SKAG Google advertising structure that will bring millions

Hello! My name is Yana Lyashenko, and I’m a Google logistics specialist. My job is to deliver traffic with the right parameters to businesses.

Today we’ll talk about things that are often overlooked in both courses and conferences. Even Google itself prefers to stay silent about this. A simple restructuring of your advertising account can double sales. Your campaign organization affects both costs and profit.

Let’s start with the basics. Structure is the way of grouping advertising campaigns and keywords within them. After six years of working with advertising and optimizing hundreds of campaigns, I’ve come to the conclusion: the best grouping option is SKAG (one ad group — one keyword).

Yes, I understand that some specialists will start protesting and snorting right now. But if you approach advertising systematically or work with foreign markets — I haven’t yet encountered a more organized structure than SKAG.

Is the “1 group — 1 keyword” approach still relevant?

Let’s figure this out step by step. There’s a basic rule written by Google itself: the more precisely your ad matches the user’s search, the better the result. Sounds elementary, right?

Now imagine this. As a specialist, you’ve seen hundreds of different search queries that people enter for a single product in a month (assuming the budget isn’t $20 per week, of course). And then there’s voice search. One-time queries.

Research shows that about half of queries are already voice-generated. Which means most have no search volume. We can’t add them to advertising as separate keywords — but they are often conversional. And what does Google do? It lowers relevance. This isn’t an exaggeration — this is reality. A vicious circle, in short.

So what do you think of that, dear PPC specialist?

Suppose you’re thinking right now: “Yana, are you serious? Making a ‘1 group — 1 keyword’ structure is hellish amount of work. Why torture yourself?”

Don’t close the page just yet. In this article, I’ll tell you what advertising account structure will be relevant and what SKAG has to do with it.

Let’s clarify one important point. We optimize campaign structure not for “magic.” Forget such formulations — there’s only systematic work, period.

The structure doesn’t need to be maximally complex. The goal is to find a balance between campaign detail and time for their optimization. We need a structure where we can both have coffee before work and save the client a decent amount so they’re delighted with the cooperation.

In this article — unlike many specialists who can’t clearly explain advertising campaign structuring — I’ll lay it all out: what SKAG is, whether the approach is relevant now, its pros and cons.

What is SKAG: 1 group — 1 keyword?

The essence is simple. In an ad group, you place exactly one keyword — regardless of match type. The ad text itself is written with maximum detail: the keyword must be included in the headline, description, and paths.

Important! The main danger of such detail is that it can be heavy to manage. And it’s not about the number of keywords or ads. The problem is with changes to Google’s match types.

Let’s look at an example.

Here’s the situation:

• Search query: interior design;
• Keyword: [interior design];
• Campaign name: Design_Interior_General;

Now compare:

• Search query: interior designer;
• Keyword: [interior design];
• Campaign name: Design_Interior_General;

Or:

• Search query: apartment design;
• Keyword: [interior design];
• Campaign name: Design_Interior_General;

See the catch? Exact match is set. After updates to match types, Google itself decides what to consider a synonym or close variant.

Okay, you’ll say, so “design” and “designer” got mixed up. At large semantic volumes, this moment needs to be controlled. “Designer” and “design” can bring completely different conversion. And if there’s high competition on keywords — you’ll start overpaying.

Suppose the minimum bid is $0.8, but in reality you pay $2–3 because the system “thinks” the ad doesn’t have the word “designer.” Or the word “apartment” is missing. Or maybe you don’t work with apartments at all, but with houses. Google doesn’t care.

Now imagine that this kind of mess occurs throughout the entire account. Semantics can amount to thousands of keywords. Perhaps as a specialist, you manage five to ten projects. Would you want to constantly clean up after the system?

Everything described above is indeed a problem. But SKAG structure still has plenty of advantages. Though, the pros we’ll discuss will be appreciated more by order fanatics and pedants. Read carefully 🙂

Advantages of SKAG

This structure has three key pluses. Let’s examine each point in detail.

Maximum possible CTR

If you do SKAG properly, the keyword is written in the ad with maximum detail — in headlines, descriptions, paths. Thanks to this detail, your ad looks much more relevant compared to groups where keywords are dumped in a pile. Hard to argue with me here.

