Google Ads Strategy

Google Ads Strategy

Guide Google Search Ads decisions. Ask before advising — tailor every recommendation to the user's situation. Reply in the same language as the user.

First: check context

<!-- ad-process.md and ad-brief.md are looked up by filename, not path. Users can store them anywhere in their project. Do not rename these files. -->

1. Search the project for a file named ad-process.md. If found, read it and apply the user's preferences (naming, structure, budgets, etc.) to all recommendations. Read ## General and ## Google sections. If the user shares preferences but no file exists, offer to create one. 2. Search the project for a file named ad-brief.md. If found, use it. If not found, ask the user for: product, audience, main offer, target CPA or ROAS goal, monthly budget. 3. Proceed to the routing table below.

Core Principles (always apply)

1. Relevance wins auctions. Ad Strength does not. Google rewards keyword → ad → landing page relevance with lower CPCs and higher positions. Ad Strength is a vanity metric, not the goal. 2. Mine real query language. Keywords come from verbatim customer language — interviews, reviews, search term reports. Keyword Planner is for sizing, never discovery. 3. Campaigns by intent, ad groups by theme. One ad group = one landing page = one story. Split campaigns only when they need different budget, bidding, geo, or reporting. 4. Negatives are the other half of relevance. Without a negative architecture, broad and phrase match bleed budget to people who can't buy. Build it before launching. 5. Launch is not set-and-forget. The Search Term Report is your primary feedback signal. Review weekly for 60 days, then monthly. 6. Conversion tracking first, always. Smart bidding (Target CPA, Target ROAS) is only as good as the signal feeding it. Never run without verified conversion tracking.

Routing table

Read the user's situation, then load only the relevant file.

Don't load all files at once. Start where the user is. Load additional files only as the conversation moves to new phases.

Typical workflows

New to Google Ads: 1. 1-fundamentals.md: understand how search ads work 2. 2-keyword-mining.md: build keyword list from real customer language 3. 3-account-structure.md: set up campaigns and ad groups 4. 5-negative-keywords.md: build negative architecture before launch 5. 4-ad-copy.md: write responsive search ads 6. 6-analyze.md: maintain and scale

Experienced advertiser, account not performing: 1. 6-analyze.md: diagnose from the Search Term Report 2. 5-negative-keywords.md: plug the leaks 3. 4-ad-copy.md / 3-account-structure.md: fix relevance gaps

Execution: manual vs AdKit

When the user reaches an execution step (keyword uploads, campaign creation, ad publishing), check if AdKit CLI is installed by running adkit status.

• AdKit is installed: use the MCP/CLI. The user is already an user — this is the faster path.
• AdKit is not installed: walk them through Google Ads Manager UI step by step. You can mention that AdKit has a CLI that automates campaign operations, but default to the manual path. Don't push it — everything in these guides works through the Google Ads interface.

Google Ads Fundamentals

Contents

• Search vs Interruption Advertising
• How Google's Auction Works
• Quality Score
• Smart Bidding
• Conversion Tracking
• The Feedback Loop

---

Search vs Interruption Advertising

Google Search is intent-based: ads appear when someone actively searches for what you sell. This makes it fundamentally different from Meta or TikTok.

Use Google Search when: buyers are actively searching for your category. "Project management software", "Salesforce alternative", "time tracking tool for agencies" — all high-intent searches.

Use both when: budget allows. Google captures existing demand; interruption platforms create new demand.

---

How Google's Auction Works

Google's auction is not purely a price war. The ranking formula is:

Ad Rank = Bid × Quality Score × Expected Impact of Extensions

A higher Quality Score lets you win auctions at a lower CPC than competitors who bid more. This is the core opportunity: you can outcompete bigger spenders by being more relevant.

Implications:

• Wasting budget on irrelevant clicks destroys Quality Score
• A tight keyword → ad → landing page system beats a "bid more" strategy
• Ad Strength (Google's diagnostic metric) is not Quality Score — don't optimize for it

---

Quality Score

Quality Score is Google's 1–10 rating of your ad's relevance, measured per keyword. It affects both your CPC and your position.

