We Spent $1,185 Running Google Ads for Our Own Agency. Here's Every Dollar and Every Lead.

$1,185.48. 2,312 impressions. 132 clicks. 7 leads.

Most agencies show you what works. They pick the best case study, crop the dashboard screenshot so the losing keywords are out of frame, and call it a "win."

We're going to do the opposite.

This article is every dollar we've spent running Google Ads for our own agency, Growin, over the last 50 days. Every keyword. Every click. Every conversion. Every mistake. Including the campaign that got zero impressions because nobody in New Zealand was searching for what we thought they were.

If you've ever read an agency case study and wondered what they weren't telling you — this is the article where we tell you.

Why we're publishing this

I'll be honest: I'm slightly nervous about this article going live.

Publishing the real numbers from your own lead-generation campaigns breaks an unwritten rule of the marketing industry. You're supposed to show polished results, not $12 clicks that didn't convert and campaigns you had to pause after four days.

But here's the thing: when I ask prospective clients what they hate about dealing with agencies, the answer is always the same — they don't feel like they're being told the truth. Monthly reports full of vanity metrics. Jargon that obscures what actually happened. "Impressions up 240%" when conversions are flat.

So we're going the other way. If Growin is going to be the agency that tells clients the full story — the good numbers, the bad numbers, and the mistakes we made along the way — we have to be willing to do that about ourselves first.

This is that article.

If you want the broader context on what lead generation actually costs in NZ, we've also written the full pillar at /insights/lead-generation-nz-complete-guide-2026 and a costs-specific breakdown at /insights/lead-generation-cost-nz-2026. This article is the receipts.

The setup: what we built

In early February 2026 I decided to stop recommending Google Ads to clients without running it on our own domain first. We had the $600 NZD Google Ads new account credit sitting there. Our own lead flow was inbound-referral only. Time to eat our own cooking.

Here's what we built:

• 3 campaigns covering three different funnel stages
• $75/day total budget split across the three
• Maximize Clicks bidding with a $10 CPC cap
• All paused by default, enabled one at a time
• GTM-NR4W3QW7 for tracking, G-V559JNM0FG for GA4, AW-17952872053 for Google Ads conversions
• Landing pages on growin.nz (/switch, /services/digital-marketing-auckland, /services/revenue-generation)
• Conversion actions: assessment_request ($150), contact_form ($100), phone_click, email_click

The three campaigns were built around a hypothesis about our buyer's journey:

Looks clean on paper. In practice, only one of these was anywhere near reality.

The headline numbers (50 days, lifetime)

Here are the actual lifetime stats straight out of the Google Ads account (customer ID 4910534354). No rounding, no smoothing.

$169.35 cost per lead. That's the number I stare at most mornings. It is higher than our $100 target. It is also, as we'll get to, defensible when you do the math on lifetime value — but I'm not going to pretend it was the number I was hoping to see on day 50.

Every lead, every dollar

Seven conversions in 50 days. Here's all of them.

A few things worth calling out:

• The best click cost us $5.54 and converted. The worst converting click cost $59.95. That 10x range is real — not every click in a campaign is priced the same, and the cheap ones are often the best ones.
• There's a 24-day gap between conversion #5 (Mar 2) and #6 (Mar 26). That gap nearly broke my nerve. If you're running your own campaign and you hit a 2-3 week dry spell, it is not necessarily broken.
• The last two conversions cost $18 combined. Once the algorithm started finding its people, efficiency jumped hard. This is why "pause after 2 weeks if it's not working" is usually the wrong call.

What actually worked: Service_Seekers_NZ

Of the three campaigns, one is carrying the entire load.

Service_Seekers_NZ was built around the theory that there are real people in New Zealand typing "lead generation NZ" and similar phrases into Google, looking for someone to do exactly what we do. It turns out there are. Not many — but enough.

Winning keywords:

Search terms that actually converted:

• "lead generation nz"
• "lead generation companies new zealand"

Search terms that got clicks but didn't convert:

• "pay per lead nz" — the intent there is "give me leads for a flat fee," which isn't our model
• A handful of generic "marketing services" queries

Interesting free real estate: We picked up impressions on competitor searches like "blue ocean marketing nz" and "growthpartnersgpx com" — people literally looking for our competitors saw our ad. We didn't bid on these deliberately; they came in via broad match expansion. Free brand impressions on competitor intent are a nice side effect.

Looking at QS 5-6 for our best keywords — that's the story of this campaign. Our landing page has Landing Page Experience: Below Average despite loading fine, because the growin.nz domain is new and has no Chrome User Experience (CrUX) field data. Google's algorithm is penalising us for being new, not slow. We're paying roughly 30-40% more per click than we'd pay with a QS of 7-8. That's the single biggest cost lever we don't yet control.

For context on how Quality Score warps pricing, the same phenomenon is broken down at /insights/google-ads-cost-nz-2026.

What failed: Agency_Switchers_NZ and Problem_Aware_NZ

Now for the uncomfortable part.

