How to Run Google Ads in 2026: A 13-Step Launch Guide

By Theo Werner 2026-07-11 21 min read

Running Google Ads in 2026 comes down to one repeatable sequence: install conversion tracking, launch a tightly structured Search or Performance Max campaign, feed it first-party data, and let Smart Bidding optimize against at least 30 conversions per month. Google's average cost per click reached $5.42 in 2026 and the average conversion rate hit 8.18%, so the gap between a profitable account and a wasted budget is decided by setup choices you make before a single ad serves. This guide walks through all 13 launch steps, from the Google tag to weekly negative keyword mining, with the exact thresholds, config blocks, and scripts that separate campaigns that compound from campaigns that oscillate.

How Google Ads works in 2026, and what actually changed

Google Ads is an auction, and that mechanic has not changed in a decade. Every time someone searches, Google ranks eligible ads by a combination of bid, expected click-through rate, ad relevance, and landing page experience, then charges the winner no more than the minimum needed to hold its position. What changed is how much of that auction Google now runs on your behalf. In 2026 you are not setting keyword bids by hand. You are engineering the inputs, conversion data, audience signals, budget, and creative, that Google's models optimize against.

Three campaign types dominate the year: Search, Performance Max, and the new AI Max for Search. AI Max is the headline shift. Google confirmed that Dynamic Search Ads are being upgraded to AI Max, with voluntary upgrades open since April 2026, automatic upgrades beginning in September 2026, and the final Dynamic Search Ads sunset extended to February 2027. According to Google's own announcement, campaigns using the full AI Max feature suite, search term matching, text customization, and final URL expansion, see an average of 7% more conversions or conversion value at a similar CPA or ROAS.

The money behind the platform explains the pace of change. In Q3 2025 alone, Google Search advertising generated $56.6 billion and YouTube ads added $10.3 billion, part of Alphabet's $102.3 billion in quarterly revenue, as Storyboard18 reported and Alphabet's Q3 2025 Form 10-Q confirms. When a channel prints those numbers, Google can afford to rebuild the interface every year, and it does.

Performance Max matured too. Through 2026 Google added asset group-level reporting, channel-level budget transparency, placement exclusion controls, and campaign-level negative keywords, the exact levers advertisers spent three years requesting, detailed in Google's Performance Max steering updates. A June 2026 bidding update introduced Bidding Target Optimization, which pulls budget-limited campaigns that consistently over-deliver their targets back toward those targets instead of letting performance drift. The practical takeaway: get the inputs wrong and automation amplifies the mistake, which is why this guide spends more time on tracking and structure than on ad copy. If you want a team to run this end to end, our paid advertising practice builds these systems for B2B accounts daily.

Prerequisites: what you need before you spend a dollar

Before you create a single campaign, assemble the stack below. Launching without it is the most common reason new accounts burn budget in week one, because Google's automation starts optimizing immediately, even against broken or missing signals. Every item includes a version or specification so you can confirm you are current as of July 2026.

• A Google Ads account in Expert Mode, not Smart Mode. Smart Mode hides the settings this guide depends on. During account creation, select the text link 'Create a campaign without a goal's guidance' to stay in Expert Mode.
• Google Analytics 4 (GA4), linked to Google Ads. Universal Analytics stopped processing data in 2023 and its properties were shut down in 2024. If your reports still reference it, that data path is dead.
• The Google tag (gtag.js). This is the current unified tag that replaced the older global site tag. Deploy one tag ID site-wide.
• Google Tag Manager (web container). Optional but recommended so you can manage tags without a code deploy for every change.
• At least one primary conversion action (purchase, qualified lead, demo request) with a dollar value attached, not a bare contact event.
• Enhanced conversions enabled. This hashes first-party data with SHA256 before sending. Advertisers typically report 5% to 15% more conversions after enabling it, per Google's setup documentation.
• Consent Mode v2 if you serve users in the EEA or UK, so tags honor the ad_storage and analytics_storage consent states and still model conversions when consent is denied.
• A conversion-ready landing page: mobile-first, Largest Contentful Paint under 2.5 seconds, and one primary call to action above the fold.
• A seed negative keyword list of 50 to 100 terms prepared before launch, not bolted on after you have wasted spend.
• Billing configured with a valid payment method and, ideally, an account-level monthly budget cap.

