Read the whole Google Shopping guide:
In the first part we reviewed 7 winning strategies for Google Shopping. Now let’s analyze each strategy in detail:
A small reminder! When we talk about strategies, we contemplate not only structuring strategies but also a few extra techniques that allow you to make more profit from Google Shopping campaigns.
Which means that you can combine a few strategies with one another. Of course, not all techniques make a perfect match, so here’s the cheat sheet to guide you:
*Strategies that involve splitting to several campaigns don’t combine with each other
Shall we look at our 7 shades of strategies a little bit closer?
The name of the strategy “All goods in one group with bidding per unit” speaks for itself.
Who is it suitable for?
Important! The most suitable option is stores with a range up to 10-15 products. Or if there is a lot of time to manage the campaign, even up to 50 products. But if there are more goods, optimization can turn into a nightmare: it will be difficult to keep track of it, compare and analyze at least something.
For example, there is Carrie — the owner of a small online store of backpacks. Carrie decided to set up a Shopping campaign ‘cause she heard that it had worked for a competitor, although she personally had no experience in setting up. Carrie started with one best-selling model in different colors. She created a campaign (and this is nothing difficult at first sight), launched and enjoyed it. Now she has 1 campaign with 1 group and 1 product group (“All products”). Carrie gets some statistics, that’s just things are going so-so. Most likely, Carrie will soon say “this your Shopping is garbage”. Or maybe she will read about the strategy “All products in one group with bidding per unit” and everything will be fine.
So, the good news is that the statistics that you have already received will come useful.
The main idea of the strategy is that products can not work the same way. One color may bring traffic to the website well but not convert, while the other may simply not be enough budgeted. Therefore, it is important to set up the bids for each color depending on their effectiveness. This strategy allows you to do so. As soon as the conditional Carrie introduces it, she will immediately notice a positive result.
Such strategy is suitable for stores of tourist backpacks, gift socks, colored lenses for eyes. It is suitable for an online store of sweaters, which plans to advertise only reindeer sweaters in Google Shopping. It is suitable wherever the range is not in the hundreds. If you have a situation similar to one of these options, feel free to try.
What’s the trick of this strategy? It significantly improves conversion rates and reduces conversion costs. So the profit becomes higher. Very nice, isn’t it?
To set up you will need to set bids at the product-level and for this you will have to change the campaign structure. How to do it:
And you can see the statistics for each product separately!
Now, based on historical data, you can set individual bids (if necessary, exclude some of the products), but across from “Everything Else” instead of the bid choose “exclude” (is required).
Pay attention to the changes in the main metrics: conversion rate and conversion cost. Also, monitor the level of traffic and the number of clicks: it should not decrease significantly. After such bids updating leave it for some time, while products are receiving new statistics for further optimization.
Here are the main mistakes which you can make in this strategy and what you should avoid:
SPAG is a KlientBoost finding, Single Product Ad Groups. This is an equivalent of SKAG, only for Google Shopping. SPAG helps to control campaigns and achieve better advertising results.
Hearing about the concept of “one product — one group”, most often we hear objections like “Oh well, it will be very difficult to manage”.
So, there may be some truth in this but our main answer is: “To each his own”.
If you sell 1-100 products, SPAG will help clarify how different products show themselves in the conditions of Google Shopping campaign. Campaign structure will take time and effort but one specialist will be able to carry it out.
If you sell 100+ products, using SPAG for all products becomes impractical. Firstly, the account itself will be very “bloated”. Secondly, to bring the account to the desired form according to SPAG principle will require a lot of energy and one specialist – it will be more difficult to cope here. Thirdly, to manage such a mass is almost impossible and it will be like the nightmare for the PPC-specialist.
Your task is not to ensure equal traffic for each product. Your task is to isolate bestsellers and use them as a traction force of sales, put them ahead of sales.
But yeah, even with a large range you can still apply SPAG.
If you manage thousands of products and work with Google Shopping, it’s important for you to identify 2 categories of products:
It often happens that one top product with a good payback drags on itself 5 “non-profitable” goods, which advertising is more expensive than the profit received. SPAG helps to avoid such situations by allocating separate groups with its own budget for bestsellers. Thus, most of your products remain in the campaigns with the old structure, which you got the hand to manage. And 5/10/20 of the best-selling and profitable products are managed on the principle of SPAG.
