Ruslan Galba AI x Google Ads
Stanford graduate. For the past decade I’ve been consulting and executing marketing strategies for the boards for DTC businesses like Onnit, Goli, AG1, Mudwtr. AI x Claude Code x Google Ads ◾️Work inquiries - tegra.co ◾️Consultations - app.cal.com/tegra
- 24 часа
- 0
- 7 дней
- –
- 30 дней
- −12
- Реакции
- 0
- Репосты
- –
- Комментарии
- –
Динамика подписчиков
Подписчиков у канала «Ruslan Galba AI x Google Ads»: 1 046. Отток за последние 30 дней: 12 (-1,1 %). Средний охват поста за сутки около 15, вовлечённость 1,4 %.
По дням, последние 30 дней
| Дата | Подписчики | Изменение | Посты | Ср. просмотры поста |
|---|---|---|---|---|
| 12 ноя | 1 046 | 0 | 0 | – |
| 11 ноя | 1 046 | 0 | 1 | 11 |
| 10 ноя | 1 046 | 0 | 1 | 19 |
| 9 ноя | 1 046 | 0 | 1 | 21 |
| 7 ноя | 1 046 | −12 | 1 | 26 |
| 7 окт | 1 058 | – | 0 | – |
Просмотры по возрасту поста
Средние просмотры в каждом возрасте, посты за последние 30 дней.
- 12 ч
- 12
- 24 ч
- 15
- 48 ч
- 18
- 3 дня
- 20
- 7 дней
- 39
Когда публикует
Время по UTC. 18 постов за 90 дней.
Последние посты
-
When a Google account stalls, I split brand and non-brand by week, 13 weeks back, before touching the target. Blended ROAS can stay flat while the mix underneath it changes. Put spend, orders and conversion value for each side in one table, then look for: • Brand spend up, brand revenue flat • Non-brand spend down while total spend holds • New-customer share sliding week after week Test brand before you call those orders free. If non-brand keeps shrinking, the account has not stalled. It has quietly become a brand defense budget.
11 просмотров -
Add the Original conv. value column next to Conv. value before your next budget review. Customer lifecycle goals and conversion value rules can both raise the value Google bids on. Original conv. value shows the number before those adjustments. If the two columns differ, the gap is value you told Google to add. It is a bidding instruction, not money in the bank. Report the original column, reconciled to store orders, to whoever pays the ad bill.
19 просмотров -
A higher ROAS is less reassuring when the campaign barely spends. After a tROAS change, read delivery before deciding whether the ratio improved for a useful reason. A target set too high can limit traffic. Check spend, purchase volume and conversion value over a period that has had time to finish reporting. Keep any stock, price or tracking changes beside the target edit. Then decide whether the campaign is still acquiring enough customers at workable economics. A higher ratio alone does not answer that question.
21 просмотров -
Фото
Смотреть в Telegram
17 просмотров
-
An AI creative system should be able to kill a beautiful ad. One of the variants in a recent proof run passed the image checks. Separate AI readers kept flagging the message. The headline was a customer's complaint. In the new layout, it sounded like the brand was admitting the product didn't work. Fixing it would have meant changing the locked headline or adding an unsupported claim. That version got dropped. The attached pair shows a version we could repair by putting the complaint in the customer's speech bubble. I want that decision inside the process before an ad gets anywhere near a budget. While wiring my partner @andreilunev's creative methods into my system, I followed actual...
26 просмотров -
Your best seller can be your worst Shopping ad. It already sells. Lots of people search for it by name. So PMax pours budget into it, ROAS looks great, and you learn nothing new. Meanwhile the products that need ads to be found get a trickle. Two ways to see it: • Split product-level spend into "already searched by name" vs "found through ads" • Check how much of the hero's revenue comes from branded or product-name queries If most of it does, the hero is a brand campaign in disguise. Give it a ceiling and give the rest of the catalog a chance to learn.
25 просмотров -
One product photo, 3 variations, 4 Demand Gen shapes. A 12-image set to check before you upload. OpenAI shipped Images 2.6, the new image model inside ChatGPT, on Nov 4. Skip the GIFs everyone is posting. What a store needs is the product staying identical from edit to edit. Sunburst, the higher-quality version in OpenAI's API, is tuned for exactly that. The prompts below work in ChatGPT too. The 4 ads attached came out of two client batches we ran through Sunburst this week. Same product and the same label in every scene. Google allows up to 20 images per ad and recommends 3 in each of 4 shapes for placement coverage. This workflow aims for those 12 files, not a full 20-image set. Eve...
