Skip to end of banner
Go to start of banner

Your first 12 Google Ads Optimization Opportunities

Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 32 Next »

In this post, we are reviewing the key opportunities of optimization you should consider after exporting your data to Google BigQuery with the following services:

1. Optimize your Google Shopping Product Data

There are many aspects of your product data that will affect their performance in Google Shopping:

  1. Title

    1. Add additional information like brand, color, size, or gender.

    2. Remove unwanted parts of the title or modify it otherwise(e.g ensure correct upper and lower case).

    3. If the title is too long you can define a rule for the priority of the attributes to appear.

  2. Description

    1. Ensure no Description is empty. At the very least we can use the title as a description.

    2. We can check different attributes and append them together if necessary.

  3. Link

    1. Add a URL parameter indicating the specific SKU (so the product is pre-selected with the proper size or color)

  4. Main Image Link

    1. Define a rule to select what is to be used as the main image (the rule can be optimized based on Clicks)

  5. Additional Image Links

    1. Make sure that up to 10 additional image links are provided

  6. Availability

    1. Go beyond in-stock/out-of-stock: add products to the feed which are available for pre-/backorder

  7. Unit pricing measure

    1. Use the unit pricing measure attribute to define the measure and dimension of your product. This value allows users to understand the exact cost per unit for your product. e.g. 10 Sfr. / 100 ml

  8. Google Product Category

    1. Add it where missing based on the shop categories (if at least one item in that category has the google product category)

  9. Product Type

    1. The naming of shop categories is not always accepted by google. We can modify the product type so it will be accepted by Google.

    2. If it is desired to send more than one product category the setup can be very complex. It is possible to use up to five product types in the feed.

  10. Identifier

    1. Make sure the proper identifier is always provided for all cases based on Brand/GTIN/MPN

  11. Age Group

    1. Infer it from other attributes with a mapping to another attribute like e.g. target_group. "Frauen/Männer" will change to "adult" and "Mädchen/Jungs" to "kids"

  12. Item Group ID

    1. Use the item group ID attribute to group product variants in your product data. When you use an item group ID to group your products, you can ensure that your product and its variants are shown to customers as a group, instead of separately.

  13. Shipping

    1. Some products may not comply with the account shipping settings in Merchant Center. e.g. products which include free shipping in the price. With this attribute it is possible to overwrite the default values per product.

A full review of your Google Shopping data is a good place to start to identify optimization opportunities.

2. Manufacturer and custom product Bundles

Having bundles in Google Shopping offers several benefits and adds to the overall appeal of the platform. Firstly, manufacturer bundles with their standard GTIN provide convenience to shoppers by offering a curated package of products that are commonly purchased together. This simplifies the shopping experience, saves time, and encourages customers to make larger purchases. Moreover, these bundles can attract potential buyers who may not have considered purchasing all the individual items separately, thereby increasing sales opportunities for manufacturers. On the other hand, custom bundles allow retailers to showcase their creativity and unique offerings. They enable businesses to create tailored packages that meet specific customer needs or target niche markets. Custom bundles provide flexibility and differentiation, helping businesses stand out from competitors and capture the attention of shoppers seeking personalized solutions. Overall, having both manufacturer bundles and custom bundles in Google Shopping diversifies the product selection, enhances the shopping experience, and boosts sales for manufacturers and retailers alike.

Merchant-defined Bundles can be rejected by Google due to difficulties with filling the other values properly (e.g. GTIN of main product should be used). We apply the is_bundle logic on all necessary fields and monitor continuously if your bundles are all approved according to their importance in terms of sales results.

3. Optimize the link from Google Shopping to your PDP

If your e-shop offers products in different variants (size, color, …) it might be that you are losing customers when they land on your Product Detail Page if they come to be unsure they are arriving on the right product.

For example, if a customer searched for a specific color or motif of cloth and clicked on a specific picture in Google Shopping, it might be a good idea to preselect this variant on the landing page, without the customer having to select the right variant again.