Accordingly, there are several main objectives that the SPAG strategy allows to achieve:
Step 1. Open Google Ads → Campaigns → “+” button → Select the “Sales” goal.
Step 2. Create a Shopping campaign.
Step 3. Select the Merchant Center account with products to advertise and the country where products are sold.
Step 4. This is where the settings for the first ad group begin. By default, Ads offers to start with a single ad group that contains all products — and yes, we use exactly this option. No, this doesn’t contradict the SPAG principle. We just take the first step in the standard scheme and then we will stop following the Ads recommendations and go our own way. So we create one common group and move on.
Step 5. After creating the group, you will enter the interface and see the “All Products” ad group. Now we need to turn it into SPAG. To do this, click on the “+” button next to the group name.
Step 6. Subdivide “All products”.
You can split a group according to several parameters: category, brand, identifier, condition, type of product…
To create a SPAG, you need to divide by item ID. Item ID is a unique identifier that each product has. It never repeats, so if you divide the group by Item ID, you’ll have a single-product group.
Step 7. After clicking on the Item ID you will see a list of available products. Click one of them, save — and everything is almost ready.
You will see a new line for the product you have selected along with “Everything else”. Be careful here: the item “Everything Else” contains other products. If you don’t exclude it, you still get all the products in one group — we don’t need it.
Accordingly, we set bidding and exclude “Everything else” through the “Excluded” button.
That’s it, your first SPAG is created!
This structure is easier to organize using Excel and AdWords Editor but if you have carried out an advertising campaign with all the products in one ad group several times, the easiest way to start the transition is to create a SPAG for the best-selling products and more precisely adjust the bids for them.
What you need to monitor after setting:
The most common mistakes:
Query-level bidding structure is that tricky scheme of splitting all advertising into 3 campaigns (general, brand, product) so that each gets its own priority and bids.
Such a strategy won’t work for a business with a low budget — it should be enough for at least 20-30 clicks from campaign per day, taking into account the cost of a click in your market segment. For the strategy you need to set up 3 campaigns.
Query-level bidding structure helps to achieve, perhaps, one of the main goals of the PPC specialists work — a rational allocation of resources and getting a return on investment. This strategy allows you to avoid situations where you put maximum resources into a campaign and then you realize that it gives you a minimal result: little traffic, few conversions, few sales. It works on the principle of “the potentially more profitable a query is, the more money you can invest in it” by setting different bids for general and narrow queries. And this is already a huge plus in itself.
To start, create one Shopping campaign with the structure you need. For the next step, the structure should be completely ready for launching a campaign, so after that we will copy the campaign and make 2 others based on it, making a few changes.
To structure a campaign, you can use any principle: either PAG or any other. This is not significant in bidding-based query-level structure.
Further setting up algorithm:
Step 1. As soon as the first campaign is ready, go to the Settings → Shopping settings tab and set the “High” priority to the campaign. Since this is your general campaign, you should also set the lowest bids for these product groups.
Step 2. Duplicate the campaign so in the end you have to set up 3 identical campaigns.
Step 3. Update campaign names: you should easily distinguish between “General”, “Brand” and “Product” campaign.
Step 4. Set the medium priority for the brand campaign and low — for the product campaign.
Step 5. Correct the bids in each campaign: in the general they should be the lowest, in the brand — the average, in the product — the highest.
The actual bids, of course, customize to reflect the real products that you sell and trends in your market segment.
Tip! We recommend bidding in that way that they differ from campaign to campaign by at least 50%. Make sure to keep bid changes consistent between items. Otherwise, the ads you planned could not be showed and the efficiency of campaign would be worse.
Step 6. Set up general negative keyword lists so these campaigns can properly divide traffic between them.