25 просмотров -
Фото
Смотреть в Telegram
15 просмотров
-
ROAS drops on a Black Friday morning. Before touching a bid, I run four checks in this order: 1. Is it real? Compare platform conversions to Shopify orders for the same hours. Conversion lag looks like a crash until about lunchtime. 2. Is it price? Pull impression share lost to rank on your top 20 products. Competitors' sale prices hit Shopping first. 3. Is it stock? Look for best sellers that went out of stock in one size or color. The ads keep sending people to the gap. 4. Is it the promo? Check the promotion is approved and the sale price is in the feed, not only on the site. Only after all four do I touch targets. Most "ROAS drops" on BFCM morning are lag.
26 просмотров -
Your weakest Shopping product may just be opening the wrong variant. The ad shows the blue one. The landing page loads the black one, at its own price. The shopper sees a color and a price they didn't click on, and leaves. Smart Bidding only sees fewer conversions, pulls back, and the product gets labelled weak. Merchant Center asks for the advertised variant to be preselected on the landing page. Shopify can link straight to a variant, but a theme can still load the default one. Open your three lowest-converting Shopping products from the ad itself, on your phone. If the page shows a different variant, fix the theme before you touch the bid.
32 просмотров
-
Google Ads AI Agentic System v1.5.0 is live ⚠️ Codebase owners: re-clone to update The repo history was replaced with one clean commit, so git pull will fail. 1. Back up your brands/ folder, .env, .project-config.json and any files you changed. 2. Clone the repo again into a new folder (recommended), then copy those back in. Or reset in place: git fetch origin then git reset --hard origin/main This deletes your local commits, uncommitted edits, and any committed file that isn't in the new release, including brand files. Back up first. Roster: 178 → 129 skills • 72 narrower skills now run as modes of one main skill. Nothing was dropped. • roast surface=copy / image / video / page • demand...
53 просмотров -
Google's own help page now says budget and target changes don't trigger a learning period. Google rewrote its learning doc on Oct 21 and renamed it "About Smart Bidding learning." The visible Learning status now lasts 7 days at most. What triggers it, per the new page: • A new or reactivated bid strategy. • Campaigns, ad groups or keywords added to or removed from the strategy. • A significant change to the conversion goals you optimize for. What doesn't: budgets, targets, creatives, and targeting settings like location, audience or language. Google says those never remove historical data or trigger Learning. The best practices are the useful part. Keep one consistent primary conversi...
46 просмотров -
Some of your Search campaigns became AI Max in September. Black Friday will be their first peak week on the new settings. Google ran the upgrades through September. The warning went out by email on Aug 5, with one way out: switch the old settings off first. It landed in the inbox nobody reads, and the upgrade went through. What got switched on depends on what you had: • Campaign-level broad match → search term matching • Automatically created assets → search term matching plus text customization, and on some accounts final URL expansion In a normal week search term matching is exploration. On Black Friday, with odd queries and peak CPCs, it's exploration at peak prices. Final URL expa...
41 просмотров -
PMax's new asset test locks your asset group until it ends. Google recommends at least 4 to 6 weeks. To finish before Black Friday, a test had to start by Oct 29 at four weeks, or Oct 15 at six. One started this week ends after Black Friday. The toggle is "Save asset group changes as an A/B test," spotted by Joey Bidner and written up by Search Engine Roundtable on Oct 21. It splits traffic inside one asset group, which beats reading a before-and-after chart. Three things on Google's help page, in the order they'll bite: 1. Nothing in the asset group can be edited, added or removed until the test ends. That includes the Black Friday headlines you planned to swap in, and the hero image of...
40 просмотров -
Smart Bidding Exploration is hard to judge in Q4, when demand moves weekly. Set it up so you can tell if it worked. Google's help page says to start with a 10% ROAS tolerance. It also says your updated "effective" ROAS target shows up in the bid strategy report. Exploration won't be effective if the campaign is limited by budget. Google shows an alert when that's the case. Q4 makes the testing problem harder, because demand moves by the week. Google's own suggestion is a campaign experiment with a pre-test and post-test, to separate the setting from seasonal shifts. Before I touched it: • Confirm the campaign isn't limited by budget. If it is, skip it for now. • Write down the current...
