Common Settings

These settings will stay the same across all of your Campaigns.

Campaign:
Branded Search

This campaign targets your Branded search terms. It is mostly defensive in nature: It prevents your competitors from bidding on your brand term cost effectively. This campaign will generate the majority of your conversions, but you want to keep your cost-per-acquisition costs low, as most of the sales generated by this campaign are not incremental sales.

Ad Groups

This campaign initially contains only two ad groups: One to target the exact match version of your brand name, and another to target the phrase match version.

Campaign:
Generic Search

This campaign will target high-performing non-branded search terms. It will initially be empty, but you will populate it as you discover high performing search terms.

Eventually this campaign will contain one ad group for each performing exact match search term, as well as a second ad group with the phrase match version of the performing term.

Ad Groups

Keyword Exclusions

Your Generic Search Campaign may target your Branded search terms. Create a Negative Keyword List containing your branded terms, and add it to the campaign to ensure this does not happen.

Campaign:
Keyword Prospecting

The goal of the Keyword Prospecting Campaign is to find new high performing keywords.

When such a keyword is found it should be:

1. Added to the Generic Search Campaign
2. Excluded from this Keyword Prospecting Campaign

Ad Groups

This campaign will have a single dynamic ad group called “Catch-all”, targeting all pages on the site.

Keyword Exclusions

This campaign will also target your Branded terms unless you explicitly exclude them. Create a Negative Keyword List containing your branded terms and add the list to the campaign.

Multi Country / Currency / Language

If you are targeting multiple countries, currencies, and languages, then you will generally want to replicate this campaign structure for each Currency x Language combination.

• Select “Save as Preset” and name your column set “ROAS”
• Click the column icon and select “Set as Default”

Optional:
Custom Metrics

Columns > Customize Columns > Create Custom Metric

Optionally, create the following custom metric

Cost of Sales

The percentage revenue spent to acquire a sale. This is a more business centric metric than ROAS, and is simpler to understand at a glance if a particular campaign is profitable or not.

Calculating based on Conversion Rates

The simplest way is to calculate bid adjustments based on the individual device conversion rates relative to the Campaign average

Calculating based
on Click Value

I believe a more correct way is to calculate bid adjustments based on the relative per click values of each device (the Conv. Value / Click column)

Example Using Real Data

Sample Campaign

Automating Bid Adjustments

Below is a Google Ads Script that will automatically make these adjustments for you (based on relative conversion rate). You can download the latest version of this script on GitHub here:
Google Ads Device Bid Adjustment Script

// Version: 1.13.1 Muppet
// Latest Source: https://github.com/Czarto/Adwords-Scripts/blob/master/device-bid-adjustments.js
//
// This Google Ads Script will incrementally change device bid adjustments
// based on conversion rates using the Campaign's average conversion rate
// as a baseline.
//

/***********

MIT License

Copyright (c) 2016-2021 Alex Czartoryski

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

**********/

var LABEL_PROCESSING_DESKTOP = "_processing_desktop";
var LABEL_PROCESSING_MOBILE = "_processing_mobile";
var LABEL_PROCESSING_TABLET = "_processing_tablet";

var BID_INCREMENT = 0.05;       // Value by which to adjust bids
var MIN_CONVERSIONS = 10;       // Minimum conversions needed to adjust bids.
var MAX_BID_ADJUSTMENT = 1.90;  // Do not increase adjustments above this value



function main() {
    initLabels(); // Create Labels

    /*****
      Device performance *should* theoretically not vary over time
      (unless a site redesign has been performed) and so it makes
      most sense to use a relatively long time period (1 year)
      on which to base adjustments.

      Shorter time periods included for reference, but commented out
    *****/

    //setDeviceBidModifier("LAST_7_DAYS");
    //setDeviceBidModifier("LAST_14_DAYS");
    //setDeviceBidModifier("LAST_30_DAYS");
    //setDeviceBidModifier(LAST_90_DAYS(), TODAY());
    setDeviceBidModifier(LAST_YEAR(), TODAY());

    cleanup(); // Remove Labels
}


//
// Set the Processing label
// This keeps track of which bid adjustments have already been processed
// in the case where multiple time-lookback windows are being used
//
function initLabels() {
    checkLabelExists();
    cleanup();

    var itemsToLabel = [AdWordsApp.campaigns(), AdWordsApp.shoppingCampaigns()];

    for (i = 0; i < itemsToLabel.length; i++) {
        var iterator = itemsToLabel[i].get();

        while (iterator.hasNext()) {
            campaign = iterator.next();
            campaign.applyLabel(LABEL_PROCESSING_DESKTOP);
            campaign.applyLabel(LABEL_PROCESSING_MOBILE);
            campaign.applyLabel(LABEL_PROCESSING_TABLET);
        }
    }
}



//
// Create the processing label if it does not exist
//
function checkLabelExists() {

    var labels = [LABEL_PROCESSING_DESKTOP, LABEL_PROCESSING_MOBILE, LABEL_PROCESSING_TABLET];

    for (i = 0; i < labels.length; i++) {
        var labelIterator = AdWordsApp.labels().withCondition("Name = '" + labels[i] + "'").get();
        if (!labelIterator.hasNext()) {
            AdWordsApp.createLabel(labels[i], "AdWords Scripts label used to process device bid adjustments");
        }
    }
}


//
// Remove Processing label
//
function cleanup() {
    var cleanupList = [AdWordsApp.campaigns(), AdWordsApp.shoppingCampaigns()];

    for (i = 0; i < cleanupList.length; i++) {
        var iterator = cleanupList[i].get();

        while (iterator.hasNext()) {
            campaign = iterator.next();
            campaign.removeLabel(LABEL_PROCESSING_DESKTOP);
            campaign.removeLabel(LABEL_PROCESSING_MOBILE);
            campaign.removeLabel(LABEL_PROCESSING_TABLET);
        }
    }
}