To configure negative keyword lists, go to the Shared Library and create three lists:
Important! There may be confusion when naming lists, since, for example, the list “brand” is applied to the campaign “general”, and the list “product” is applied to the campaigns “brand” and “general”. In order not to complicate your life, you can take it as a rule to put its purpose in the name of the list. That is, the first item from the list above ↑ will be “All campaigns”, the second “General campaign”, the third “General + Branded campaign”.
Step 7. Adjust the budget.
This is probably one of the easiest steps to set up. Simply go to the budget page, assign a budget name, apply it to your Shopping campaigns, set the amount and save it. Bingo!
What is important to remember:
As you understand, you need to monitor certain rules in order everything to work as needed. Some insidious dirty tricks are not expected for you but sometimes we can see such mistakes due to inattention:
You’ve probably heard about the Alpha/Beta campaigns but for search advertising. The main idea is to split queries (keywords) into 2 campaigns: alpha — converting (higher probability of purchase), beta — all the rest. In Shopping, you can also do this, despite the fact that you can’t add keywords. You can combine this strategy with other ones (for example, #1 and #2).
The Alpha/Beta strategy of campaigns will suit those who already have some data: when you know which queries bring more conversions and which only bring traffic. Unlimited budget is also important for it. It is suitable for small and medium stores with an assortment of up to 200 products.
For the Google Shopping Alpha/Beta strategy, it doesn’t matter which campaign structure it has, so the products can be both similar and completely different:
Alpha/Beta-strategy will help to split up the bids: higher for conversion queries, lower for the rest. Accordingly, Alpha campaign ads will be in a better position and are more likely to receive a click, so it should lead to increase in conversion.
Alpha/Beta campaigns in Shopping split traffic into converting and general. From the Beta campaign, we minus the search queries in the exact match that usually brought in conversions with a good cost per conversion.
Next, we can periodically review the search queries in the Alpha campaign and add to negative keywords those that don’t correspond to the list of converting phrases. Or install a special script for Google Ads, which will save our time and perform the task automatically.
The most common mistakes in Alpha/Beta strategies for Google Shopping:
Structure by device is quite a specific thing, which is useful if you really have a noticeable difference in the results of the campaign work on desktops, tablets, and mobiles.
The device strategy has no restrictions of using — it is suitable both for single-brand stores for 10 products and for huge retailers with a range of thousands of products. True, there are some subtleties: it is not recommended to combine it with query-level structure, about which we have already written in the first article of the guide.
The Google Shopping device strategy will help you control budgets on devices and set up the budget to get maximum results from the most effective device of your store.
To set up a campaign with a structure by devices, again, firstly create one campaign with the desired ad group structure and settings at the campaign level.
Suppose the first campaign we are doing is the desktop campaign. In the settings, set the bids for mobile and tablets at -100%.
Create 2 copies of the campaign: one will be mobile, the second — tablet. In mobile one set a bid adjustment of -100% for desktops and tablets. For the tablet — -100% for mobiles and desktops. Be sure to accurately name the campaign to easily navigate them. And now you already have campaigns with segmentation by device! Easy, right?
What to monitor after setting up a strategy for Google Shopping? Watch for changes in statistics, cost of conversion and their number. Perhaps some time later it will become clear that you should not spend money on some of devices. Then it is better to redistribute this budget to other devices.
Test this option and if the results don’t get better, try to return to the general campaign back but make different adjustments to the bids for the devices: +% for those that have higher efficiency, -% for those that are worse. Sometimes this option works better as the campaign gets more statistics and conversions, the cost per click starts to decrease and the number of impressions increases as the quality of the campaign grows in Google’s eyes.
The most common mistakes of this strategy for Google Shopping:
RLSA is Remarketing lists for search ads. It’ll help to interact with those users who have already touched the brand and entered the funnel: viewed the website, clicked on the “Buy” button, performed another valuable action.
This strategy is suitable for e-commerce websites which already have traffic. This is important, because you need to have at least 1000 users in your remarketing list for Google Search Network to launch such campaign.
Important! For remarketing you have a few separate lists available: for Display network, Search, and so on. The system itself generates lists — and the volume of the audience may differ. Usually, the Display network audience turns out to be noticeably bigger than the Search, however, it happens and vice versa.