40 просмотров -
If you sell on Google, Universal Cart can resurface your product weeks after a shopper saved it. Check what it shows. Google launched Universal Cart in the U.S. on Oct 15, in Search and the Gemini app. It saves products from different stores in one place, tracks price drops and restocks, and shows promo codes from the cart. Google says the cart is built on the Shopping Graph, which it puts at 60+ billion product listings. Checkout can happen with Google Pay in the cart, or on your own site, and the merchant stays the seller of record. That makes a wrong price or availability value more visible than it was in a one-time search. What I'd check this week: • Save your ten best sellers to a...
39 просмотров -
Google Ads has two Black Friday buttons for tROAS campaigns. Both make the bidder push harder. Google hasn't documented what happens when you press both. Seasonality adjustment: • Tells the bidder conversion rate is about to jump • Your ROAS target stays where it is • Smart Bidding already plans for seasonal events, so Google says use it only for "major changes," ideally 1-7 days Promotion mode (beta on Search and PMax, access through your Google rep): • Tells the bidder you'll accept a worse ROAS for a while • Loosens the ROAS target and lets you add daily budget • Runs 3 to 14 days, then reverts on its own The first one is a forecast about your shoppers. The second is a decision...
36 просмотров -
Google is moving broad-match and ACA Search campaigns into AI Max. You don't get a vote on the migration. You do get a vote on the settings. I turned the checks I run into a 40-point audit. Five parts: • AI Max control surface: Final URL Expansion, text customization, locations of interest, brand lists • Auto-apply lockdown: every category that rewrites your account while you sleep • Signal quality: Enhanced Conversions match rate, offline imports • AI Overview visibility • PMax vs AI Max cannibalization It scores 0 to 200. Takes about five minutes if you have the account open. Reply AUDIT and I'll send it.
32 просмотров -
Your weakest Shopping product may just be opening the wrong variant. The ad shows the blue one. The landing page loads the black one, at its own price. The shopper sees a color and a price they didn't click on, and leaves. Smart Bidding only sees fewer conversions, pulls back, and the product gets labelled weak. Merchant Center asks for the advertised variant to be preselected on the landing page. Shopify can link straight to a variant, but a theme can still load the default one. Open your three lowest-converting Shopping products from the ad itself, on your phone. If the page shows a different variant, fix the theme before you touch the bid.
32 просмотров -
An AI creative system should be able to kill a beautiful ad. One of the variants in a recent proof run passed the image checks. Separate AI readers kept flagging the message. The headline was a customer's complaint. In the new layout, it sounded like the brand was admitting the product didn't work. Fixing it would have meant changing the locked headline or adding an unsupported claim. That version got dropped. The attached pair shows a version we could repair by putting the complaint in the customer's speech bubble. I want that decision inside the process before an ad gets anywhere near a budget. While wiring my partner @andreilunev's creative methods into my system, I followed actual...
26 просмотров
Вопросы о канале «Ruslan Galba AI x Google Ads»
Сколько подписчиков у канала «Ruslan Galba AI x Google Ads»?
У канала «Ruslan Galba AI x Google Ads» (@hustle365) 1 046 подписчиков на 12 ноября 2026 г.
Растёт ли канал «Ruslan Galba AI x Google Ads»?
Сейчас нет. Отток за последние 30 дней: 12 (-1,1 %).
Сколько человек видят пост в канале «Ruslan Galba AI x Google Ads»?
Средний пост набирает около 15 просмотров за 24 часа, это примерно 1 % подписчиков (ERR).
О чём канал «Ruslan Galba AI x Google Ads»?
Канал в категории «Блоги», язык постов: английский. Описание канала: «Stanford graduate. For the past decade I’ve been consulting and executing marketing strategies for the boards for DTC businesses like Onnit, Goli, AG1, Mudwtr. AI x Claude Code x Google Ads ◾️Work in...»
Какое место занимает канал «Ruslan Galba AI x Google Ads»?
По числу подписчиков: №67 в категории «Блоги», №7 736 среди каналов на языке «английский», №8 687 в мире.
Публикует ли канал «Ruslan Galba AI x Google Ads» рекламу?
За последние 30 дней ни один из 18 постов не помечен как реклама (#ad, Sponsored, «Реклама»). Рекламу без пометки определить нельзя.
Когда публикует канал «Ruslan Galba AI x Google Ads»?
Пик публикаций: воскресенье, около 15:00 по UTC. Выходит около 4 постов в неделю.
Статистика обновлена 18 минут назад.