//
// Set Device Bids
//
function setDeviceBidModifier(dateRange, dateRangeEnd) {

    var STANDARD = 0;
    var SHOPPING = 1;

    for (i = 0; i < 2; i++) {
        //Logger.log('---  ' + (i==STANDARD ? 'Standard Campaigns' : 'Shopping Campaigns'));

        var labels = [LABEL_PROCESSING_DESKTOP, LABEL_PROCESSING_MOBILE, LABEL_PROCESSING_TABLET];

        for (l = 0; l < labels.length; l++) {
            //Logger.log('     ' + labels[l]);

            var campaigns = (i==STANDARD ? AdWordsApp.campaigns() : AdWordsApp.shoppingCampaigns());
            var campaignIterator = campaigns.forDateRange(dateRange, dateRangeEnd)
                .withCondition("Status = ENABLED")
                .withCondition("Conversions >= " + MIN_CONVERSIONS)
                .withCondition("LabelNames CONTAINS_ANY ['" + labels[l] + "']")
                .get();

            while (campaignIterator.hasNext()) {
                var campaign = campaignIterator.next();
                var baseConversionRate = campaign.getStatsFor(dateRange, dateRangeEnd).getConversionRate();
                var platforms = [campaign.targeting().platforms().desktop(),
                    campaign.targeting().platforms().mobile(),
                    campaign.targeting().platforms().tablet()
                ];

                //Logger.log('    CAMPAIGN: ' + campaign.getName());

                var targetIterator = platforms[l].get();
                if (targetIterator.hasNext()) {
                    var target = targetIterator.next();
                    var stats = target.getStatsFor(dateRange, dateRangeEnd);
                    var conversions = stats.getConversions();
                    var conversionRate = stats.getConversionRate();
                    var targetModifier = (conversionRate / baseConversionRate);
                    var currentModifier = target.getBidModifier();

                    //Logger.log('    Conversions: ' + conversions);
                    
                    if (conversions >= MIN_CONVERSIONS) {
                        if (Math.abs(currentModifier - targetModifier) >= BID_INCREMENT) {
                            if (targetModifier > currentModifier) {
                                target.setBidModifier(Math.min(currentModifier + BID_INCREMENT, MAX_BID_ADJUSTMENT));
                            } else {
                                target.setBidModifier(Math.max(currentModifier - BID_INCREMENT, 0.1));
                            }
                        }

                        campaign.removeLabel(labels[l]);
                        //Logger.log('    Remove Label: ' + labels[l]);
                    }
                }

            }
        }
    }
}

//
// Date range helper function
// Returns today's date
//
function TODAY() {
    var today = new Date();
    var dd = today.getDate();
    var mm = today.getMonth() + 1; //January is 0!
    var yyyy = today.getFullYear();

    return { year: yyyy, month: mm, day: dd };
}

//
// Date range helper functions
// Returns date 90 days ago
//
function LAST_90_DAYS() {
    var date = new Date(); 
    date.setDate(date.getDate() - 90);
    
    var dd = date.getDate();
    var mm = date.getMonth()+1; //January is 0!
    var yyyy = date.getFullYear();
  
    return {year: yyyy, month: mm, day: dd};
  }

//
// Date range helper functions
// Returns date 1 year ago
//
function LAST_YEAR() {
    var today = TODAY();

    today.year = today.year - 1;
    return today;
}

The latest version of this script is available on GitHub here:
Google Ads Device Bid Adjustment Script

Install the script and have it run once per week.

Every week, the script will look at the past 365 days of data for each of your Search and Shopping campaigns, and make device bid adjustments based on conversion rates.

The script will raise and lower your device bid adjustments by up to 5% each week (you can change this value if you prefer bigger adjustments). You can also adjust the date range for which you want the script to run, and what the minimum number of conversions are required to make a change.

Additional Resources

• Beginner’s Guide to AdWordsGoogle Scripts [WordStream.com]
• How to Automatically Apply Bid Modifiers [SearchEngineLand.com]
• Automated In-Market Audience Bidding [SearchEngineLand.com]
• Bidadjustments for Audiences [AdScripts.com]
• Bidadjustments for Devices [AdScripts.com]

You have been running a campaign for 7 days with a daily budget of $50. Over those 7 days you have spent $350 with an ROAS of 1500%. You are using a “Maximize ROAS” bid strategy, and want to hit an ROAS of 2000%.

What should your daily budget be?

Here is the formula:

This formula calculates what your daily spend should have been to hit your target ROAS. Although there are no guarantees that future performance will be the same as past performance, it does provide a baseline from which to work.

If you are running a smart bidding campaign in either Google Ads or Facebook Ads, and are aiming for a target Return on Ad Spend (ROAS), this is a great formula to use for adjusting your budgets.

The Answer

Using the formula above we get:

New Daily Budget = $350 ÷ 7 × 1500 ÷ 2000 = $37.50

By reducing our budget from $50 to $37.50, we should be in better shape to hit our target ROAS of 2000%.

Related Reading

• Adjusting Facebook Ad Budgets towards a target ROAS Goal

ROAS

Return On Ad Spend — It is surprising that this column doesn’t exist by default, since Google provides target ROAS based bidding strategies.

This metric tells you how much sales are generated for each dollar in ad spend.

• 100% ROAS: For every $1 spent, you generated $1 in revenue.
• 200% ROAS: For every $1 spent, you generated $2 in revenue.