The technical details of Google’s work are very complex, but it’s necessary to understand a few reasons, why audiences size is significantly different:
The reason for the significant advantage over of Display Network audience may be the fact that GDN campaigns traditionally receive more traffic than search campaigns. To understand the details, we recommend reading the discussion “Remarketing: Display vs Search audience” on AdvertiserCommunity.
You can’t gather 1000 users in the list if your website has small traffic, so only for start-up stores the RLSA strategy will not work. If you have such a case, put off the RLSA for the future and move on to the next strategy.
RLSA for Google Shopping helps not to miss those users who have already visited your website. If they are not allocated using a bid or individual campaigns, you can lose them. Maybe there is not enough budget and maybe there will be a loss in reach and our ads will not be shown. Remarketing for Google Shopping increases our chances of being shown and therefore the chances that visitors decide to convert in our online store.
There are two options for setting up RLSA for Google Shopping — Targeting or Observations. Perhaps these names will surprise you. Indeed, earlier these features were called Bid only and Target&bid.
Target audiences in addition to new users that visited the website
Target and bid
Target audiences exclusively, without new visitors
Cheat Sheet with old and new names
Let’s look at both options.
The “Observations” option helps to increase the effectiveness of all Shopping campaigns, taking into account the data of the audience that you have collected.
Read how to create and edit audiences in Google Analytics Help. We recommend creating audiences from both micro-conversions (for example, page views, time on the website) and from macro-conversions (for example, purchases). The higher the person in the sales funnel, the lower the probability of conversion. And remarketing better converts people from the final stages of the funnel, since their purchasing intentions are clearly stronger.
Users from the first stages of the funnel can watch the page, for example, for the sake of “window shopping”, i.e. “just look”. They still have a weak intention to buy but they are already potentially interested in the product — they are looking for it!
To increase the conversion rate of users at the “research” stage, you can use long-term remarketing, for example, with a duration of more than 1 week. The specific duration should be chosen based on the specifics of your product, its price, characteristics of your target audience. For example, usual pattern — the lower the price of the goods, the easier it is to decide on a purchase.
We recommend using a lower bid modifier for the audience at the “research” stage, since this user doesn’t cost much yet.
Remarketing will also help in the final stage of the sales funnel. For example, a user with an abandoned cart is still in the funnel but he/she didn’t make a purchase and it’s not necessarily that he/she will complete the buying cycle. To encourage him/her to convert, use remarketing with a short duration, for example, for 7 days. Bid modifier can be set higher to ensure that the user won’t forget about you. And the payback of this approach is higher than from first stages in the funnel.
Important! The audience for the search network must have at least 1000 users. If the website has a low traffic, such “segmented” lists simply won’t set up, especially in a short time.
The easiest way to add audiences to all ad groups at once is to use AdWords Editor.
Then you will see a list of all possible audiences that you can add. You will notice the columns “no. of users” and “no. of search users.” Since Shopping ads are still shown to people on the search network, you need to be sure that there are at least 1000 search users in the audience. Otherwise, the audience will not trigger any ad impressions.
Let’s imagine the situation. You sell shoes in your online store “Super Shoes”. Surely the basic query “shoes” is on your general list of exceptions, because using it for e-commerce advertising is completely inappropriate, right? So, what if you can show ads for “shoes” query to users who ALREADY visited your website? Since they were on your website, it means that their interest in shoes is higher than of the rest of people. So, in fact, they are already in the sales funnel — somewhere at the very top, at the research level. And the ads even on such a basic query can work, isn’t it?
This is where the “Target and bid” approach is useful, which has now become simply Targeting.
Since this campaign targets only users who already visited your website, don’t be afraid to use a wider range of search queries. Even if they are common as “shoes”. This is your chance to remind potential customers about you and convert them.
According to the advice of Google experts, it is worth setting these bid adjustments:
For mass adding audiences, it’s convenient to use the AdWords Editor.
What to monitor after setting up the strategy? Bid only is very unpretentious: it is enough to set up once and from time to time to check traffic and other campaign performance indicators. Periodically, you can test the adjustment bids. Since it is difficult to get a lot of statistics for a short time, it often makes no sense to change this adjustment. We advise you to add such audiences to this campaign and for any other chosen strategy.