Skipping enhanced conversions or Consent Mode is not cosmetic. In 2026, iOS privacy defaults and cookie consent gaps cause unmodeled accounts to under-report conversions by double digits, which starves Smart Bidding of the exact signal it needs to bid accurately. Fix measurement first, because everything downstream depends on it. Teams that get conversion tracking right also invest in conversion rate optimization so the clicks they pay for actually turn into pipeline.

The 2026 benchmark numbers to calibrate against

You cannot judge a campaign without a baseline, and guessing at good performance is how accounts get paused prematurely or scaled recklessly. WordStream's 2026 benchmark study analyzed more than 13,000 search advertising campaigns across 23 industries running from April 2025 through March 2026. The headline averages: a $5.42 cost per click, a 6.64% click-through rate, and an 8.18% conversion rate. Use these to sanity-check your own numbers, not as targets to hit precisely, because the spread between industries is enormous.

WordStream's 2026 study of more than 13,000 campaigns across 23 industries puts the average Google Ads cost per click at $5.42, the average click-through rate at 6.64%, and the average conversion rate at 8.18%, with year-over-year results staying broadly stable from 2025.

The table below pulls the verified data points from that study. If your cost per click is triple your industry benchmark, your Quality Score or match types are the first suspects, not your budget.

Read the table as a distribution, not a scoreboard. High-intent verticals such as pet care, automotive repair, and education convert at 13% to 16% because searchers arrive ready to act. Regulated and high-ticket categories such as legal and finance carry expensive clicks and lower conversion rates, so their economics depend on lifetime value rather than first-purchase margin. Pull your own account's trailing 90-day CPC, CTR, and conversion rate, line them up against the WordStream 2026 benchmarks, and you will know within minutes whether your problem is traffic cost, ad relevance, or landing page conversion.

Choose your campaign type: Search, Performance Max, or AI Max

The single most consequential decision at launch is campaign type, because it determines how much control you keep and how much budget you need to feed the algorithm. For B2B lead generation with a defined keyword set, a Search campaign with Smart Bidding is the correct default. For e-commerce with a product feed and more than 50 conversions per month, Performance Max earns its keep by spanning Search, Shopping, Display, YouTube, Gmail, Maps, and Discover from one campaign. AI Max sits between them: it layers Google's query discovery onto a Search campaign and is the mandatory successor to Dynamic Search Ads, so adopting it deliberately now beats being force-upgraded in September 2026.

The comparison below maps each type to its ideal use, control level, and realistic minimum monthly budget. Minimum budgets assume you need enough conversion volume to clear Smart Bidding's learning thresholds within a month.

Two rules cut through the noise. First, do not start with Performance Max if you have no conversion history, because a black-box campaign with no data to learn from will spend against Google's priors, not yours. Start with Search, accumulate 30 or more conversions, then expand. Second, always run a separate brand exact-match campaign. It protects your branded traffic at a low cost per click and keeps competitors from buying attention you already earned. Everything after this section assumes you are building a Search campaign, which is the foundation every other type builds on.

Building your Search campaign, steps 1 through 6

This is the core build. Follow the six steps in order. Each one lists the exact click path, what you will see on screen, and the output or config you should end up with. Do not skip the without-guidance step, because starting from a goal preset silently locks several settings you will want to change.

1. Create a campaign without guidance. Screenshot: the campaign creation screen shows a grid of objective cards (Sales, Leads, Website traffic, and so on). Scroll below the grid and click the text link 'Create a campaign without a goal's guidance'. Output: the campaign opens with no objective preset, giving you full access to networks, bidding, and location settings.
2. Select Search as the campaign type. Screenshot: a row of campaign-type tiles appears; click the magnifying-glass Search tile, then choose Website visits or Leads as the desired result. Output: campaign type locked to Search.
3. Turn off partner and Display expansion. Screenshot: under the Networks panel, uncheck 'Include Google search partners' and 'Include Display Network'. Output: pure Search delivery, which keeps early data clean so Smart Bidding learns from qualified intent, not cheap Display impressions.
4. Set locations, language, and budget. Under Locations, click Location options and choose 'Presence: People in or regularly in your targeted locations', not the default 'Presence or interest', which leaks budget to people merely searching about your region. Use the config below as a template.