Agency_Switchers_NZ: 0 impressions, 0 clicks, 0 everything

This was the campaign I was most excited about. The thesis was beautiful: people frustrated with their current agency will search "switch marketing agency" or "leave google ads agency" or "new marketing agency auckland." We'd show up, offer a free audit, and steal the business.

Here's what actually happened:

Zero impressions. Not low impressions. Zero. Over multiple days with a $33.75/day budget and reasonable bids.

The reason is the single most useful lesson of this whole experiment: people do not search "switch marketing agency." That search intent barely exists in New Zealand.

When an NZ business is unhappy with their current agency, they don't Google "switch marketing agency." They Google "lead generation" or "google ads auckland" or — more often — they ask a friend. My entire campaign was built on a customer journey that doesn't happen on a search engine.

Money lost: $0 (you can't waste budget that never serves). Time lost: four days of watching a dashboard full of dashes.

Problem_Aware_NZ: nearly as bad

Problem_Aware targeted queries like "why am I not getting leads," "marketing not working," "how to get more customers NZ." The theory: catch people at the moment of frustration, before they know what solution they need.

The reality: even when those searches happen, the intent is wildly mixed. Some people want a blog article. Some want to vent on Reddit. A small minority want to hire someone. Two clicks in 50 days, zero conversions. Paused.

This isn't a failure of the keywords specifically — it's a reminder that top-of-funnel search ads rarely work unless you have serious volume and a content engine behind them. For a 3-person agency, the math doesn't work.

The 5 mistakes I made

In the interest of full transparency, here are the things I'd do differently if I were starting this on day one.

Mistake 1: Launching all three campaigns at once

I enabled all three campaigns in the same week. In hindsight this was ego — I wanted to feel like I was "doing it properly" with a full-funnel strategy. What I should have done was enable Service_Seekers_NZ only, give it 2-3 weeks to settle, and then decide whether the other two were worth enabling based on real data.

By launching three at once I muddied the learning signals and burned my mental energy watching three dashboards instead of one.

Mistake 2: Assuming search demand where there was none

The entire Agency_Switchers campaign was an assumption dressed up as a strategy. I never validated that the searches existed before building ads, keywords, and landing pages around them. Five minutes in Google Keyword Planner would have told me the search volume was effectively zero. I skipped that step because I was excited about the idea. Don't skip the keyword research step, even if you think you already know the answer.

Mistake 3: Landing Page Experience is Below Average and I can't fix it fast

Our landing pages load in under 2 seconds on mobile and score well on Lighthouse. But Google's Quality Score uses Chrome User Experience (CrUX) field data, and growin.nz is too new to have enough real-world CrUX traffic for Google to rate us. Google treats "insufficient data" as "Below Average," which pushes Quality Score down and CPC up.

The only fix is time and traffic. We need real users hitting those pages from Chrome over multiple weeks for CrUX to register us. Paying for the clicks while this bakes is part of the price of being a new domain.

Mistake 4: Phone and email click tracking weren't wired up for the first 45 days

We have phone_click and email_click conversion actions set up in Google Ads. They were supposed to fire from GTM when someone tapped a tel: or mailto: link on the landing page. They didn't, because the GTM tags were built but not published to the live container correctly for the first 45 days of the campaign.

How many phone/email conversions we missed during that period: we don't know. Could be zero. Could be two or three. That's the point — if your tracking isn't wired up, you're not just missing data, you're missing the data you need to judge whether to keep going.

Mistake 5: Diagnostic keywords added too late

Around day 45 I added a batch of broad "diagnostic" keywords designed to pull in search term data — things like "marketing agency auckland" and "google ads nz" on broad match with low bids. The goal was to learn what people were actually typing when they were in our orbit. I should have done this on day 1, not day 45. By the time the data started coming in, I'd already made most of the big decisions about the campaign structure.

The 5 lessons for NZ businesses running Google Ads

If you take nothing else from this article, take these:

1. Validate search demand before you build the campaign

The Agency_Switchers lesson is the biggest one. Every campaign starts with a question: is anyone actually searching for this? Spend 20 minutes in Google Keyword Planner or Google Trends before you spend a cent on ads. If the volume is zero, the campaign is zero. No amount of clever copy will fix a non-existent audience.

2. "Lead generation NZ" is a real search. Most of what you think people search for isn't.

New Zealanders search differently to Americans. They use fewer qualifiers. They skip the state/suburb most of the time. They search in plain English, not marketing-speak. "Lead generation NZ" is a phrase real people type. "B2B demand generation services Auckland" is a phrase nobody types. Write keyword lists as if you were explaining your service to your mum, not pitching it to a CMO.

3. Landing Page Experience affects QS even when your page is fast

If you're launching Google Ads on a brand new domain, know this upfront: Google will likely mark your Landing Page Experience as Below Average for the first 1-3 months while CrUX data accumulates. This inflates your CPC. Plan for it. Budget for it. Don't panic when you see it.