The most common mistakes:
Bidding per unit and focus on ROI is the cherry on the cake. Our Penguin-team has repeatedly tested it at medium and small online stores and the result was extremely pleasing each time. Some clients come to us with a negative ROI and it is precisely that the Bidding per unit and focus on ROI allows them to take in a plus in a couple of months.
Such a Shopping strategy is suitable for stores of any size, both mono-brand and multi-brand resellers. It doesn’t matter whether your products are similar or completely different. It can be used from diapers / cosmetics / shoes to home appliances / medical equipment / electronics. So we do that.
Bidding per unit and focus on ROI will help identify weak and strong, most efficient products. In this way, it will be possible to focus on the best products to achieve the target profit level.
Step 1. For its implementation, you need to put the id in groups, as in the first strategy. This is done for bid management.
Step 2. For each product a separate bid is set based on efficiency, price and margin. It happens that there are very few statistics on a product, for example, if it is new and has not been sold yet or if it was not in demand of your audience. In this case, calculate the bid based on the average conversion rate.
Step 3. If you already know products with very low margins and low prices — you can immediately turn them off so as not to spend the budget on deliberately ineffective promotion and not to feed Shopping mobsters (The Mob Effect is a terrible thing!).
Step 4. Calculate the target CPC for products with a normal margin. This is the cost per click, at which sales for the product will bring the target profit. Only in this way can we be 100% sure that advertising will not drown business in expenses. Here is how this metric is calculated:
target CPC = product price * marginality * investment rate * conversion rate,
the normal investment rate here is the share of the revenue that the business is ready to invest in advertising (from 0 to 1). For example, an investment rate of 0.3 means that you are ready to invest 30% of revenue in advertising.
Suppose that our Carrie decided to sell backpacks through the Shopping campaign, built according to this strategy:
The average CPC of the backpack will be $50 * 0.3 * 0.5 * 0.04 = $0.3.
To manage the strategy, you will need a file with statistics on products: marginality, the price of goods and the rate of investment. It is also important to calculate the ROI, profit and target CPC. This file will help to see the real situation for each product, including their options — size, color, etc.
We often had to do such calculations by our hands — and this is a hell of a problem, we want to say. Therefore, we have gathered and developed The Panda (yes, the Penguins have created pandas, here’s the secret of the origin of the species). But ok, let’s be serious.
The Panda is a software that creates such a file with statistics ten times faster. Just link your Google Ads account and add a file with product margin. This can be done both in one campaign, and in many of them simultaneously. The system automatically calculates the remaining metrics — and voila, you are in the know and advertising will never take your business into a minus.
What to monitor after setup, how to maintain and manage the campaign? Let’s get a look.
Depending on the amount of traffic for the selected campaigns, you need to select the frequency with which the file and bids for products should be updated. Our advice — choose 2 weeks or a month, this is optimal.
After some time, you can re-enable advertising on products for which it was previously disabled. The exception is products for which margins tend to zero. Perhaps, the reasons for weak sales were previously in low seasonal demand or decline in market demand for other reasons.
After evaluation the situation on the products take out the most effective ones in a separate campaign with a separate budget. And if there are no such products, then start with the most appropriate ones, based on reports from other sources. Choose a piece of goods or a category and focus on it.
The most common mistake of such strategy is to set up the bid once and forget it. Bidding is a working tool of the PPC-specialist, so it is important to use it.
And that’s it! All important things that you need to know about 7 main Google Shopping strategies. Are you ready to roll? ‘Cause you can start now.
See our next article about Google Shopping campaigns audit to be able to make even better result.
Or request a free ROI and Profit Analysis of your Google Shopping Campaigns which will help to determine how Penguin-team can improve and scale your Google Ads account with The Panda — PPC micromanagement tool.
В версии 3.0.0
In version 3.0.0
В версии 2.0
In version 2.0:
Несколько шагов до работы в Penguin
Несколько шагов до работы в Penguin
Маркетинговая и техническая оценка
Советы для роста и оптимизации