Taikun Digital

We help companies make money with Google and Meta ads!

[Taikun Digital @TaikunDigital]

19% of all Meta ad impressions are served to users who have already seen the same creative five or more times in 30 days. 

Most accounts have no system for catching this before it quietly drains budget.

We wrote the audit framework we use to find it: → https://t.co/qbp6g48TcB

[Collin Slattery @CJSlattery]

On paper, $1M in Google spend should return $5.6M. We dug into the numbers across $20M in spend and found that many brands are only getting $4.75M. The missing $850,000 has a name:

Hidden search terms.

Up until September 2020, Google showed you 90-95% of the search terms your ads paid for.

If a term was bad, you marked it as negative and stopped paying for it. Then they hid most of them for "privacy reasons."

Now, a big chunk of what you pay for gets lumped into a single line called "other search terms."

You can't see the individual queries, just one aggregated row, and that performance is reliably worse than what's still visible. It works in Google's favor to hide the stuff you'd otherwise exclude.

We looked at $20M in search and shopping spend to size it:

For every dollar you run through Google, they skim almost a full dollar of efficiency off the top. And this problem is only going to get worse as Google hides more data with more automated campaign types.

You can't unhide it, but you can claw back efficiency with n-gram analysis. Hidden terms are usually long-tail queries that Google tags as low-volume.

You want to aggregate the visible data, find the terms that consistently underperform, and exclude them. You'll usually pull out some hidden queries with the same term, too.

It’s not a complete fix to the problem, but it is critical for improving efficiency on Google.

[Taikun Digital @TaikunDigital]

A standard Google Ads audit tells you the account is healthy.

It does not tell you whether doubling your budget will double revenue or destroy efficiency.

We wrote the guide to the audit that actually answers that question: https://t.co/WwSw24PQBg

[Collin Slattery @CJSlattery]

Creative is the stupidest use case for AI right now.

Video editors aren't that expensive, and neither are graphic designers. Hell UGC is just getting cheaper and cheaper.

So, why are you spending resources trying to solve very affordable problems?

As @DigitalSamIAm says: "They just look at the end invoice and they're like, we spent $75k on creators and editors. AI could do it for free. And then you start adding up all the monthly subscriptions and the compute and the credits and the time.

“And all of a sudden you spent $85,000 to save $75,000."

Plus, you still need people to use the AI, build the systems, maintain the systems, and fix the outputs that aren't quite right

[Taikun Digital @TaikunDigital]

YouTube is the most-watched streaming platform on US TV screens, even bigger than Netflix.

Most DTC brands are still not advertising there. Full-screen, non-skippable, living room attention is sitting largely unclaimed.

We wrote the honest guide to YouTube CTV: formats, targeting, costs, and how to measure actual lift: https://t.co/mMDN6mZs2S

[Phil Kiel @PhilKiel]

Long time listener, first time caller on @andrewjfaris's podcast this week.

Like many I've been listening since the 4x400 days.

Spilled the secrets on how we use CPMr as an indicator (not an optimizer) and why sometimes the issue is in the business... not the ad account.

https://t.co/Cy0RM0jVZw

[Phil Kiel @PhilKiel]

How to remove Audience Network from your ad account without editing each ad set individually 👇

[Taikun Digital @TaikunDigital]

Most Google Ads accounts are producing more creative and getting worse results.

The system doesn't need volume.
It needs contrast.

Scaling accounts test distinct angles: price, quality, speed, status, outcome.

We wrote about the 2026 Google Ads Creative Guidelines: https://t.co/Pg8l6gcCfh

[Taikun Digital @TaikunDigital]

Here's how Google Ads is taking credit for conversions it didn't cause:

For most mature brands, up to 90% of branded search conversions are non-incremental. Lifts typically fall between 10% and 50%, meaning 67 to 91% of attributed conversions would have happened anyway.

We wrote up how to actually measure what your ads are doing → https://t.co/SdBvoQrAAv

[Taikun Digital @TaikunDigital]

We restructured Jones Road Beauty's Shopping campaigns from individual SKUs to product categories and saw a 40% revenue increase at the same spend. 

When data pools, the algorithm learns faster. Fragmented SKU groups starve it. Category groups feed it.

Full audit breakdown →  https://t.co/hP5XKpnoCt

[Collin Slattery @CJSlattery]

What got you where you are today is not going to get you where you want to go.

This a good life lesson more broadly, but if you're a brand in the 20-40mm revenue range, it's super applicable to your ad stack and your measurement.

When you're predominantly purchase-optimized Meta => DTC, it's simple. You know where your customers are coming from and the Meta data is reliable. Understanding what works is easy.

That's not going to get you to $100mm. Your ad stack is going to get more complex. Meta's share of ad dollars will go down.

Google, YouTube, TikTok, Podcasts, TV. A larger share of your ad dollars will be on other channels. You're going to be selling on Amazon and moving into retail. If you keep relying on the deterministic measurement that got you where you are today, you're very likely to plateau.

You have to be able to evaluate the contributions of each channel on their own and that means looking at new data in new ways. Running holdout/inverse holdout tests. Leaning into probabilistic measurement.

We've been through this journey with clients many times. It requires a fundamental shift in how you evaluate and think about performance. It's hard but it's good.

[Collin Slattery @CJSlattery]

More creative, more creative, more creative.

That's what every brand is being told right now.

But anyone who has tried to ramp up creative output knows: it's not just creative strategy that gets challenged. It's the entire production line. People management. Quality control. Delivery timelines. Everything.

We hit that wall last year.

So we built a custom creative studio inside our project management software. Now we track the status of each brief (which produces 6-10 assets), briefs per client, briefs per designer, completions by date, time in status, and more.

Aside from compliments on how well our assets perform, the most frequent compliment from paid social and creative clients? How quick and reliable our delivery is.

And this system isn't just for production. The creative strategy side is just as organized to ensure every brief that hits production has a strong thesis behind it.

This is how we produce incredible ads reliably at scale.

[Collin Slattery @CJSlattery]

Want to wish @PhilKiel a very happy 1 year anniversary here at Taikun.

I can't even believe that he agreed to join us to run the paid social department. And to say it has been transformative has been an understatement.

The company looks completely different than it did a year ago. Phil is even better at what he does than it looks. Phil is one of the rare people who's bite is bigger than his bark.

So excited for year 2 of building together. And Phil is absolutely cooking up some massive stuff for year 2. Y'all have just gotten a small taste of it.

[Taikun Digital @TaikunDigital]

When a brand hires an agency, they instantly tap into what’s basically a million-dollar payroll working on their account.

Not full-time, of course but that's actually a good thing.

[Taikun Digital @TaikunDigital]

Making your Cyber Monday offer better than your Black Friday offer doesn't make you cute. It just makes you an asshole.

Nothing erodes trust faster than a bait-and-switch.

Customers definitely notice.

[Taikun Digital @TaikunDigital]

https://t.co/zymP1aCOd0

[Taikun Digital @TaikunDigital]

Most advertisers suspect hidden search terms waste money. We wanted to be certain.

So we analyzed $20 million in ad spend across dozens of accounts.

The results were staggering:
👉 For every $1 you spend, Google siphons off $0.85 in inefficiency purely through hidden search terms.

This isn’t about user privacy. It’s about Google selling low-quality inventory without scrutiny.

We also found that hidden search terms had:
📈 CPCs that were 52% higher
📉 CTRs that were 44% lower

It took months to compile and clean this data. If you're spending on Google Ads, you’ll want to see the full breakdown.

We’ve laid it all out in the blog post which is linked in the first comment.