Many of the automated bidding strategies in Google Ads let you specify a target ROAS goal, so this custom column allows you to see at a glance if your campaigns, ad groups, and keywords are hitting your ROAS targets.

COS (Cost of Sales)

COS stands for Cost Of Sales

This metric is the inverse of ROAS and it displays ad spend as a percentage of revenue. Cost of sales is a very intuitive metric when comparing against your profit margins. If you know your profit margins, then it’s easy to see when your Cost of Sales exceeds those margins, and when you start to lose money on every sale.

• 20% Cost of Sales: For every $1 in revenue, we spent $0.20 on ads.

% COS CPC

“What is the max Cost per Click bid that will keep me below the specified % Cost of Sales target.“

If you use Manual CPC bidding for a campaign, creating a few of these columns for various profit margins will give you access to quick calculation for your max CPC bids.

Eg:

The number of columns you create and the values you use will depends on your business and profit margins. Create one column for some typical cost targets.

Optimistic % COS CPC

“What is the max Cost per Click bid that will keep me below the specified % Cost of Sales target, under the assumption that the next click will result in a sale“

This custom column is similar to the % COS CPC column above, except that it is optimistic in assuming that the “next click” will result in a sale, and includes the next sale’s revenue in the Max cost per click calculations.

You can either hard-code your average conversion value, or you can calculate it based on past conversion data.

This column is useful when you are starting a new campaign where you don’t have much conversion data yet, and you want to be optimistic with your bidding.

Further Reading

• 7 Useful Custom Columns in Google Ads
• How to Setup Custom Columns in AdWords

• Additional Images by Variant: More control over which product images get associated with which variant.
• Exclude Out of Stock items from Google Programs. If you are paying to advertise, make sure you are only paying to display products you can sell.
• Exclude Low Stock variants from Google Programs. Exclude color ways that have low size options. Show other color options or products instead.
• GTIN: Optionally set GTIN to be your barcode value

To add the above functionality, you must create and upload a supplemental data feed to Google Merchant Center.

Step 1
Create a Supplemental Data Feed in Shopify

The first step is to create a data feed in Shopify. We can accomplish this by creating a custom Shopify Collection Template that will output XML data instead of HTML:

1. Create a new Collection Template called collection.google-update.liquid with the following code:

{% layout none %}<?xml version="1.0"?>
<rss xmlns:g="http://base.google.com/ns/1.0" version="2.0">

{% comment %}
Google Shopping / Merchant Center + Shopify Product Update Feed by Alex Czartoryski
https://business.czarto.com/2020/10/14/enhance-shopify-google-shopping/

This version: Aug 30, 2022
The latest version of this script available here:
https://github.com/Czarto/ShopifyScripts/blob/master/Templates/collection.google-feed-update.liquid

TODO: Test & handle products without variants
TODO: Test & hanlde products without color
TODO: Specify Sizes to never exclude. eg: Women's 7,8
{% endcomment %}

{% comment %} Settings {% endcomment %}
{%- assign exclude_unavailable_variants = true -%}
{%- assign exclude_variant_colors_with_limited_availability = false -%}
{%- assign ignore_x_smallest_sizes = 1 -%}
{%- assign ignore_x_largest_sizes = 1 -%}
{%- assign minimum_percentage_availability = 50 -%}
{%- assign filter_variantImages_byColor = false -%}
{%- assign use_barcode_as_gtin = false -%}

{%- comment -%}
TODO: Move this into a snippet and use capture to assign the variable
{%- endcomment -%}}
{%- case shop.currency -%}
{%- when 'USD' -%}{%- assign CountryCode = 'US' -%}
{%- when 'CAD' -%}{%- assign CountryCode = 'CA' -%}
{%- when 'GBP' -%}{%- assign CountryCode = 'GB' -%}
{%- when 'AUD' -%}{%- assign CountryCode = 'AU' -%}
{%- else -%}{%- assign CountryCode = 'US' -%}
{%- endcase -%}

<channel>
<title>{{ shop.name }} {{ collection.title | strip_html | strip_newlines | replace: '&', '&amp;' }}</title>
<link>{{ shop.url }}</link>
<description>{{ collection.description | strip_html | strip_newlines | replace: '&', '&amp;' }}</description>

{%- paginate collection.products by 1000 -%}
{%- for product in collection.products -%}

    {%- comment -%} Get color option {%- endcomment -%}
    {%- for option in product.options -%}
      {%- if option == 'Color' -%}{% capture option_color %}option{{ forloop.index }}{% endcapture %}{%- endif -%}
    {%- endfor -%}

    {%- comment -%} Make a list of Colors to exclude {%- endcomment -%}
    {%- assign colors_to_exclude = "" -%}
    {%- if exclude_variant_colors_with_limited_availability -%}
        {%- for color in product.options_by_name['Color'].values -%}
            {%- assign variants = product.variants | where: option_color, color -%}
            {%- assign variants_to_process_count = variants.size | minus:ignore_x_smallest_sizes | minus:ignore_x_largest_sizes -%}
            {%- assign available_count = 0 -%}
            {%- assign total_processed_count = 0 -%}
            {%- for variant in variants offset:ignore_x_smallest_sizes limit:variants_to_process_count -%}
                {%- assign total_processed_count = total_processed_count | plus:1 -%}
                {%- if variant.available -%}{%- assign available_count = available_count | plus:1 -%}{%- endif -%}
            {%- endfor -%}
            {%- if total_processed_count == 0 -%}
              {%- continue -%}
            {%- endif -%}
            {%- assign percentage_availability = available_count | times: 100.0 | divided_by: total_processed_count | round -%}
            {%- if percentage_availability < minimum_percentage_availability -%}
            {% capture colors_to_exclude %}{{colors_to_exclude}}#{{ color }}{%endcapture%}
            {%- endif -%}
        {%- endfor -%}
        {%- assign colors_to_exclude = colors_to_exclude | split: "#" -%}
    {%- endif -%}

