level negatives.
How much does a Google Ads consultant cost in 2026?
USD 1,800 to 4,000 per month for a solo consultant retainer covering strategy, conversion setup, weekly optimisation, and CRM-tied reporting. One-off audit and rebuild: USD 2,000 to 5,500. Agencies quote USD 3,500 to 10,000 monthly for the same scope. For accounts below USD 5k monthly spend, a consultant retainer is usually not economical: an audit plus DIY handoff makes more sense.
What is broken in most Google Ads accounts?
Seven patterns recur. Conversions counted per session not per transaction. Enhanced Conversions off. Brand and non-brand mixed in the same PMax asset group. Search Terms report untouched for 60+ days. Bid strategy stuck on Maximize Conversions with no CPA ceiling. Ad-schedule and geo-bid modifiers untouched since account launch. Reporting anchored to Google Ads CPA rather than blended CAC or cost per SQL.
How is a Google Ads consultant different from an agency account manager?
A consultant is accountable to pipeline or contribution margin, not to Google Ads CPA. Deliverables include CRM-tied reporting (Google Ads to HubSpot, Salesforce, or Shopify), a documented target ROAS or target CPA ceiling, and monthly incrementality or holdout review on brand and Performance Max. An agency account manager usually reports platform metrics inside a dashboard and rotates every 8 to 14 months. Both can run ads; only one is accountable to profit.
Do you work with Bangladesh clients on Google Ads, or only international accounts?
Both. BD clients typically run Search plus PMax for local lead-gen and ecommerce with BDT budgets USD 800 to 6,000 monthly. International retainers cover US, UK, EU, GCC, and AU accounts with English creative, CRM-tied offline conversion upload, and monthly incrementality tests on brand and Performance Max.
When should I switch from Maximize Conversions to Target CPA or Target ROAS?
Once the campaign has 30+ conversions in the last 30 days with stable conversion action definitions. Set Target CPA at your current 30-day average, then tighten by 10 to 15 percent every two weeks while watching volume. If volume falls more than 25 percent, hold the target and rebuild ad copy or landing pages before tightening further.
Want a teardown of your Google Ads account?
Send me read-only access to Google Ads plus one export from your CRM or Shopify. I reply with a written root-cause diagnosis inside four working hours, no sales pitch attached.