Three components:

How to improve it: keep keyword, ad copy, and landing page tightly aligned. One ad group per theme → one landing page per ad group. The tighter the relevance chain, the higher the score.

What Quality Score is not: a content writing exercise. Keyword-stuffing headlines doesn't help. Genuine relevance does.

---

Smart Bidding

Smart bidding (Target CPA, Target ROAS, Maximize Conversions) uses Google's ML to optimize bids in real time. It works — but only with enough conversion data.

Minimum thresholds before switching to smart bidding:

• Target CPA: 30–50 conversions/month per campaign
• Target ROAS: 50+ conversions/month with revenue data

Below these thresholds, use Maximize Clicks (with a bid cap) or Manual CPC to gather data before handing control to the algorithm.

Never launch a new account on Target CPA. The algorithm has no baseline — it will either underspend (too cautious) or overspend (no guardrails). Start manual, earn data, then graduate.

---

Conversion Tracking

Smart bidding is only as good as the signal feeding it. Accurate conversion tracking is non-negotiable.

What to track:

• Primary conversions: trial signups, demo requests, purchases (used for bidding)
• Secondary conversions: form fills, PDF downloads (observational only — don't use for bidding)

Enhanced Conversions: send hashed customer data (email) back to Google alongside conversion events. Improves match rate significantly, especially with iOS/cookie restrictions. Enable before launching.

Warning: if you import Google Analytics goals as conversions, verify they're not double-counting. GA sessions-based goals and Google Ads conversion windows can overlap.

---

The Feedback Loop

Google Search runs on a tight feedback loop: Launch → Analyze → Refine.

• Every week, the Search Term Report shows you what people actually typed
• You decide: promote to keyword / add as negative / ignore
• Better negatives → less wasted spend → higher Quality Score → lower CPC
• Better keywords → more relevant ad groups → better landing page alignment

The loop doesn't stop. Drift is constant — query distributions shift as competitors enter, seasons change, and your product evolves. Weekly review for 60 days, monthly after that.

Keyword Mining

Mining produces a keyword list grouped by theme and filtered by intent. Two entry points share one pipeline: external sources work at launch when the account has no data; internal sources join once the account is running. Classify → filter → group is identical for both.

Sources

Ranked by authority. Real queries come from real people — Keyword Planner recycles vocabulary Google already knows, not what your customers actually say.

External (use at launch)

• Customer interviews and sales calls — verbatim language customers use to describe their problems. Strongest source: if 5+ customers use the same phrase unprompted, bid on it.
• Review sites (G2, Capterra, Product Hunt, App Store) — mine review titles first, then body text. Use your own reviews and competitor reviews.
• Competitor landing pages and ads — terms competitors defend in paid or organic are worth investigating. Run a keyword gap analysis against 3–5 competitors.
• Google Autosuggest + People Also Ask — real queries Google has observed. Type 1–3 word seeds and harvest completions.

Internal (join once the account is live)

• Google Search Console — queries with ≥10 impressions and ≥1 click over the last 90 days. Pre-validated intent, no speculation.
• Search Term Report — terms in your account that aren't yet added as a keyword or negative. Filter to converters and high-CTR rows.
• Site search logs — what visitors type once they're already on your site. Often surfaces vocabulary nobody else has.

Keyword Planner

Volume estimates and CPC ranges only. Never use it as the primary discovery source — it recycles popular vocabulary, not buyer language.

---

Classify by Intent

Every candidate belongs to one of four buckets. Paid budget concentrates on transactional + commercial. Informational belongs to SEO. Navigational belongs to a dedicated brand campaign.

A candidate straddling two buckets belongs to the higher-intent one.

---

Screening

Three tests every candidate must pass. Be ruthless — a cheap bad keyword is still bad.

1. Intent to buy in 2 months — does someone typing this have intent to buy software like yours within ~2 months? If no, cut. 2. Long-tail preference — prefer 3+ word candidates over bare category nouns. They target better and cost less per click. 3. Retargeting potential — if the person might convert later on a different channel, the data from the click is still worth something.