    {%- for variant in product.variants -%}
<item>
    <g:item_group_id>shopify_{{ CountryCode }}_{{ product.id }}</g:item_group_id>
    <g:id>shopify_{{ CountryCode }}_{{ product.id }}_{{ variant.id }}</g:id>
    <g:mpn>{{ variant.sku }}</g:mpn>
    <g:barcode>{{ variant.barcode }}</g:barcode>
    {% if use_barcode_as_gtin %}<g:gtin>{{ variant.barcode }}</g:gtin>{% endif %}

    {%- comment -%} Additional Images by Color {%- endcomment -%}
    {%- assign additional_images = product.images -%}
    {%- for option in product.options -%}
    {%- if option == 'Color' -%}{% capture variant_color %}{{ variant.options[forloop.index0] }}{% endcapture %}{%- endif -%}
    {%- endfor -%}
    {% if filter_variantImages_byColor %}{% assign additional_images = product.images | where: "alt", variant_color | sort: 'attached_to_variant' | reverse%}{% endif %}
    {% if additional_images.size > 1 %}{%- for image in additional_images offset:1 limit:10 -%}
    <g:additional_image_link>https:{{ image.src | product_img_url: 'master' }}</g:additional_image_link>
    {% endfor %}{% endif %}

    {%- comment -%} Exclude Out of Stock Variants {%- endcomment -%}
    {% if exclude_unavailable_variants and variant.available == false %}
    <g:pause>ads</g:pause>
    {% elsif exclude_variant_colors_with_limited_availability and colors_to_exclude contains variant_color %}
    <g:pause>ads</g:pause>
    {% endif %}
</item>

  {% endfor %}
{% endfor %}
{% endpaginate %}
</channel>
</rss>

Available on github here

2. Create a new collection called “google-update” and choose google-update as your collection template.

3. Preview the collection and copy the url. Your url should look something like this: yourstoredomain.com/collections/google-update

Step 2
Configure your Feed

There are a few options you can configure in your feed, all located towards the top of your template file in the “configuration” section.

Exclude Out of Stock Variants

exclude_unavailable_variants = true

By default, any variants that are out of stock will be excluded from Google Shopping, Google Shopping Actions, and Dynamic Remarketing. Change the value of exclude_unavailable_variants = false if you want to disable this behaviour.

Exclude Limited Stock Color Variants

exclude_variant_colors_with_limited_availability = true

This setting is intended for apparel products, where you may have many colors of a product, but limited sizes available in a specific product. Setting this value to true will cause the script to attempt to exclude the colors with low availability (so that alternative colors or products can show instead).

There are a few additional configuration items for this setting that you can change:

Filter Variant Images by Color
and Alt Text Matching

filter_variantImages_byColor = true

Setting this value to true will assign additional images to the current variant where the image’s Alt Text matches the variant’s color. Note that the primary variant’s image will always be included in the feed regardless of Alt text.

Set GTIN to Variant’s Barcode Value

use_barcode_as_gtin = true

Setting this value to true will assign your barcode value to GTIN. If you would like the option to set a different value than Barcode, it should be pretty straightforward to edit the code.

Step 3
Add a Supplemental Data Feed in Google Merchant Center

1. Open Merchant Center and go to
Products > Feeds > Supplemental Feeds > Add Supplemental Feed

Leave everything else as default values and click Continue

2. Make sure there’s a checkmark beside Content API and click Create Feed

3. You should now see your newly created feed in the Supplemental Feeds section. Click on your feed’s name and then click on Fetch Now to update your product data now.

Testing

It may take up to 30 minutes for your main feed to be updated. It is a good idea to review your products and feed to ensure that everything is coming through as expected, and tweak as required.

If everything looks good, your new sale pricing, variant images, and program availability should now be updated once per day.

Related Reading

• Add Sale Price to Shopify’s Google Shopping feed [deprecated]
• DIY Google Shopping Feed for Shopify [deprecated]
• sag_organic and product_sync in Google Analytics
• Google Dynamic Remarketing for Shopify

STEP 1
Install the Google Shopping Channel

This guide assumes that you have the Google Shopping Channel installed on your Shopify store. If this is not the case, then:

1. Install the Google & YouTube Shopping Channel
2. Connect your Google Account
3. Turn on Conversion Tracking

STEP 2
Remove Old Conversion Tracking Code

By connecting Shopify to Google Ads via the Google Shopping Channel, Shopify will begin sending conversion data to your Google Ads account. If you were already tracking conversions in Google Ads, then you need to make sure you are not duplicating your conversion data:

• If you previously had a Google Ads conversion tracking script installed in your checkout, then remove that code.
• If you were importing conversions from Google Analytics, stop importing those conversions.

STEP 3
Conversion Windows and Attribution Models

Although Shopify created multiple conversion actions, you only need to worry about the Google Shopping App Purchase conversion event.

Click on the Google Shopping App Purchase conversion action and update the following settings:

One Conversion per click

If you count “Every Conversion” you increases the risk of double attribution across your various channels.

For example: A user clicks on your Google ad and makes a purchase. It makes sense to count and attribute this conversion to Google Ads. If this same user later receives your newsletter, and purchases again, then you probably want to attribute that sale to the Newsletter and NOT to Google. Setting conversion count to “One” instead of “Every” ensures that only the first sale gets attributed to google, and not the second.

Click-Through Conversions

Arguably, you should set your Click-Through Conversion window to 30 days. 90 days seems extremely long, and could increases your chances of double attribution across multiple sales channels.

