The Google Ads account structure I run for clients in 2026

The biggest mindset shift in Google Ads over the past two years: your job stopped being to pick keywords and write headlines, and started being to feed the algorithm clean signals. Get the inputs right and Smart Bidding outperforms manual every time. Get them wrong and Performance Max is the fastest way to set your budget on fire.

Here's the account structure I deploy when I take over a new account in 2026.

The 4-campaign baseline

1. Brand search (manual CPC)

One campaign, one ad group, exact + phrase match of your brand name. Manual CPC at a low ceiling (₹15-30 in India). Why manual: smart bidding will overspend here because brand traffic converts by default. Cap it.

2. Non-brand search (Maximize conversions, target CPA)

This is where 60% of your budget goes. Organized by intent cluster, not by keyword. Typical ad groups:

• Buyer-intent commercial: "{service} cost", "{service} near me", "best {service}"
• Problem-aware educational: "how to fix {problem}", "why does {symptom} happen"
• Comparison: "{competitor} vs", "alternatives to {competitor}"

One responsive search ad per ad group. Pin headline 1 to your brand promise. Let Google rotate the rest. Bidding strategy: Maximize conversions with target CPA once you have ~30 conversions in a 30-day window.

3. Performance Max (one campaign, audience signals + asset groups by buyer)

PMax is no longer optional in 2026 — it's where the algorithm finds incremental volume your search campaigns miss. But it only works with structured signals:

• Audience signals: Upload your customer email list (CRM customer match), recent purchasers, and 1% lookalikes. PMax uses these as starting points.
• Asset groups: One per buyer segment. Different headlines, descriptions, and images per segment.
• Account-level negative list: Brand terms, irrelevant queries. Otherwise PMax will cannibalize search.
• Conversion goal weighting: Tell Google a Purchase is worth 10× a Lead. Otherwise it optimizes for whatever's easier.

4. Retargeting (manual CPM display)

Standard display campaign, cap at 1-2 impressions per user per day, audiences segmented by funnel stage. Critical: exclude converted users at the account level so you don't waste impressions.

The non-negotiable conversion stack

If your conversion tracking is wrong, none of this works. The 2026 stack:

• Server-side GTM (Stape or self-hosted). Browser-only pixels are losing 25-30% of conversions to ad-blockers and privacy modes.
• Enhanced conversions turned on. Send hashed email at minimum; phone if you have it.
• Offline conversion imports for any lead-gen account. If a lead becomes a customer, upload the conversion value back to Google within 30 days. This is how Smart Bidding learns to find higher-value leads, not just cheaper leads.
• Single Purchase / Lead event as the only "primary" conversion. Everything else is "secondary." Don't let secondary conversions bid against primary.

What I monitor weekly (and what I ignore)

Watch: Search terms report, conversion volume by campaign, cost-per-conversion by ad group, asset performance ratings, search lost (rank), search lost (budget).

Ignore: Impression share alone, CTR (it's a vanity metric in Smart Bidding), quality score (matters less than people think for non-brand).

Bid strategy timing

This is the part most people get wrong. New campaigns start with Maximize Clicks for the first 7-10 days to gather data. Once you have 30+ conversions in the last 30 days, switch to Maximize Conversions. Once that's stable, set a Target CPA 10% above the average. Then drop the CPA target by 10% every 14 days until volume starts shrinking — that's your true cost ceiling.

The brands beating their competition in Google Ads in 2026 aren't writing better headlines. They're feeding the algorithm cleaner signals and giving it room to optimize. Set the structure right, then get out of its way.