4. The new-domain penalty is real

Related but distinct: new domains get a general trust penalty across Google Ads. Higher CPCs, lower QS, fewer impressions on the same bids. This isn't a conspiracy — it's Google being appropriately cautious about advertiser quality. The fix is time, volume, and content. For growin.nz, that means we're paying a "new domain tax" until the site has been actively serving ads and receiving organic traffic for long enough.

5. $169 per lead is fine if the lead is worth enough

This is the lesson that lets me sleep at night. A $169 cost per lead sounds bad next to our $100 target. But here's the actual math:

• Average engagement value for a Growin client: $15,000 - $40,000/year
• Lead-to-customer close rate (realistic): 15-25%
• Expected revenue per lead: $2,250 - $10,000
• Current cost per lead: $169
• Expected ROAS: 13x - 59x

At those numbers, $169 per lead is not the problem. The problem would be if we started getting leads at $50 and couldn't close any of them. High CPL with high close rate beats low CPL with low close rate every time. For a deeper breakdown on this math for NZ businesses, see /insights/lead-generation-cost-nz-2026.

This is also why I'm suspicious of any agency that brags about low cost per lead without telling you the close rate and lifetime value. Cheap leads that don't close are worse than expensive leads that do.

What happens next

Here's the honest plan for the next 60 days:

1. Keep Service_Seekers_NZ running. It works. The data is improving week on week. The last two conversions cost $18 combined. The algorithm is starting to find its people.
2. Leave Agency_Switchers and Problem_Aware paused. Possibly forever. The search demand isn't there, and no amount of optimisation fixes a zero-impression problem.
3. Rebuild the /switch landing page as a top-of-funnel content piece, not an ad destination. If people won't search for "switch agency," we'll get them there from this article and other content instead.
4. Focus QS on Service_Seekers. Tighter ad groups, more specific ad copy, dedicated landing page for "lead generation NZ" specifically.
5. Kill date: if CPL is still above $150 at the 90-day mark (early May), we pause everything and rethink. I'm not going to quietly burn budget for another six months to save face. That's the whole point of publishing this article — when we said "transparent," we meant it.
6. NZ$600 Google credit still active. That's reducing our real out-of-pocket spend significantly through the test period. When the credit runs out, the economics get reevaluated from scratch.

If you want to see a more polished writeup of how we help other NZ businesses do this properly, our revenue generation service page is the best starting point.

Frequently Asked Questions

Why are you publishing this?

Because we're betting our entire positioning on being the agency that tells clients the truth about performance, and we can't ask clients to trust that if we're hiding our own numbers. Also: nobody else in the NZ market is doing it, and "nobody else is doing it" is usually where the most valuable content lives.

Is $169 per lead good or bad in NZ?

It depends entirely on what a customer is worth to you. For a $50 product, $169 per lead is catastrophic. For a professional services engagement worth $15,000+, it's excellent. The right question isn't "is my CPL good?" — it's "is my CPL less than my lifetime customer value divided by my close rate?" If yes, keep spending. If no, stop. For context on what other NZ industries pay per lead, see /insights/google-ads-cost-nz-2026 and /insights/lead-generation-cost-nz-2026.

Why did Agency_Switchers get zero impressions?

Because nobody in New Zealand searches "switch marketing agency" or the related phrases I built the campaign around. The search volume is effectively zero. Not low — zero. Google can't serve an ad against a search that nobody runs. This was entirely on me for not validating the keyword volume before launching.

What would you do differently if you started today?

Launch only Service_Seekers_NZ. Start with a $30/day budget. Tighter ad groups — one per core keyword theme. Build a dedicated landing page for "lead generation nz" rather than sending traffic to the main services page. Get phone/email click tracking working before day 1, not day 45. And — most importantly — check Google Keyword Planner for actual search volumes before designing the campaign structure, not after.

Can I see the actual Google Ads account?

I can't open up full account access publicly, but if you're a prospective client I'm happy to share a read-only walkthrough on a discovery call. The numbers in this article come directly from customer ID 4910534354 in our Google Ads Manager account and are accurate as of the publish date.

Are you going to update this article with new numbers?

Yes. I'll come back at the 90-day and 180-day marks and add the new data — including whether we killed the remaining campaign, pivoted, or scaled it up. The "kill date" isn't marketing language; it's a real date on my calendar.

What's the single most useful thing in this article?

Probably mistake #2. Assuming search demand where there isn't any is the most expensive, most common, and most invisible mistake in Google Ads. If this article saves you from designing a campaign around a keyword that nobody actually types, it's paid for itself.

The closing thought

I don't have a clever pitch for you at the end of this.

The point of this article isn't to sell you on hiring us. The point is to show you what transparency looks like when an agency actually means it — the real dashboard, the real mistakes, the real math, and the real plan for what comes next.

If that's the kind of honesty you want from whoever is running your lead generation, you know where to find us: /services/revenue-generation.

And if you just wanted to see what $1,185 of real Google Ads spend looks like from the inside — now you have. Use it. Share it with your team. Show it to the agency currently invoicing you and ask them for the same.

That's the whole article.