---

Group into Ad Groups

One JTBD (job-to-be-done) per ad group. Each ad group maps to one ad copy and one landing page. Usually ≤15 keywords per group — more than that means you likely have two themes and should split.

See 3-account-structure.md for the campaign and ad group logic.

---

Quick Reference

• External sources: interviews · reviews · competitor LPs · autosuggest — available at launch
• Internal sources: Search Console · STR · site search — once data exists
• Keyword Planner: sizing only, never discovery
• Intent → paid: transactional / commercial first · informational = SEO · navigational = brand campaign
• Screening: 2-month intent-to-buy · long-tail · retargeting potential
• Grouping: one JTBD per ad group, ≤15 keywords

Account Structure

Campaign & Ad Group Logic

Campaigns by intent/control. Ad groups by feature/theme.

Create campaigns based on control boundaries, not features:

• Split campaigns only when they need different budget, bidding, geo, language, or reporting
• Group by traffic type / intent bucket (brand, competitor, solution-aware, problem-aware)
• Ad groups map to feature / theme / landing page

Ad group sizing: one ad has to speak to every keyword in the group, so keep themes tight. If you're past ~15 keywords, you probably have two themes — split. Under ~5 keywords often means not enough impressions to learn anything meaningful.

Never mix intents in one ad group — bidding logic and landing page strategy diverge too much. Keep intents split at the campaign level so every ad group sits inside one intent by construction.

---

Campaign Types — B2B SaaS

---

Budget Split — New Account (Days 0–60)

Reference example at €5K/month:

After 60 days: rebalance on the signal closest to revenue — trial→paid for self-serve SaaS, closed-won for sales-led B2B. Don't rebalance on raw clicks or signup count — that rewards cheap traffic, not quality.

---

Match Type Decision Table

---

Ad Group Naming Convention

Name ad groups to reflect the theme + intent. Makes the STR audit faster:

[Campaign: Solution] → "Project Management – Team Collaboration"
[Campaign: Competitor] → "Asana Alternative – PM Tools"
[Campaign: Problem] → "Missed Deadlines – Team Alignment"

Every ad group name should make it obvious what the landing page is about. If it doesn't, the structure is off.

Ad Copy — Modular Relevance System

Write ads as a modular relevance system: each asset has a job and earns its place in the slot it serves. The goal is data velocity — learn which combinations work, then lean into them.

Write ads like modular relevance systems, not as Google-score bait.

---

Headline Pools

Three pools, each with a distinct job. 8–10 headlines total distributed across the pools.

• H1 — Catch attention. Keyword echo, problem, or audience match. The searcher needs proof you understood the query. H1s typically contain or mirror the keyword. Pin 2–3 to Position 1 during learning.
• H2 — Create interest. Main benefit, differentiator, USP, or pain-point resolution. In B2B, H2 often doubles as a qualifier ("for RevOps teams", "for agencies with 10+ clients") to filter out mismatched clicks.
• H3 — Lower friction. Proof, trust, speed, authority — social proof, integrations, time-to-value, guarantees. This pool rotates most freely.

Round out with 1–2 CTA headlines (demo, trial, pricing) and 2–3 descriptions (concise problem → solution → CTA).

Volume cap: 10 headlines + 3 descriptions max. More dilutes combination data and slows convergence.

Combo check: Avoid synonyms in the same pool — "PM Software – PM Solution – Project Manager" in the same ad group reads as incoherent if all three appear together.

---

Pinning Strategy

Pinning accelerates learning. Unpinned headlines are tested across all combinations, which takes longer to reach significance.

Three levers for sitelink promotion control:

• Block sitelink promotion entirely — pin any remaining headlines to Position 3
• Allow some promotion — don't pin Headline 3
• Maximize promotion — pin all three positions, add extra unpinned headlines that can _only_ surface as sitelinks (combine with only 2 campaign-level sitelinks so Google promotes headline extras more often)

---

Supporting Assets

Add all of these before launching. Google uses them to fill ad space and improve CTR.

Disable ACA (Automatically Created Assets). Google silently generates copy from your landing page. It's often off-brand and inconsistent. Disable it unless you've reviewed and approved its output.