Suppose you type “graphic design courses” into Google. Of four advertisers in the results, the first and fourth bothered to write the key phrase in headlines and descriptions. They pay less per click than others. Moreover, the first advertiser used the key phrase the maximum number of times. While the second will pay more — they sell software, not courses. Their ad doesn’t match the query at all.

Flexible bid management

Let’s be honest: keywords with minimal competition are rare. Take graphic design. If you check semantics in the keyword planner, it becomes obvious — competition is high.

And this means one thing: we need maximum control over bids. Otherwise, budget drain is guaranteed.

If you set bids correctly — based on site conversion, product margin, and cost per click — you get truly profitable traffic. By the way, we have a separate video on YouTube about how to calculate the maximum cost per click.

In certain niches, cost per click reaches $10 and above. Imagine what happens without bid control at such volumes.

Easy to learn yourself and teach the system

The “one keyword per group” structure is maximally simple and effective. It’s not difficult to create, and if necessary, quite realistic to quickly scale to thousands of ads.

Another bonus: if a group gathers enough statistics, you can connect automatic bidding strategies — and be confident they’ll work to the maximum. This data is clean, without “garbage” from mixed keywords.

Disadvantages of single-keyword ad groups

It would be dishonest to talk only about pluses. SKAG has weak sides too, which are worth knowing in advance.

Manual campaign creation — long and tedious

If you’re adding not 20 keywords to an account, but thousands — for many this is simply unmanageable volume. And it’s not about campaign grouping. Each group needs at least three ads. Multiply this by hundreds of groups — and the scale of the task becomes obvious.

Tip for pedants: if you want to work with SKAG structure — get yourself an Excel file and a couple of macros. Or, at worst, delegate the routine part to a freelancer and breathe easy 🙂

Labor-intensive maintenance

Tracking all search queries, analyzing ad effectiveness, adjusting CTR and texts — this is indeed not easy.

But honestly, the problem here isn’t the structure itself. Not all specialists know how to manage large semantic volumes. In my time, I “survived” the fashion for million-keyword accounts — in both Yandex Direct and Google Ads. For me, these structures are maximally organized and fast to manage. But for some, a similar data array will seem like a haystack where it’s easy to get lost.

Slow experiment execution

Suppose you launch an A/B test of new ad variants. For a keyword with 20 impressions per month, you might wait half a year for results. Not the most inspiring prospect.

Here’s what I’ll say to pedants: don’t get fixated on A/B test speed. Your task is to get traffic at the most favorable price, not to test for testing’s sake. If the keyword is written in the ad in detail, then a prefix with a promotion or price can well wait until enough clicks accumulate.

Reduced optimization score

We know that the optimization score (and its drive toward 100%) affects the overall ad rating in the account. This can’t be ignored. Google directly warns: keywords marked “Low search volume” are associated with minimal search traffic. The system considers them irrelevant for most users and temporarily makes them inactive — your ads for them will simply be turned off.

So Google intentionally limits impressions on keywords regardless of your wishes. It does this to save its own system resources on moderation and calculations. And we have to reckon with this.

Will single-keyword ad groups be relevant in the future?

The SKAG structure was super popular for a long time — and deservedly so. It worked great. I still love it hotly and consider it one of the best ways to organize a Google Ads account. But. In recent years, Google has made so many changes (and even more are coming) that we have to adapt.

Back in 2014, Google updated the rules on close variants. Remember the example with “designer” and “design”? Well, that was just the beginning.

In modern competition conditions, it’s critically important to control every penny of spending and not drain traffic on what Google considers “close” but in fact is far from real business processes.

In 2017, changes were made to exact match type — added functionality for reordering keywords and the ability to substitute new formulations. Yes, “apartments” and “interior” are now close variants in the system’s opinion.