4. Target your Google Ads with Audiences

Boxalino offers 3 out of the box Google Ads Audiences which you can use for your campaign and we can add many others upon request:

  • Customers who bought during the last year but not during the last month

  • 25% percentile (top quartile) of customers with the best customer lifetime value (CLV) to promote

  • 25% percentile (bottom quartile) of customers with the worst customer lifetime value to demote

Check out the business impact of this measure for Lehner-Versand without solution!
https://www.deptagency.com/de-ch/case/daten-sei-dank-optimiertes-timing-bei-der-kundenreaktivierung-fuer-lehner-versand/

(article from our partner DEPT with whom we collaborated on this project)

5. Create Google Shopping Performance Campaigns

Typically your Google Shopping campaigns will be optimized based on target ROAS.

While this is a good start, the ROAS typically only considers the generated revenue.

In addition, other aspects could be used to influence the ROAS optimization:

  • Achieved Margin: not only the margin of the advertised products but the real margin generated

  • Customer Acquisition: the expected lifetime value of acquired customers (to which degree are products good at acquiring new customers versus getting existing customers again through a repetitively paid channel? And how much does the first product(s) this new customer buys indicate their future purchase potential?)

  • Promotion: strategic products which you want to promote

  • Zombie: products (products that have had no clicks recently in Google Shopping are likely to be demoted and simply circling them through another campaign can have a positive effect on their promotions) (as recommended by Google Best Practice: Targeting Low Volume SKUs)

  • %Sales Availability: Current variants availability according to their sales contribution (a product like a running shoe might be in stock only in sizes 10% of your customer's order, while this makes the product out-of-stock for most customers it would still be considered as active and in-stock by your systems and bring paid traffic)

  • Out: while Google Shopping avoids to some extent bidding on non-performing products, it is relevant to explicitly exclude segments of your products that are not possible to promote in a profitable way

Segmenting your products in different campaigns based on these performance factors allows you to set different target ROAS for each campaign and then go beyond the limitation of the simple ROAS optimization which only considers your revenue and nothing else.

Concrete steps:

  1. Validate the standard attribution of the products to the Margin, Customer, Zombie, Normal and Out Performance groups (tuning is possible if needed)

  2. Add a custom field in your Google Shopping feed with this new performance campaign groups

  3. Configure several Google Ads Campaign (one for each group) and associate different target ROAS with the following guideline

    1. 20%-30% lower target ROAS for the Margin and Customer Campaigns

    2. base-line target ROAS for the Normal and Zombie Campaigns

    3. filter out or at least put higher ROAS for the out campaign

6. Correct your Conversion Value

While segmenting your products in several performance campaigns (see point 4) is a good start, it is also interesting to modify the Conversion Value which is used by Google Ads to define your ROAS.

Indeed, while the definition of the cost is quite clear, the definition of the conversion value is not that simple and cost and conversion value are the only two components of the ROAS (e.g.: selling a product for 500.- for which you get only 1% margin is not the same value for your business as selling a product for 300.- on which you get 10%).

Therefore, returns and cancellations are often not considered, but they should be.

We can correct the conversion value if the order was canceled, or returned, or if another event reduced the revenue value for you.

Concrete steps:

  1. Define a corrected value for a conversion based on the effective ERP revenue (canceled value should be 0.- , returned article should lower the value accordingly, etc.)

  2. Integrate the updated conversion value in GAds with the following method: (https://support.google.com/google-ads/answer/7686280?hl=en )

    1. Look up the “Conversion action” name you use for the conversions that should be adjusted.

    2. Boxalino will prepare a datafeed.

    3. Use the feed to schedule uploads automatically. https://ads.google.com/aw/conversions/uploads/schedules

  3. (optionally) implement an API real-time call to Boxalino in your Tracking Data Layer to send the adjusted conversion value right away when available to the Google Ads tracker.