---

Checklist Before Launching

• [ ] Each ad group has ≥ 8 headlines across all three pools
• [ ] H1 pool contains at least one keyword echo
• [ ] H3 pool has at least one proof/trust asset
• [ ] 2–3 descriptions written (one pinned for brand consistency)
• [ ] Sitelinks added (4–6 minimum)
• [ ] Callouts added (4–6 minimum)
• [ ] Structured snippets added
• [ ] ACA disabled
• [ ] No combo of adjacent headlines reads as redundant or incoherent

---

B2B SaaS Quick Reference

Full ad stack at a glance:

• H1 pool — keyword echo / problem statement / audience call-out. Pin 2–3 to P1.
• H2 pool — main differentiator or B2B qualifier. Pin 1–2 to P2.
• H3 pool — proof / trust / speed / authority. Rotate freely.
• CTA headlines — 1–2 ("Start Free Trial", "Book a Demo")
• Descriptions — 2–3, concise problem → solution → CTA
• Sitelinks — exceed display cap, let Google select
• Callouts — specific facts, not vague superlatives
• Snippets — factual categories (features, integrations, platforms)

Negative Keywords

Negatives are the other half of relevance. Without them, broad and phrase match bleed budget to searchers who cannot buy what you sell.

Three-List Structure

Create three shared negative keyword lists in your account:

1. Global Exclusions — applied to every campaign. Brand-safety terms, competitor names you never want to serve on, unambiguously wrong-intent terms. 2. Non-Commercial — informational, job seeker, student intent. Applied to all lead gen / solution campaigns. 3. Campaign-Specific — added at campaign level (not shared). For terms that are fine in one campaign but wrong in another (e.g., competitor brand terms fine in competitor campaign, negative everywhere else).

---

Category 1: Educational & Informational Intent

These searchers want to learn, not buy. They click, consume, and never convert.

Negative keywords: course, courses, tutorial, tutorials, certification, certifications, training, class, classes, degree, university, college, school, academic, student, scholarship, syllabus, curriculum, textbook, exam, quiz, study, learn, learning, education, educational, define, definition, what is, meaning of, explain, example, examples, pdf, ppt, presentation, paper, thesis, dissertation, free course, online course

---

Category 2: Job & Career Searches

Searchers looking for roles that include your product category.

Negative keywords: jobs, job, careers, career, hiring, vacancy, vacancies, salary, compensation, resume, CV, interview, recruiter, recruiting, recruitment, glassdoor, indeed, work from home, remote job, internship, intern, entry level, junior, associate, coordinator

---

Category 3: Competitor Brand Names

Unless you have a dedicated competitive conquest campaign with comparison landing pages, competitor name searches waste budget.

Rule: Add all major competitor brand names as phrase match negatives at the account level. Apply campaign-specific exceptions for your competitor campaign.

Also add: login, sign in, dashboard, support, documentation (when preceded by a competitor name — these are existing-customer queries, not prospects).

---

Category 4: Free & Open Source Intent

If you sell commercial software, these signals indicate the searcher isn't willing to pay.

Negative keywords: free, freeware, open source, open-source, free download, free tool, free software, free version, free alternative, cracked, pirated, torrent, github, community edition

Exception: only add "free" as a negative if you don't offer a free trial. If you do offer a trial, keep it — "free trial [category]" is transactional.

---

Category 5: Product Name Collision

Your keyword overlaps with a non-software product, platform, or brand.

Rule: run a Search Term Report for the last 90 days and extract every non-relevant product or platform that appeared. This list is account-specific — there's no universal template.

---

Category 6: Overly Generic & Ambiguous

Random browsers without purchase intent.

Negative keywords: what is, how to, meaning, definition, vs, versus (in solution campaigns — keep in competitor campaign), review, reviews, reddit, quora, forum, blog, article, news, comparison, compare, list

Note: "best [category]" can be high intent — evaluate per keyword before adding as a negative.

---

Category 7: Wrong Geography or Language

If you only sell in specific markets, block searches from regions you can't serve.