In 2019, synonyms and intents were merged into phrase and modified broad match. In short, Google wants to increase its profit. Don’t think all these changes were made for advertiser convenience. Google does this for itself, to earn a couple more billion 🙂

It would seem that the problem with mixed queries and synonyms could be solved with cross-negative keywords. But I encountered another trouble — cross-negative keywords no longer work as before. Often they limit clicks and impressions by default. Especially for queries with about 50 impressions per month — there the system cuts reach too strongly. Plus Google already reduces impressions due to low keyword frequency (remember, I wrote about this above).

And another headache — budget distribution. Google is much more willing to give budget and impressions to high-frequency keywords. Low-frequency ones remain “on starvation rations.”

In light of all this, I’ve come to the conclusion: keywords need to be structured differently.

My solution: Intent Keyword Ad Group (IKAG)

I haven’t come up with a simpler name yet, but the essence is this — it’s grouping by user intents. Yes, this is what I talk about in every one of my videos: keywords need to be adjusted to the sales funnel.

The main task is not to squeeze maximum clicks, but to focus on what the client is really looking for. So we’ll structure all semantics by user intent. And for this, we’ll have to strain our brains a bit.

Why specifically “by intent”? Because semantics with low competition have practically disappeared in the world. And Google itself limits us in this. So it’s better to figure out how our client thinks, adjust advertising to the right moment, and catch them — like a fisherman catches ocean tuna. Here you need experience, dexterity, and attention. This isn’t going after bullheads with a net 🙂

All advertising should be built around customer avatars. So we compose target audience portraits, analyze them — and build structure based on this data.

Sales funnel in Google Ads

To avoid speaking abstractly, let’s break down the intent funnel with specific examples.

Take the query “buy shrimp.” Seems simple. But shrimp come in different types: frozen, boiled, fried, dried, tiger — and this is far from a complete list. By such a general query, we don’t understand the final intent: what specific product the person is looking for and how much they’re willing to pay.

If we sell tiger shrimp, our keywords will look different: “buy tiger shrimp with delivery,” “tiger shrimp delivery,” and so on. Specifics are key.

Now let’s break down a more complex example — selling tuna fillet. Suppose our client is a man aged 25–45 who likes to cook at home, relax outside the city, has a barbecue, and wouldn’t mind surprising friends with something unusual. What does his journey look like?

Awareness stage

Queries here are usually high-frequency and maximally general. For example, simply “tuna.” The person isn’t even thinking about buying yet — they’re just getting acquainted with the topic.

Interest stage

Queries appear related to studying: “benefits of tuna,” cooking recipes.

I understand, by frequency this isn’t an iPhone — not everyone buys tuna :). But for our niche, these are working queries.

Study stage

The person has already looked at recipes and realized that choosing fillet requires more attention. There are varieties of tuna, and they differ in taste. Queries like “tuna for sushi” appear — and this is already a signal that the client is looking for a quality product.

Comparison stage

Here the person starts looking for where to buy and comparing options. Phrases with price are just right. At this stage, they study cost, delivery conditions, assortment from different sellers.

Final chord

Query “tuna fillet with delivery.” The person is ready to buy.

This is what the user’s journey through intents looks like. The main conclusion: focus on intent, not on the formal keyword type — whether it’s cold, warm, high-frequency, or low-frequency.

If a person types “trends in design” — they’ve just started their journey in studying the topic. But if “design courses price” — they’re already comparing options and thinking about who to buy from. The difference is huge.

How to structure keywords by intent?

When you’ve collected semantics, it needs to be grouped. And here the logic of intents helps again. Suppose you have several types of queries within one product or service:

• Queries with price/cost;
• Queries with geolocation;
• Queries with “buy”/”order”;
• Queries “brand + model”;
• Queries with “inexpensive”/”cheap”/”affordable.”

Essentially, this is your grouping. By selecting keywords maximally close to the user’s “wallet,” you increase the probability of conversion to sale. Collecting general traffic — even if targeted — is too costly in both time and budget.

Similarity between keywords works in two directions. On one hand, it allows combining similar keywords into one group (for example, all price queries). On the other — it provides the opportunity to write a maximally relevant ad that responds to user intent. And this, as a consequence, means a sale.

The principle works equally for products, services, B2C, and B2B.

Want to explore the topic deeper? Write — and I’ll do it!