However, if you are using Performance Max or Automateed Bidding, the bidding algorithm can only take into account conversions that have occurred within the specified conversion window. So theoretically, the longer your conversion window, the more data for Smart Bidding to optimize with. So the 90 day click-conversion window can probably stay.

View-Through Conversions

Set the View-Through Conversion window to only 1 day. A longer View-Through conversion window is dangerous, and will lead to over attributing sales to Display Remarketing, YouTube, and Performance Max. This will invariably cause you to overspend on those campaigns.

Further Reading

• The Value of View-Through Conversions
• YouTube Attribution Poaching
• Display Remarketing Over Attribution, Poaching, and Fraud

If you install the Google sales channel on Shopify, you may notice a rise in strange referrals in your Google Analytics reports.

Medium = product_sync and Campaign = sag_organic

Where do sag_organic and product_sync come from?

When the Google sales channel uploades your product feed to Google Merchant Center, it appends the following UTM tags to the product url:

The following querystring is appended to your product’s link url:

&utm_medium=product_sync&utm_source=google&utm_content=sag_organic&utm_campaign=sag_organic

Why do we need these UTM parameter?

Today, you no longer need these UTM tags and you should remove them from your product urls.

These UTM parameters were created by Shopify to track organic free traffic from Google Shopping. They were meant to distinguish between organic traffic from Shopping vs organic traffic from Search.

Although well intentioned, this implementation altered the default behavior of Google Analytics, which was unexpected and not desirable.

Another issue with the Shopify implementation is that they didn’t include a canonical link in the feed. A canonical link ensures that only the “clean” version of the product link (without UTM tags) is included in Google’s search index. Without this, there is a risk that your organic search traffic will also include these UTM tags (which will really mess up your analytics).

What does sag_organic mean?

SAG stands for Surfaces Across Google.

Surfaces Across Google was the original “free shopping listings” in Google. It has since been renamed to “free shopping listings” and the name “SAG” is no longer used.

Fixing your Product Urls

If you set up Auto-Tagging correctly in both Google Ads and Google Merchant Center, Google Analytics will accurately separate Paid and Organic Shopping traffic.

Step 0
Switch to Classic Mode

To follow these steps, you need to be in Google Merchant Center “Classic” Mode. If you are instead in “Google Merchant Center Next” mode, follow these steps first:

1. Click on the Help logo in the top right
2. Click the last menu option “Use classic Merchant Center“
3. Confirm by clicking “Switch Now“

Step 1
Remove the UTM tags

To remove the UTM parameters from your Product Links

1. Go to Merchant Center > Products > Feeds
2. Click on the Content API feed
3. Click on the Feed Rules tab
4. Click the big blue PLUS + button, type link in the field, and select link from the drop down.
5. For Data Source, select Set to, type link and select link from the Primary Feed: Content API list.
6. Click OK
7. Set the default behaviour if the link is blank to “Leave Blank”
8. Now go to Modifications > Add Modification
9. Choose Optimize URL > Set Parameter

8. Click OK and look in the right side column preview to see if the link has been properly updated.
9. Click Save as Draft and Apply

Step 2
Set your Canonical Link

“If you use tracking parameters in your link attributes, it is recommend that you use the canonical_link attribute to provide a canonical URL. Use the canonical_link attribute to ensure that products are associated with the correct URL in the Google Search index.”

Google Merchant Center Help
https://support.google.com/merchants/answer/188492?hl=en

To ensure that only the “clean” non-utm version of the product link is included in Google’s search index, follow these steps.

1. Go to Merchant Center > Products > Feeds
2. Click on the Content API feed
3. Click on the Feed Rules tab
4. Click the big blue PLUS + button to add a new Rule and type canonical and choose canonical link from the drop down.
5. For Data Source, select Set to, type link and select link from the Processed Attributes list.
6. Click OK
7. Change the default behaviour if canonical_link has no value to “leave blank”
8. Add Modification > Find and Replace
   Now add two Find & Replace operations: One to remove the UTM tags and another to remove the Currency parameter

9. In Advanced Options, select “Search as regular expression“
10. Click OK and confirm in the right side column preview that the new link looks nice and clean, without any parameters appended.
11. Click Save as Draft and Apply

Step 3
Configure Auto-Tagging

This will configure Google Ads and Google Merchant Center to properly tag your product urls.

1. Google Ads (in the left hand menu)
   Settings > Account Settings > Auto Tagging: ON
2. Google Merchant Center:
   Gear Icon > Conversion Settings > Auto Tagging: ON
3. Google Analytics:
   Property Settings > Advanced Settings > Allow Manual Tagging to Override: OFF

Done!

You may need to wait a few hours for your feed to update with the new values. You should now have the default vanilla setup that Google Analytics expects.

Create a Price Feed in Shopify

The first step is to create a data feed in Shopify containing your products sale and regular prices. We can accomplish this by creating a custom Shopify Collection Template that will output XML data instead of HTML:

1. Create a new Collection Template

{% layout none %}<?xml version="1.0"?>
<rss xmlns:g="http://base.google.com/ns/1.0" version="2.0">
{% paginate collection.products by 1000 %}
{%- case shop.currency -%}
{%- when 'USD' -%}{%- assign CountryCode = 'US' -%}
{%- when 'CAD' -%}{%- assign CountryCode = 'CA' -%}
{%- when 'GBP' -%}{%- assign CountryCode = 'GB' -%}
{%- when 'AUD' -%}{%- assign CountryCode = 'AU' -%}
{%- else -%}{%- assign CountryCode = 'US' -%}
{%- endcase -%}
<channel>
<title>{{ shop.name }} {{ collection.title | strip_html | strip_newlines | replace: '&', '&amp;' }}</title>
<link>{{ shop.url }}</link>
<description>{{ collection.description | strip_html | strip_newlines | replace: '&', '&amp;' }}</description>
{% for product in collection.products %} 
  {% for variant in product.variants %}
    {%- if variant.compare_at_price > variant.price -%}
      {%- assign OnSale = true -%}
      {%- assign Price = variant.compare_at_price -%}
      {%- assign SalePrice = variant.price -%}
        <item>
            <g:item_group_id>shopify_{{ CountryCode }}_{{ product.id }}</g:item_group_id>
            <g:id>shopify_{{ CountryCode }}_{{ product.id }}_{{ variant.id }}</g:id>
            <g:price>{{ Price | money_without_currency }} {{ shop.currency }}</g:price>
            <g:sale_price>{{ SalePrice | money_without_currency }} {{ shop.currency }}</g:sale_price>
        </item>
    {%- endif -%}
{% endfor %}
{% endfor %}
</channel>
</rss>
{% endpaginate %}

Also available on github

2. Create a new collection called “google-feed-sale-price” based on your Collection Template

• IMPORTANT Choose xml-pricing-feed as your collection template
• Add products to the collection (or you can create an automatic collection with Compare At Price is Greater than 1)

3. Preview the collection and copy the url.

• Your url should look something like this: yourstoredomain.com/collections/google-feed-sale-price
• When you preview your page, it should look like a bunch of unformatted text on your page. If you see images, then you probably skipped the first bullet point in Step 2.

Add a Supplemental Data Feed in Google Merchant Center

4. Open Merchant Center and go to
Products > Feeds > Supplemental Feeds > Add Supplemental Feed

• Name: Sale Pricing Update
• Feed Type: Scheduled Fetch
• File Name: google-feed-sale-price
• File Url: yourstoredomain.com/collections/google-feed-sale-price

Leave everything else as default values and click Continue

5. Make sure there’s a checkmark beside Content API and click Create Feed

6. You should now see your newly created feed in the Supplemental Feeds section. Click on your feed’s name and then click on Fetch Now to update pricing data immediately.

Done

It may take up to 30 minutes for your main feed to be updated. Any new sale pricing will now be uploaded once per day.

Related Reading

• Enhance Shopify’s Google Shopping feed
• DIY Google Shopping Feed for Shopify [deprecated]
• sag_organic and product_sync in Google Analytics

An Evergreen Campaign Structure is one that does not change much year over year: Ideally the Campaigns and Ad Sets are continually in use, and only the ad creative is changed.

Evergreen Campaigns add some sanity to your Facebook Ad management. They prevent the typical disorganized Campaign sprawl that plagues most Facebook Ad Accounts. Evergreen Campaigns encourage continuous improvement and optimization, facilitates data analysis, and lowers maintenance costs.

It usually takes a few iterations before you land on your ideal evergreen structure, but here are some recommendations to get you started on the right foot.

Campaign Structure Overview

The campaign structure will look like this

Audiences

Before you can create your campaigns, you will need to define some base audiences to capture purchasers and website visitors. A list of essential Facebook audiences that should be created is available here.

Building the Campaigns

Common Settings

All the campaigns you create will have the following settings

Conversion Events & Attribution

Initially select “Purchase” as your converion event, with a 1-day or 7 day click only converison window.

If, however, you are unable to obtain at least 50 purchase events per week for an ad set, you will want to instead choose “Add to cart” as your conversion event, with a 1-day click only conversion window.

Campaign #1:
Dynamic Product Ads

This campaign will contain all your Dynamic Product ads and ad sets. This is the most automated campaign type, with generally very little management required.

You will need to have your Product Catalog and Product Feeds setup before you can create this campaign.

Ad Sets

1. Dynamic Remarketing < 30d

2. Dynamic Remarketing 30d-180d

3. Dynamic Past Purchasers > 30d

4. Dynamic Prospecting

For high-volume sites, or for periods of high sales such as Black Friday, you can also consider creating additional more granular ad sets, with shorter time windows, or explicitly targeting cart abandoners.

For lower-volume sites, you may need to change your conversion event to “AddToCart” instead of “Purchase” if you have less than 50 purchases per week.

Campaign #2:
Past Purchasers

This campaign targets people who have previously purchased from you.

Ad Sets

At it’s most basic, the campaign contains a single Ad Set containing all your past customers:

1. Past Purchasers All Time

For high volume businesses that have large amount of customers, multiple Ad Sets could be created to have more granular control based on past purchase date.

Campaign #3:
Remarketing

This campaign targets website visitors and people who have previously engaged with your brand on facebook, instagram, or otherwise, but who have never purchased from your business.

Ad Sets

Ad it’s most basic, the Remarketing Campaign contains only two Ad Sets: One targeting site visitors up to 30 days ago, and another targeting site visitors up to 180 days ago (the Facebook maximum)

1. Remarketing < 30d

2. Remarketing 30d-180d

For higher volume sites, you can consider adding more granular Ad Sets for 10, 60, 90 day etc… Or separating ad sets into stronger/weaker pruchase intent such as Add to Carts, Product Views, Social Engagement, etc…

Campaign #4:
Prospecting

This campaign will target “brand unaware” customers. People that have never purchased from you, and that have not interacted with your brand or site in at least 180 days.

This campaign will feed your Remarketing campaigns. You can expect your Cost per Acquisition (CPA) to be relatively high, but the more you can manage to spend here, the more you will be able to spend on Remarketing.

Ad Sets

At it’s most basic, the prospecting Campaign should target a 1% look-a-like audience (based on the Purchase pixel event).