Additionally we can use another metric then the revenue, like for example the margin of the transaction, as the definition of the conversion value. This gives you the following advantages:

  1. More accurate bids for optimizing the profit.

  2. Margin labels per product are a good start but will not work if the customers buys another item then he clicked in the Ads.

  3. Profit tracking is the natural progression of revenue and ROAS tracking.

  4. By optimizing directly the profit you don't need to guess what changes in ROAS/Revenue will lead to increased profitability.

Concrete steps:

  1. Define a corrected value for a conversion based on the effective ERP margin (canceled value should be 0.- , returned article should lower the value accordingly, etc.)

  2. Integrate the updated conversion value in GAds with the following method: New Conversions (per google click id) (https://support.google.com/google-ads/answer/7014069?hl=en )

    1. Create a new Conversion action in your Google Ads Account. https://ads.google.com/aw/conversions

    2. Select Import > Manual Import using API or uplaods > Track conversions from clicks

    3. Select a Category and the conversion action name.⚠️ Select the “Secondary action not used for bidding optimisation“ first when setting up the conversion action. This allows you to see the impact in GAds before using it for the bidding and serves as a test.

    4. Select different values for each conversion and set the default value to 0. For the count we want to have each conversion separately. The Click-through conversion window should be set to 90 days and the attribution should be data-driven.

    5. Hit create and continue to confirm your settings.

    6. Let Boxalino know the name of the conversion action and we will prepare a datafeed.

    7. Use the feed to schedule uploads automatically. https://ads.google.com/aw/conversions/uploads/schedules

    8. Set up custom columns to show profit and POAS. The metric “All Conversion Value” must be selected or else it will not work. Then, choose the conversion action that you created for the formula.

    9. Set up custom columns to show revenue and ROAS. The reason is that once you change your primary conversion action in the account, then you can’t really use historical data to compare year-over-year.

    10. Observe the POAS performance and define a good starting point for the new Smart Bidding portfolio.

    11. ⚠️Do not continue before consulting an expert from Boxalino first!!!

    12. Switch the profit action to “primary“ and the revenue action to “secondary“ conversion actions.

    13. Once you’ve changed your conversion action to profit, you need to immediately make changes to your Smart Bidding setup. Set up a new Smart Bidding portfolio based on the current POAS performance.

    14. (optionally):

      You can do the same thing with just a single campaign by changing the conversion action at the campaign level and creating a portfolio bid strategy only for that campaign.

      This way, you can get a sense of what happens when you switch the conversion action.

  3. (optionally) implement an API real-time call to Boxalino in your Tracking Data Layer to send the adjusted conversion value right away when available to the Google Ads tracker.

7. Extend your Conversion Value

We can add additional values when a new customer is acquired and an even higher value when the new customer is expected to have a high future customer lifetime value e.g. based on the first product(s) bought. There are two ways to adjust the conversion value. Either directly modify it when sending the Google Ads event or adjust it by uploading the extended value with the order identifier at a later point.

Use Case:

Additional value for new customers based on their predicted CLV.

Online Concrete steps:

  1. Request the adjusted Conversion Value from the Boxalino API by providing the information that should be used in the Calculation. 

    curl --location 'https://track.bx-cloud.com/narrative/renovero_api/api/1' \
--header 'Content-Type: application/json' \
--data '{
    "username": "your_eshop",
    "apiKey": "***",
    "apiSecret": "***",
    "dev": false,
    "test": false,
    "language": "de",
    "sessionId": "value of cems cookie",
    "profileId": "value of cemv cookie",
    "widget": "search",
    "hitCount": 1,
    "filters": [
        {
            "field": "id",
            "values": [
                "$$$"
            ]
        }
    ],
    "correlations": [
        {
            "origin": "boxalino_std_bq_custom",
            "type": "conversion-value-new-customers",
            "source": "26-1474"
        }
    ]
}'

    The type needs to be configured by Boxalino. In this example we expect a category_id and the zip code concatenated by “-“ to be in the source.

