Acquisition.md
← Playbooks
Advertising · 18 min

Google Ads for B2B

Intent-first search strategy for high-consideration products.

In B2B, the click you want costs $80 and the click you don't want costs $12. Guess which one gets the traffic by default.

tl;dr
  • Use Google Ads only for intent capture; Meta Advantage+ handles demand creation in parallel.
  • Build 3–5 intent clusters in Clay, then score accounts before spending 1 pound.
  • Send paid clicks to Framer + Relume pages, not a generic homepage or 12-step maze.
  • Connect PostHog, Northbeam, and Attio before launch, or every CAC number is fiction.
  • Use 11x + Clay to route high-intent leads in 5 minutes; cold SDRs are dead in 2026.
Action plan
Action plan0/3 · 0%
Progress saved locally. Sign in to sync
Week one
  1. 01Day 1: Export 500 target accounts, tag 3 intent clusters, and drop the rest.
  2. 02Day 2: Draft 3 keyword sets in Google Ads with exact and phrase match only, then add 25 negatives.
  3. 03Day 3: Build 1 landing page per cluster and instrument CTA clicks and form submits.
  4. 04Day 4: Route every high-intent lead to a rep in under 5 minutes via workflow automation.
  5. 05Day 5: Launch 3 campaigns, 1 per buying motion, with separate budgets and tCPA targets.
  6. 06Day 6: Pull the first cohort, review search terms, and negate anything that produced 0 meetings.
  7. 07Day 7: Write 1 SQL audit and decide which cluster gets 20% more budget.

Design the account around intent, not product categories

Google Ads for B2B in 2026 starts with 3 intent layers: problem-aware, solution-aware, and vendor-aware. Use Clay with Apollo AI and Common Room to build 1 list for each layer, then use an LLM enrichment pass to tag buying stage, company size, and trigger events. Keep the account narrow: 10–20 exact and phrase keywords per cluster, 1 RSA per cluster, and 1 Framer page per intent level. PostHog should tell you which queries convert, while Attio should hold the account-level source of truth. If you are still bidding on broad nouns like “software” or “platform,” you are donating to Google’s margin.

  1. 01

    Map 3 intent clusters

    Use Clay to pull 500 accounts from Ocean.io and Apollo AI, then ask Claude 4.5 Sonnet to label them by pain, urgency, and buyer role.

    Prompt: `Tag each account as problem-aware, solution-aware, or vendor-aware; return 1 line per account in CSV.`
  2. 02

    Build 1 page per cluster

    Create Framer + Relume pages with 1 primary CTA, 1 proof section, and 1 objection-killer above the fold.

    Example: `framer.com/ppc/b2b-data-revops` with a 14-word headline and 1 Calendly-style booking CTA.
  3. 03

    Score before launch

    Use Clay to assign a 0–100 account score, then only send budget to clusters above 70 and exclude the rest.

    Rule: if a cluster has <15 high-fit accounts in Clay, do not buy the keyword.
IfThen
a keyword has high volume but low purchase intentpause it and move budget to exact-match problem and vendor queries.
a cluster cannot map to a Framer page and 1 CTAdo not launch it; build the page first or kill the cluster.

Run 1 tight structure and let Smart Bidding do the grinding

In 2026, the winner is a clean Google Ads structure with 3–7 campaigns, 1 theme per ad group, and aggressive negative keyword hygiene. Pair Google Performance Max for capture with Search for precision, then use Demand Gen only for retargeting once PostHog shows engaged visitors. Clay should feed company-size and ICP filters into audience segments, while n8n pushes qualified leads into Attio and notifies 11x for immediate follow-up. Use AdCreative.ai or Pencil for 3–5 test variants, but stop pretending creative volume fixes weak intent. If your CTR is high and qualification is low, the query set is wrong, not the landing page.

  1. 01

    Split campaigns by buying motion

    Create 1 campaign for problem keywords, 1 for category terms, and 1 for competitor terms; keep budgets separate so signals do not blur.

    Example: `Search | Problem`, `Search | Category`, `Search | Competitor`, each with its own tCPA target in Google Ads.
  2. 02

    Automate lead routing

    Use n8n to send form fills from Framer to Attio, then trigger 11x Alice or Lindy within 60 seconds for first response.

    Workflow: `Framer -> n8n -> Attio -> 11x -> Slack` with a `high-intent` tag from Clay enrichment.
  3. 03

    Prune waste weekly

    Use PostHog query paths and Northbeam spend data to exclude terms with clicks but no qualified meetings after 30 days.

    Query rule: if `clicks > 25` and `qualified_meetings = 0`, add the term as a negative keyword.
IfThen
a campaign mixes competitor and non-competitor trafficsplit it immediately; mixed intent corrupts bidding and reporting.
tCPA rises 25% without qualified pipeline growthfreeze expansion and rework the query set before raising budget.

Make the landing page a conversion system, not a brochure

Every B2B Google Ads landing page in 2026 should have 1 promise, 1 proof block, and 1 conversion path. Use Framer + Relume for speed, then wire PostHog to track scroll depth, CTA clicks, and form abandonment. If you need account-level personalization, use Mutiny or Intellimize for 1:1 modules driven by Clay fields like industry, employee count, and tech stack. Keep forms short: 4 fields max, with company email, work role, company size, and pain selector. Relevance AI can triage the inbound lead and classify urgency before Attio updates the record. A 3-second CTA decision beats a 3-minute case study read every time.