   Campaign name: US-B2B-Search-Leads
   Networks:      Search only (partners OFF, Display OFF)
   Locations:     United States [Presence]
   Language:      English
   Bidding:       Maximize Conversions (move to tCPA $120 at 30 conv/mo)
   Daily budget:  $100.00   # averages about $3,040/mo

5. Build ad groups by tight theme. One theme per ad group keeps Quality Score high and ad relevance obvious. Structure it like this:

   Ad group: Project Management Software
     Exact:  [project management software], [project management tool]
     Phrase: project management platform, project management app
     Negatives: [free], [jobs], [template], [salary]

6. Write Responsive Search Ads. Provide 15 headlines and 4 descriptions per ad group, pin one brand headline to position 1, and aim for an Excellent Ad Strength. Output example:

   Headlines (15):
     1. Project Management Software
     2. Ship Projects On Time
     3. Free 14-Day Trial, No Card
     ...
   Descriptions (4):
     1. Plan, track, and ship work in one place. Start your free trial today.
   Ad Strength: Excellent

By the end of step 6 you have a Search-only campaign, one or two tightly themed ad groups, exact and phrase keywords, and at least one high-strength Responsive Search Ad. Do not publish yet. A campaign without conversion tracking is a campaign that cannot optimize, so the next three steps wire up measurement before you spend anything.

Conversion tracking and the Google tag, steps 7 through 9

This is the section that decides whether Smart Bidding works or wanders. If Google cannot see conversions accurately, every automated bid is a guess. Complete all three steps and verify each one with Google Tag Assistant before launch.

7. Install the Google tag. In Google Ads, open Tools, then Data manager, then create or copy your tag ID (format AW-XXXXXXXXXX). Place the base snippet in the head of every page. Screenshot: Tag Assistant shows a green Tag detected badge on page load.

   <!-- Google tag (gtag.js) -->
   <script async src='
   .com/gtag/js?id=AW-XXXXXXXXXX'></script>
   <script>
     window.dataLayer = window.dataLayer || [];
     function gtag(){dataLayer.push(arguments);}
     gtag('js', new Date());
     gtag('config', 'AW-XXXXXXXXXX');
   </script>

8. Create a conversion action and fire the event. In Data manager, create a conversion action, set the category (for example Submit lead form), assign a value, and choose different values per conversion if they vary. Place the event snippet on the thank-you page or trigger it on form submission. Output: the conversion registers within 3 to 24 hours in the Conversions column.

   <script>
     gtag('event', 'conversion', {
       'send_to': 'AW-XXXXXXXXXX/AbC-D_efGh12_34',
       'value': 120.0,
       'currency': 'USD',
       'transaction_id': ''
     });
   </script>

9. Turn on enhanced conversions. Enhanced conversions send hashed first-party data so Google can match conversions it would otherwise lose to cookie and privacy gaps. Pass user-provided data with the tag, hashed with SHA256. Starting in April 2026, Google Ads accepts user-provided data from website tags, Data Manager, and API connections at the same time, so you no longer pick one method.

   gtag('set', 'user_data', {
     'email': '[email protected]',
     'phone_number': '+15551234567',
     'address': {
       'first_name': 'john',
       'last_name': 'doe',
       'postal_code': '10001',
       'country': 'US'
     }
   });

Verify before you launch. Open Google Tag Assistant, load your site, submit a test lead, and confirm three things: the base tag fires on every page, the conversion event fires once on the confirmation page, and the enhanced conversions status in Google Ads reads Recording rather than Needs attention. Skipping this verification is the number-one cause of campaigns that spend for a week and report zero conversions, which then trains Smart Bidding on garbage. Measurement is the foundation; treat it like production code.

Bidding, audiences, and launch, steps 10 through 13

With structure and tracking in place, the final four steps set the bidding strategy, attach audience signals, block junk traffic, and publish. Resist the urge to over-configure. The goal at launch is a clean campaign with accurate signals, not a maze of manual adjustments Smart Bidding will ignore anyway.

10. Set the bidding strategy. If the account has no conversion history, start with Maximize Conversions or, for pure traffic and keyword discovery, Maximize Clicks with a max CPC cap. Once the campaign clears 30 conversions in a rolling 30-day window, switch to Target CPA. Screenshot: the Bidding panel shows Maximize conversions selected with an optional Set a target cost per action checkbox left unchecked at first.
11. Add audience signals in Observation mode. Attach your Customer Match lists, in-market segments, and remarketing lists in Observation, not Targeting. Observation feeds the model without restricting reach; Targeting narrows delivery to only those users. Output: audiences appear in reporting with observed performance but no reach limitation.
12. Add negative keywords and a shared list. Apply your 50 to 100 seed negatives, then attach an account-wide shared negative list for permanent junk.