  2. Ensure you modify the conversion value and currency dynamically in your event snippet.

  3. Depending on the setup it might be necessary to adjust the GA4 event to ensure the correct data. Make sure to set the currency and the correct dynamic value you receive from Boxalino for the value.

Offline Concrete steps:

  1. Define an additional conversion value for new customers (e.g.: +20% of AOV) and an even higher value for customers predicted to have a high customer lifetime value (e.g.: +50% of AOV)

  2. Integrate the updated conversion value in GAds with the following method: New Conversions (per google click id) (https://support.google.com/google-ads/answer/7014069?hl=en )

    1. Create a new Conversion action in you Google Ads Account https://ads.google.com/aw/conversions

    2. Select Import > Manual Import using API or uploads > Track conversions from clicks

    3. Select a Category and the conversion action name.💡 We highly recommend to select the “Secondary action not used for bidding optimization“ first when setting up the conversion action. This allows you to see the impact in GAds before using it for the bidding and serves as a test.

    4. Select different values for each conversion and set the default value to 0. For the count we want to have each conversion separately. The Click-through conversion window should be set to 90 days and the attribution should be data-driven.

    5. Hit create and continue to confirm your settings.

    6. Boxalino will prepare a datafeed.

    7. Use the feed to schedule uploads automatically. https://ads.google.com/aw/conversions/uploads/schedules

  3. (optionally) implement an API real-time call to Boxalino in your Tracking Data Layer to send the adjusted conversion value right away when available to the Google Ads tracker.

8. Extend your Search Ads Keywords

Boxalino can do an analysis of all the Keywords in your Google Search Ads on the basis of:

  • Google Ads Keywords

  • Google Analytics Searches (searches in your web-site)

  • Semrush SEO and SEA of your e-shop and of your competitors

  • External AI (ChatGPT & Deepl) to expand your keywords with relations and synonyms

These additional Keywords can be automatically added to your existing campaign to increase their reach.

Boxalino can also connect these new Keywords with dedicated Destination URLs to matching Topic Landing Pages to avoid any “loss of context” when the user clicks on them.

9. Switch Search Ads Destination URLs to Topic Landing Pages

Boxalino can create many different types of automated Semantic Topic Landing Pages.

These pages are related to one or several keywords and show the most relevant products (and content) related to these keywords.

For cases you don’t have a clear destination URL on a category or a brand page, you might want to try to use a Topic Landing Page instead.

Even if you already have a category page, you can still add the topic landing page in rotate mode to see in which case it performs better than your normal category pages and let google keep these ones automatically.

Finally, you can use Topic Landing Pages with a combination of topics to have a more precise destination URL and avoid any “loss of context”.

10. Generate Search Ads Sitelink & Product Extensions

You might want to use more Sitelinks but don’t have the time to set them up all manually.

Boxalino Winning Interaction Platform can extend your existing Search Ads with site links automatically based on what is most often searched together with the advertised keyword and with destination URLs going directly to optimized Topic Landing Pages.

The title and short text of the Sitelinks can be generated based on rules and the value of the combined semantic topic, but can also be generated or enriched by external AI like ChatGPT.

Product Extensions

Another way to extend a search ad is to position specific products in it.

This is something you can optimize and automate with Boxalino which can directly set the most performance Product extensions for each ad and update it when needed (out of stock, price drop or increase, seasonality, etc.).

11. Generate New Responsive Search Ads automatically

While as indicated in point 6, Boxalino can help you automatically extend your Search Ads Keywords, it will happen that none of your existing Search Ads is well suited for these automatically generated keywords (or only vaguely related).

In such case, you can generate new RSA completely automatically with Boxalino (and with the help behind of external AI like ChatGPT) and have them added to your Search ads, including:

  • Headline

  • Description

  • Destination URLs

  • SiteLinks (with Headline, Description and Destination URLs)

You can put these new Search Ads in specific AdGroup and/or Campaign and track their performance compare to your other ads.

12. GPT-enriched dashboards

Interested, ask more to Boxalino about it (smile)

  • No labels