  1. 01

    Ship 1 page per cluster

    Build each page in Framer with Relume sections, then clone the structure across 3 clusters instead of inventing a new layout.

    Template: `hero -> proof -> use cases -> objection -> CTA`, repeated with cluster-specific copy.
  2. 02

    Instrument conversion events

    Track page view, CTA click, form start, form submit, and booked meeting in PostHog, then sync the events to Northbeam.

    Event names: `ppc_cta_click`, `ppc_form_start`, `ppc_meeting_booked`, all sent with UTM and gclid.
  3. 03

    Use AI for lead triage

    Run Relevance AI or Clay Chat on each submission to score fit, extract pain, and route urgent leads to 11x.

    Prompt: `Classify this lead as urgent, normal, or junk; cite 2 evidence fields and 1 next action.`
IfThen
a landing page needs more than 1 CTA to make sensethe offer is unclear and the page should be rewritten, not decorated.
form completion is below 40%remove 1 field and move proof above the fold before changing traffic.

Measure qualified pipeline, not vanity lead count

Google Ads for B2B dies when marketing celebrates leads and sales ignores source quality. Use Attio as the CRM, PostHog for behavioural truth, and Northbeam for spend reconciliation; then join them with a RevOps SQL layer in Cursor or Claude Code. Pull weekly cohorts by keyword, landing page, and company fit, then calculate qualified opportunity rate, pipeline created, and payback period. If you are migrating off Salesforce, cut the extra objects and rebuild the pipeline around 6 fields, not 60. Clay should enrich every opportunity with firmographic data so you can see which industries actually convert. The rule is simple: no attribution, no scale.

  1. 01

    Build a single source of truth

    Sync Google Ads, PostHog, Northbeam, and Attio into one cohort view so marketing and sales use the same numbers.

    SQL prompt: `Join gclid, account_id, and opportunity_id; return weekly qualified opps by campaign.`
  2. 02

    Add enrichment to every opp

    Use Clay to backfill employee count, funding stage, and tech stack for each qualified opportunity in Attio.

    Example fields: `employee_band`, `funding_stage`, `has_warehouse`, `uses_segment`.
  3. 03

    Audit in 30-day windows

    Review campaign, keyword, and landing page cohorts every 30 days, because B2B cycles shorter than 30 days are usually noise.

    Decision rule: only scale if `qualified_opps >= 5` and `payback_months <= 9`.
IfThen
lead volume rises but qualified pipeline does nottreat it as a measurement failure until proven otherwise.
a source cannot map from gclid to Attio opportunityexclude it from budgeting decisions and fix the join keys first.
Pitfalls

What goes wrong

Mistake

Buying broad terms because volume looks impressive in Google Ads

Fix

Move to exact and phrase match, then use Clay to validate that each term maps to 15+ high-fit accounts.

Mistake

Sending traffic to the homepage or a generic product page

Fix

Use Framer + Relume pages per intent cluster with 1 CTA and 1 proof block.

Mistake

Letting MQLs sit for hours before follow-up

Fix

Trigger 11x or Lindy from n8n in under 5 minutes and route all urgent forms to Attio.

Mistake

Judging success by CTR, CPC, or raw lead count

Fix

Measure qualified opportunities, pipeline created, and payback using PostHog, Northbeam, and Attio.

Mistake

Running one blended campaign across every intent level

Fix

Split problem, category, and competitor traffic into separate campaigns with separate budgets.

FAQ
Should B2B Google Ads start with Search or Performance Max?▾

Start with Search when intent is clear, because Search gives you query control, negative keywords, and cleaner learning. Add Performance Max only after you have 30–50 qualified conversions and a strong offline conversion signal in Attio. If you cannot explain why a query should convert, PMax will scale confusion faster than it scales pipeline.

How many keywords should a B2B campaign use?▾

Use 10–20 exact or phrase keywords per theme, then keep each theme tied to 1 landing page. In 2026, thin structures beat bloated accounts because Smart Bidding needs clear signals. Clay can validate whether each theme has enough fit accounts, and PostHog can confirm which term paths create meetings.

What is the best way to qualify Google Ads leads?▾

Use 4-field forms, enrichment in Clay, and an AI triage layer in Relevance AI or Clay Chat. Score by company size, use case, and urgency before the lead reaches sales. If the lead is high intent, trigger 11x or Lindy immediately. If the lead is junk, suppress it in Attio and do not let sales waste time.

How should attribution work for long B2B cycles?▾

Use PostHog for behavioural events, Northbeam for spend, and Attio for opportunity outcomes. Join them with Cursor or Claude Code SQL on gclid, account_id, and opportunity_id. Monthly blended ROAS is too blunt for B2B; cohort-level qualified pipeline per keyword is the number that matters.

When should I increase budget?▾

Increase budget only when 2 conditions hold for 30 days: qualified opportunities are stable at 5+ per cluster, and payback is inside your target window. If the cluster has no negative keyword debt, a clear CTA, and a responsive 11x workflow, scale by 20% steps. Bigger jumps usually buy noise, not demand.

Get one system like this in your inbox each week

No fluff, no upsell. Unsubscribe anytime.