    Shared library > Negative keyword lists > Account-wide junk
      free, cheap, diy, tutorial, jobs, salary,
      reddit, crack, torrent, download, meme

13. Run the pre-launch checklist and publish. Confirm: conversion tracking Recording, Search-only networks, Presence-based location, budget set, at least one Good or Excellent RSA, negatives applied, and audiences in Observation. Then publish. Output: campaign status reads Eligible and, within a day, Eligible (Learning) as Smart Bidding calibrates.

The moment you publish, the learning phase begins, and your job changes from building to restraint. For the first 7 to 14 days, do not edit budgets, targets, or bidding by more than roughly 20% at a time, because larger changes reset learning and restart the clock. Watch impressions and search terms daily, add negatives, and otherwise leave the campaign alone. The most common self-inflicted wound in a new account is an impatient manager making five changes in week one and never letting the algorithm stabilize.

Smart Bidding strategies and the conversion thresholds that make them work

Smart Bidding is only as good as the conversion volume feeding it. Each strategy has a practical minimum, and these thresholds are measured at the campaign level, not the account level. If you run three campaigns at 15 conversions each, none of them reaches the 30-conversion Target CPA threshold, and all three bid erratically. Consolidation beats fragmentation almost every time. The table below maps each strategy to what it optimizes and the volume it needs to stabilize.

In 2026 the Smart Bidding learning phase typically lasts 7 to 14 days for standard Search campaigns, and Google needs roughly 30 conversions in 30 days for Target CPA and 50 for Target ROAS to calibrate. Every reset from a large budget or target change costs another one to three weeks of volatile performance.

Two operational rules follow from this. First, pick the strategy that matches your volume, not your ambition. A campaign generating 18 conversions a month belongs on Maximize Conversions, not Target ROAS, and forcing the higher-threshold strategy produces exactly the bid oscillation people blame on the algorithm. Second, respect the June 2026 Bidding Target Optimization behavior: budget-limited campaigns that consistently beat their targets are now pulled back toward those targets, so if you want more volume, raise the budget rather than expecting the system to keep over-delivering. When volume is genuinely thin, feed the model conversion value data so it learns which query and audience combinations drive revenue, not just raw form fills.

Keyword match types and the broad match debate

Match types control how loosely Google maps a search to your keyword, and in 2026 they behave differently than they did even two years ago. Broad match now uses large language models to interpret intent, which makes it far better at surfacing high-converting long-tail queries and far more dangerous on the wrong bidding strategy. The recurring debate, broad match versus exact match, has a clear resolution: it depends entirely on whether Smart Bidding is steering the spend.

• Exact match [keyword]: reserve for proven, high-intent, and brand terms. Once validated, roughly 60% of budget should sit here where intent is unambiguous.
• Phrase match (keyword): the middle ground, capturing intent that preserves word order and meaning without the tight constraint of exact.
• Broad match keyword: discovery only, and only paired with Smart Bidding. Allocate 20% to 30% of budget here to find queries you would never think to add manually.
• Never run broad match on Manual CPC. It is the fastest way to drain a limited budget on irrelevant queries, because nothing is filtering intent in real time.
• Review the Search Terms report daily for the first two weeks of any broad match campaign, then at least weekly after that.
• Promote winners. When a broad-match query clears your CPA target repeatedly, add it as an exact-match keyword in a dedicated ad group so you control its bid and copy.
• Defend your brand with a separate exact-match brand campaign so you are not paying discovery prices for people already searching your name.

Google's guidance is that broad match plus Smart Bidding plus audience signals is the modern growth pattern, and Google's broad match documentation makes the pairing requirement explicit. That is genuinely true when conversion tracking is accurate and your negative list is aggressive. It is also true that broad match with weak tracking or manual bids is the single fastest way to waste money in the platform. Both statements coexist. The safe sequence for most B2B accounts is to launch on exact and phrase, prove the economics, then introduce a controlled broad-match discovery campaign on Smart Bidding with a hard daily budget and a daily search-terms review.