More advanced campaigns can also target 2%-10% look-a-like audiences, or custom interest based audiences. But it is important to always exclude your Past Purchasers and lower funnel Remarketing audiences so that there is no overlap with your campaigns.

1. Prospecting 1% Look-a-Like

More to Explore

In a follow-up post I will give some guidelines into how to:

• Calculate your Bid Caps
• When (and how) to change your Conversion Event from Purchase to Add to Cart.
• Building your first ads and replicating your organic posts into this campaign structure.

Related Reading

• Essential Facebook Audiences

Past Purchaser Audiences

These target users that have previously purchased from you. Larger e-commerce sites may wish to get more granular on the time-spans, but the 4 audiences cover most scenarios.

Remarketing Audiences

These will target website visitors, cart abandoners, and people who have engaged with your content on Facebook and Instagram. These are all people that are “brand aware” and are the lowest hanging fruit for generating sales.

Larger e-commerce sites may want to get more granular on the time-scales.

Prospecting Audiences

Lookalike audiences are a good start for new customer prospecting.

Related Posts

• Facebook Evergreen Campaign Structure
• Google Ads Display Remarketing Best Practices
• Shopify + Google Ads Dynamic Remarketing Setup

If you use the Facebook Channel on your Shopify store, you will notice that some product attributes are missing in the product catalog:

• Google Product Category fixed!
• SEO Description
• Gender
• Material
• Additional Images

Below is the solution to add these missing details. This will give you finer control over your Facebook Product Sets and will let you create richer dynamic ads with multiple product images.

Pre-requisites

Before we start, make sure all of the following are done:

1. Install the Google Sales Channel
   The script leverages some of the metafields that are created on by the Google Sales Channel
2. Install the Facebook Sales Channel
3. Setup Empty Field Rules
   This is not actually a pre-requisite, but it’s a good backup in case things break down with your feed. This sets some default values for fields such as condition and availability.

1. Go to Facebook Business Manager and then
   Commerce Manager > Catalog > Data Sources
2. Click on your data feed and go to Settings
3. Scroll down to Data Feed Rules and click on Add Rules > Set Default Values
4. Create default values for age_group, gender, availability, condition, material, and any other fields that make sense for your business.

Step 1
Create the Facebook Product Feed Template

To create the Facebook product update feed, we will use a “hack” to transform a standard Shopify Collection page into and XML data feed:

• In the Shopify admin, go to
  Online Store > Themes > Action > Edit Code
• Under Templates, choose Add a new Template
• Choose collection from the drop down and name your template facebook-feed-template

Paste the code below into your newly created template file and click Save.

{% layout none %}<?xml version="1.0"?>
<rss xmlns:g="http://base.google.com/ns/1.0" version="2.0">
{%- paginate collection.products by 1000 -%}
{%- assign useSEOdescription = true -%}
{%- assign additionalImagesForVariants = false -%}
{%- assign includeOutOfStock = false -%}
{%- assign filterVariantImagesByColor = false -%}
{% comment %}
This template is used to add additional information to the Facebook product catalog
Documentation: https://business.czarto.com/2019/12/11/update-your-shopify-facebook-product-feed-with-missing-attributes/

<comment:title>{{ product.title}} {{variant.title}}</comment:title>
{% endcomment %}
<channel>
<title>{{ shop.name }} {{ collection.title | replace: '&', '&amp;' }}</title>
<link>{{ shop.url }}</link>
<description>{{ collection.description | strip_html }}</description>
{% for product in collection.products %}
{%- assign Gender = product.metafields.mm-google-shopping.gender -%}
{%- assign AgeGroup = product.metafields.mm-google-shopping.age_group -%}
{%- assign Material = product.metafields.mm-google-shopping.material  -%}
{%- assign Color = "" -%}

{%- if product.variants.size > 0 -%}
{%- for variant in product.variants -%}
{%- if includeOutOfStock or variant.available -%}
{%- for option in product.options -%}
{%- if option == 'Color' -%}{% capture Color %}{{ variant.options[forloop.index0] }}{% endcapture %}{%- endif -%}
{%- endfor -%}

{% assign additional_images = product.images %}
{% if filterVariantImagesByColor %}{% assign additional_images = product.images | where: "alt", Color | sort: 'attached_to_variant' | reverse%}{% endif %}

<item>
<g:id>{{ variant.id }}</g:id>
<g:brand>{{ product.vendor }}</g:brand>
{% if useSEOdescription and product.metafields.global.description_tag.size > 0 %}<g:description>{{ product.metafields.global.description_tag | strip_html | strip_newlines | replace: '&', '&amp;' }}</g:description>{% endif %}
<g:product_type>{{ product.type | replace: '&', '&amp;' }}</g:product_type>
<g:mpn>{{ variant.sku }}</g:mpn>
<g:item_group_id>{{ product.id }}</g:item_group_id>
<g:content_id>{{ variant.id }}</g:content_id>
<g:material>{{ Material }}</g:material>
<g:gender>{{ Gender }}</g:gender>
<g:age_group>{{ AgeGroup }}</g:age_group>
{% if additionalImagesForVariants %}
{% if additional_images.size > 1 %}{%- for image in additional_images offset:1 limit:10 -%}
<g:additional_image_link>https:{{ image.src | product_img_url: 'master' }}</g:additional_image_link>
{% endfor %}{% endif %}
{% endif %}
</item>
{%- endif -%}
{% endfor %}
{%- else -%}