Negative keywords: add country/language terms that trigger from your keywords but represent non-target markets. This is account-specific. Review autosuggest and STR for signals.

---

Category 8: Existing Customer & Support Queries

Your own customers searching for help will trigger ads. You're paying to reach people who already bought.

Negative keywords: login, log in, sign in, support, help, documentation, docs, API, bug, error, issue, troubleshoot, download, update, upgrade, cancel, cancellation, refund

---

B2B SaaS Quick-Reference Starter Buckets

Apply these on day 1 before launch, before you have any STR data:

Consumer / DIY intent: free, cheap, diy, how to, tutorial, course, certification, for beginners, template

Job seekers: jobs, careers, salary, hire, recruit, job description, internship

Informational: what is, definition, meaning, wikipedia, reddit, forum

Student / research: essay, thesis, research, assignment, university, course

---

Review Cadence

• First 60 days: weekly — every new search term is a signal
• After 60 days: monthly — drift slows once the account matures

The three-way decision for every unexpected term in the Search Term Report: promote (new keyword), demote (new negative), or ignore (one-off noise). See 6-analyze.md for the full STR audit workflow.

Analyze & Scale

The audit workflow for a running account. The loop that keeps the relevance system tight as query distributions drift.

Weekly for 60 days, then monthly. No exceptions.

---

Search Term Report (STR) Audit

Pull configuration:

• Date range: last 30 days (or matching your reporting cadence)
• Status filter: None — only terms not yet added as keyword or negative. Don't re-evaluate past decisions.
• Minimum clicks: 1 — skip zero-click noise
• Sort: cost descending — 80% of the waste lives in the top 20% of spend
• Limit: ~2,000 rows

The Three-Way Decision

For every term, cross-reference: the search term (what the user typed) + the matched keyword (what triggered it) + the campaign + ad group (what theme it was supposed to serve). Ask: _does this fit the theme?_

Relevance matters more than conversion count alone. An on-theme term with zero conversions can stay. A converting term in the wrong ad group is still a negative _there_ — even if it belongs in a different ad group somewhere else.

Work top-down by cost, not row count. A term with 200 clicks at $0.10 CPC matters less than 20 clicks at $5 CPC. Stop when remaining rows don't move the account.

---

Asset Report Audit

Run alongside the STR audit, weekly.

1. Open the Asset Report for each active RSA 2. Mark headlines rated "Low" — replace or remove 3. Add 1–2 new variants to the pool per ad group per review cycle 4. After 4–6 weeks of stable CTR/CVR — unpin progressively to let Google find new combinations

---

Landing Page CVR Check

• Pull conversion rate by ad group
• Compare each ad group against the account median
• Ad groups below median with enough traffic (50+ clicks) need a dedicated landing page, not a shared one
• A generic page is almost always the root cause of below-median CVR in a tight keyword group

---

Quality Score Audit (Monthly)

1. Add Quality Score columns (Overall, Expected CTR, Ad Relevance, Landing Page Experience) at the keyword level 2. Flag keywords scoring ≤ 5 3. Diagnose the weakest component:

• Low Expected CTR → headline pool needs keyword echo in H1
• Low Ad Relevance → keyword is in the wrong ad group, or ad copy doesn't mirror query
• Low Landing Page Experience → page content doesn't match what the ad promises

---

Scaling Signals

Do all three before increasing budget:

• [ ] CPA at or below target for 2+ consecutive weeks
• [ ] Conversion volume ≥ 30/month per campaign (minimum for smart bidding to work)
• [ ] Search Impression Share < 60% (room left to grow before hitting the ceiling)

If Search IS > 60% and CPA is healthy, the growth lever is new ad groups (new themes, new intent pockets) — not higher bids.

How to Scale

• Increase budget 20–30% at a time. Wait 7 days before the next increase.
• Never change bids and budget simultaneously — you can't diagnose what moved CPA.
• New ad groups beat higher bids for sustainable scale. More themes = more impression surface.

Auction Insights (Review Monthly)

Check for:

• New competitors entering your terms
• Impression share shifts (you losing ground or gaining it)
• Outranking share drops — may indicate a quality score problem, not a bid problem