First-party data, Customer Match, and audience signals

First-party data is the durable advantage in 2026, because it is the one signal your competitors cannot buy or copy. Customer Match lets you upload hashed customer lists, email, phone, and address, so Google can match them to signed-in users and use them as bidding signals. Google lowered the minimum list size to 100 active users across Search, Display, and YouTube, as Search Engine Land reported, which finally makes the feature usable for smaller B2B lists. Format the upload like this before hashing:

Email,Phone,First Name,Last Name,Country,Zip
[email protected],+15551234567,John,Doe,US,10001
[email protected],+15559876543,Anna,Smith,US,94105

Eligibility has two tiers. All policy-compliant advertisers can use Customer Match in Observation mode and for Exclusions immediately. To use it in Targeting mode with manual bid adjustments, you need at least 90 days of Google Ads history and more than $50,000 in lifetime spend, per Google's Customer Match policy. Lists must also keep at least 100 members updated within 540 days to stay eligible, and the Google Ads API documentation notes that from April 1, 2026, Customer Match upload requests fail if your developer token has never previously sent them.

Here is the correction that saves accounts real money. You will read repeatedly that you should apply a +30% bid adjustment for past purchasers or +15% for site visitors. Under Smart Bidding, those manual audience bid adjustments do not apply. Maximize Conversions, Target CPA, and Target ROAS ignore manual audience modifiers and instead use the audience as a live signal inside the model. Explicit percentage bid adjustments only take effect under Manual CPC or Enhanced CPC. So the correct move in a Smart Bidding campaign is to attach high-value audiences in Observation and let the algorithm weight them, not to set a literal +30% that will never fire. The one place to still exclude rather than observe is new-customer acquisition: in Performance Max you can now exclude existing customer lists so budget chases new logos instead of re-buying people who already converted. Feed the model good audiences, and it will out-bid your manual guesses every time.

Negative keywords and the weekly Search Terms routine

Negative keywords are where profitable accounts are separated from leaky ones. Every campaign, especially anything running phrase or broad match, accumulates irrelevant queries that quietly drain budget. The fix is a disciplined, repeatable routine, not a one-time list. Build the habit below and hold it for the life of the account.

• Review the Search Terms report at least weekly, and daily for the first two weeks of any new campaign.
• Add 10 to 20 negative keywords per review for queries that are semantically related but commercially irrelevant, for example free, salary, how to, and competitor job postings.
• Maintain a shared account-level negative list for permanent junk so every campaign inherits it.
• Separate exact and phrase negatives so you block precisely without accidentally suppressing valuable queries.
• Watch for close-variant creep on exact match, where Google matches plurals, misspellings, and reorderings you did not intend.
• Log every negative you add in a change document so you can audit false blocks if volume drops.

Manual review does not scale forever, so automate the detection. The Google Ads Script below scans the last 30 days for search terms that spent more than $25 with zero conversions and logs them for review. Schedule it to run daily and it becomes an early-warning system for budget leaks.

// Flag search terms with high spend and zero conversions
function main() {
  var q =
    'SELECT search_term_view.search_term, metrics.cost_micros, metrics.conversions ' +
    'FROM search_term_view ' +
    'WHERE segments.date DURING LAST_30_DAYS ' +
    'AND metrics.conversions = 0 ' +
    'AND metrics.cost_micros > 25000000';
  var rows = AdsApp.report(q).rows();
  while (rows.hasNext()) {
    var r = rows.next();
    Logger.log(r['search_term_view.search_term'] + ' spent $' +
      (r['metrics.cost_micros'] / 1000000).toFixed(2) + ' with 0 conversions');
  }
}

Scripts like this are the entry point to a broader automation layer: budget pacing alerts, anomaly detection, and automated bid-strategy guardrails. If you would rather have those systems built and maintained for you, our AI automation team wires Google Ads into the rest of the growth stack so nothing depends on someone remembering to check a report.

Five common pitfalls and how to fix them

Most account failures are not exotic. They are the same handful of mistakes repeated across thousands of advertisers. Here are the five that do the most damage, each with the specific fix. Screenshot each fix into your onboarding doc so new team members do not relearn them the expensive way.