<item>
<g:id>{{ product.id }}</g:id>   
<g:brand>{{ product.vendor }}</g:brand>
{% if useSEOdescription and product.metafields.global.description_tag.size > 0 %}<description>{{ product.metafields.global.description_tag | strip_html | strip_newlines | replace: '&', '&amp;' }}</description>{% endif %} 
<g:product_type>{{ product.type | replace: '&', '&amp;' }}</g:product_type>
<g:item_group_id>{{ product.id }}</g:item_group_id>
<g:material>{{ Material }}</g:material>
<g:gender>{{ Gender }}</g:gender>
<g:age_group>{{ AgeGroup }}</g:age_group>
{% if product.images.size > 1 %}{%- for image in product.images offset:1 limit:10 -%}
<g:additional_image_link>https:{{ image.src | product_img_url: 'master' }}</g:additional_image_link>
{% endfor %}{% endif %}
</item>

{% endif %}
{% endfor %}
</channel>
</rss>
{% endpaginate %}

or download from Github

Although the script above will work as is, there are three items that you can configure:

useSEOdescription = true
Setting this to true will use your product’s SEO description. Setting to false will not upload any description (default Shopify feed will be used)

additionalImagesForVariants = false
Setting this to true will upload all your product’s additional images. If you use variants, all your variants will have the identical additional images (their primary image will be as-configured in Shopify)

filterVariantImagesByColor = false
Setting this to true will only upload additional images for a variant IF the ALT text of the images exactly match that variant’s color attribute.

Step 2
Select the Products to Send to Facebook

Now select which products will be included in your feed.

• In Shopify Admin, go to
  Products > Collections > Create Collection
• Enter a Title: “Facebook”
• Add Products to the collection (either manually or using conditions). If you want to include all your products, then add a rule similar to “Inventory Stock is greater than 0”.
• IMPORTANT!
  Assign the Facebook Feed Template to this collection.
  In the bottom right column of the page, you should see a section called Theme templates. Choose collection.facebook-feed-template otherwise none of this will work.
• At the bottom of the page, click on Edit Website SEO and enter “facebook” as your collection url handle. (optional – but helps with remembering your feed url)
• Save and Preview the collection. You should see unformatted text on the screen. This is your Facebook feed. Do a “View Source” in your browser to preview the XML data.
• Copy the url of this page. It should look similar to https://www.yourstore.com/collections/facebook

Step 3
Upload your Product Feed to Facebook

• Go to your Facebook Business Manager and go to Commerce Manager
• Expand Catalogs in the left hand menu and select Data Sources
• In the Supplementary Feeds section, click on Add supplementary feed
• Click Upload Feed and choose Scheduled Feed
• Paste in the feed url generated in Step 2 and click Next (leave Login Details blank)
• Schedule your updates to occur daily or hourly. Turn off automatic updates (or try leaving this on — I’ve never tried it)
• Select the data feed to update
• Name your supplementary feed and click Upload
• Wait for Facebook to finish fetching your feed

Done

Your products should now have the missing information added. You will probably want to repeat STEP 3 anytime you modify or add new products to your shop.

It’s just a matter of time before Shopify starts uploading the full data specs to Facebook, but until that time, this is the workaround!

Related Reading

• DIY Facebook Feed for Shopify
• Ad Sale Price to Shopify’s Google Shopping Feed

Google’s Built-in Audiences

Before we create our own audiences, we first need to install the Google Ads Dynamic Remarketing Code. This triggers Google to automatically create some built-in Remarketing Audiences:

Additional Audiences

The additional audiences to create follow the same pattern used in Google’s built-in audiences, but with expanded membership durations. The main focus is on Shopping Cart Abandoners, Product Viewers, and Past Buyers.

Shopping Cart Abandoners

Google will have already created a “Shopping cart abandoners” audience with 30 day time window. We will create the following additional audiences for 7, 14, 90, and 180 day durations:

• Shopping cart abandoners: 7d
• Shopping cart abandoners: 14d
• Shopping cart abandoners: 90d
• Shopping cart abandoners: 180d

Past Buyers

Google will have already created a “Past buyers” audience with 30 day time window. We will create these additional audiences for 14, 90, 180, 365, and 520 day durations:

• Past buyers: 14d
• Past buyers: 90d
• Past buyers: 180d
• Past buyers: 365d
• Past buyers: 520d

Product Viewers

Google will have already created a “Product viewers” audience with a 30 day duration. We will create additional audiences with 14, 90, 180, 365, and 520 day durations.

• Product viewers: 14d
• Product viewers: 90d
• Product viewers: 180d
• Product viewers: 365d
• Product viewers: 520d

General Notes

• The Membership Durations are somewhat arbitrary. You can get more or less granular, and set your own intervals. It is however best to start with something simple and add granularity later on.
• A Remarketing Audience needs a minimum of 1,000 members to be eligible to serve. When creating your audience durations, consider how much time it will take to reach 1,000 memebers. eg: How long will it take before your site generates 1,000 abandoned carts, or 1,000 purchases? That will probably be the shortest duration with which you should start.
• You should generally add all these lists as “Observations” to all your campaigns.

More Reading…

• Google Ads Display Remarketing Setup Best Practices
• Shopify + Google Ads Dynamic Remarketing Setup

How I use this

1. I maintain an evergreen campaign structure so that I can easily measure past performance.
2. My performance target is generally an ROAS of 4.5
3. My lookback time period is generally 7 days

Once a week (or sometimes daily if we are in a period of high activity or if we have launched a new promotion), first thing in the morning, I look at the past 7 days and run the above formula for each of my campaigns.

• If my target ROAS is very close to my actual ROAS, then I don’t bother with the calculation, nor with adjusting budgets.
• If the results call for a decrease, then I decrease the budgets.
• If the results call for an increase, then I increase budgets, but usually by no more than 10%.

Related Reading

• Adjusting Campaign Budgets towards a Target ROAS