1. Launching before conversion tracking fires. The campaign spends, reports zero conversions, and Smart Bidding trains on noise. Fix: use Google Tag Assistant to confirm the conversion event fires exactly once on the thank-you page before you enable the campaign, then check that the Conversions column populates within 24 hours.
2. Fragmenting conversions across too many campaigns. Ten campaigns at 5 conversions each will never clear a Smart Bidding threshold. Fix: consolidate into fewer campaigns so each one clears 30 conversions per month, then split only after you have volume to spare.
3. Running broad match on manual bidding. This drains a limited budget on irrelevant queries within days. Fix: either switch the campaign to Smart Bidding or downgrade the keywords to phrase and exact until you have the data to steer broad match safely.
4. Editing budgets or targets during the learning phase. Changing budget or target by more than about 20% resets learning and restarts the 7 to 14 day clock. Fix: make changes in small increments and wait for the phase to complete before the next adjustment.
5. Letting ad copy and assets go stale. Responsive Search Ads decay as offers, seasons, and competitors change, and assets marked Low for Ad Strength drag the whole ad. Fix: refresh RSA assets on a quarterly cadence, remove low-rated assets, and rotate in new offers before performance flattens rather than after.

A sixth pitfall deserves a mention for anyone serving EEA or UK traffic: ignoring Consent Mode v2. Without it, denied-consent sessions vanish from reporting entirely, and your conversion counts collapse in a way that looks like a performance problem but is actually a measurement gap. Deploy Consent Mode v2, enable conversion modeling, and your data recovers. Every pitfall here traces back to the same root cause, which is treating Google Ads as a set-and-forget system when it is a data pipeline that rewards disciplined maintenance.

Troubleshooting: eight problems and their fixes

When a campaign misbehaves, the symptom rarely names the cause. Use the table below as a diagnostic index. Match the symptom in column one to the likely cause, then apply the fix. These ten rows cover the overwhelming majority of support tickets a new advertiser will file in the first quarter of running Google Ads.

Work the table top to bottom when performance drops, because the earlier rows are both more common and cheaper to fix. Eligible (Limited) and missing conversions account for a disproportionate share of my-campaign-is-broken complaints, and both are usually resolved in minutes. The two hardest to diagnose are the bottom rows: a learning phase that never completes almost always means someone is editing the campaign every few days, and an enhanced conversions warning usually traces to a Consent Mode misconfiguration that silently drops user-provided data. When in doubt, isolate one variable, change it, and wait a full learning cycle before concluding anything. Chasing daily fluctuations is how advertisers turn a functioning campaign into a permanently resetting one.

Advanced tactics for scaling past the first month

Once a campaign clears its learning phase and holds a stable CPA, the work shifts from launching to compounding. The tactics below are where experienced teams pull ahead, and every one of them depends on the tracking and structure you built earlier. Do not attempt these until your foundational conversion data is clean, because advanced automation on top of broken measurement just fails faster.

• Offline Conversion Import (OCI). Tie closed-won CRM deals back to the click that generated them using the GCLID, so Smart Bidding optimizes for revenue and pipeline, not raw form fills. This is the single highest-leverage move for B2B, because it stops Google from optimizing toward cheap, low-quality leads.
• Value-based bidding. Pass different conversion values by lead quality, for example a higher value for an SQL than an MQL, so the algorithm chases the leads that actually close.
• Portfolio bid strategies. Group campaigns that share a target under one portfolio strategy so they pool data and hit thresholds faster.
• Drafts and experiments. Test a new bidding strategy or landing page as a controlled experiment against a holdout, rather than changing the live campaign and guessing at the impact.
• Data-driven attribution. Now the default model, it credits every touch in the path rather than only the last click, which reshapes which keywords look profitable.
• Deliberate AI Max adoption. Move Dynamic Search Ads to AI Max on your schedule before the September 2026 forced upgrade, with brand exclusions and text guidelines set from day one.
• First-party audience layering. Feed Customer Match and high-value segments as signals, and exclude existing customers when the goal is new-customer acquisition.

Offline Conversion Import deserves the spotlight for B2B. When Google optimizes toward form submissions, it will happily find you the cheapest possible form fills, which are frequently the lowest-quality leads. Feed it closed revenue instead and the same automation starts hunting for buyers. Wiring your CRM into Google Ads this way is the core of modern lead generation, and it converts a click-buying account into a pipeline-buying one. That single shift often does more for return on ad spend than any bid or budget change.

A complete working project: a $3,000 per month B2B SaaS campaign

Here is the entire method assembled into one worked example you can copy. The scenario: a B2B SaaS company selling project management software, a $3,000 monthly budget, a target of $120 per demo request, and a sales team that needs qualified demos, not newsletter signups. Everything below reflects the steps, thresholds, and structure covered in this guide.

Account structure. Three campaigns, deliberately few, so each one clears Smart Bidding thresholds:

1. Brand-Search-Exact        | $300/mo   | Maximize Conversions
   [company name], [company name login]

2. US-B2B-Search-Core        | $2,000/mo | Max Conversions -> tCPA $120
   Ad group A: PM Software [Exact/Phrase]
   Ad group B: PM Tool [Exact/Phrase]
   Ad group C: Competitor Alternatives [Phrase]

3. US-B2B-Search-Discovery   | $700/mo   | Maximize Conversions
   Broad match + aggressive negatives + daily search-terms review

Tracking. Google tag site-wide, one primary conversion (Demo request, value $120), enhanced conversions on, Consent Mode v2 for any EEA or UK traffic, and Offline Conversion Import scheduled once the CRM confirms which demos become opportunities.

Bidding path. All campaigns start on Maximize Conversions. The core campaign moves to Target CPA $120 only after it logs 30 conversions in a rolling 30-day window. At a $120 target on $2,000, that is roughly 16 demos a month from the core campaign alone, so plan on 6 to 8 weeks before the target strategy stabilizes.

Week-by-week plan:

• Week 1: Launch all three campaigns on Maximize Conversions. Review search terms daily, add 10 to 20 negatives per day, and do not touch budgets.
• Weeks 2 to 4: Let learning complete. Promote winning discovery queries into the core campaign as exact match. Keep the negative routine running.
• Week 5: If the core campaign has cleared 30 conversions, switch it to Target CPA $120. Leave the others on Maximize Conversions.
• Weeks 6 to 8: Turn on Offline Conversion Import so bidding optimizes toward demos that become sales opportunities, not just form fills.
• Ongoing: Quarterly RSA refresh, weekly search-terms review, monthly structure audit.

Expected math. At the 2026 B2B benchmark neighborhood, a $2,000 core budget at a mid-single-dollar CPC buys enough clicks that a 6% to 8% conversion rate yields roughly 15 to 20 demos a month, landing near the $120 target once tracking is clean and negatives are tight. Scale the winning campaign 20% at a time, never all at once, and route the rest of the funnel work, landing pages and CRM handoff, into your SaaS growth program so paid clicks convert into booked revenue. That is a complete, launch-ready Google Ads account, built from the same 13 steps above.

How to get started Monday morning

You do not need a month to start. You need a clean sequence and the discipline to follow it in order. Here is the first-day checklist that turns everything above into motion, and it maps one to one onto the steps in this guide.

• Confirm your Google Ads account is in Expert Mode and GA4 is linked.
• Install the Google tag, create one primary conversion action with a dollar value, and verify it fires with Tag Assistant before anything else.
• Turn on enhanced conversions and, if you serve the EEA or UK, Consent Mode v2.
• Build one Search campaign: Search-only networks, Presence-based location, one or two tightly themed ad groups, and one Excellent-strength Responsive Search Ad.
• Load 50 to 100 seed negatives and attach a shared account-wide negative list.
• Start on Maximize Conversions, attach audiences in Observation, and publish.
• Block your calendar for a daily 15-minute search-terms review for the first two weeks.

Then do the hardest thing in paid search, which is nothing. Let the learning phase run for 7 to 14 days without editing budgets or targets by more than 20%. Add negatives, watch the data, and resist the reflex to optimize a campaign that is still calibrating. Once you have 30 conversions in 30 days, move to Target CPA, layer in Offline Conversion Import, and adopt AI Max deliberately before the September 2026 upgrade lands. Running Google Ads well in 2026 is not about clever bids. It is about accurate measurement, tight structure, aggressive negatives, and the patience to let Google's models do the bidding you used to do by hand. Build the pipeline once, maintain it weekly, and the compounding takes care of itself.

Theo Werner

Paid Media Director

Theo manages Skitrate's paid-acquisition practice — Google Ads, Meta, LinkedIn, Reddit, TikTok and the smaller channels worth testing. Manages $20M+ in active client spend. Spent six years inside Google's Performance Solutions team before going agency-side. Specializes in Performance Max diagnostics, full-funnel attribution and creative testing.

• Manages $20M+ in annual client ad spend across Google, Meta, LinkedIn, Reddit, TikTok
• Ex-Google Performance Solutions specialist (Search and Performance Max)
• Designed Skitrate's full-funnel attribution methodology and creative-iteration framework
• Frequent speaker at HeroConf, SMX and Reddit's Advertising Summit

More posts by Theo Werner →