This document provides a step-by-step walkthrough for generating keyword recommendations for Google Search campaigns using Targeting AI Playground. It empowers you with AI-driven, high-quality keyword (New, Positive & Negative) recommendations that can be added into your Google Ads campaigns.
You must have an AI group created using the Pixis Plugin.
One or more Google search campaigns should be part of your AI group.
Finding relevant keywords for your Google campaigns can be challenging and time-consuming. TAI Playground simplifies this process by offering automated analysis and personalized recommendations tailored to your campaign's historical data. Leveraging AI and data from sources like Google Keyword Planner and search trends, it identifies new keyword opportunities and suggests positive and negative keywords based on factors like user intent and competition levels. This streamlines keyword research, saving you time and effort while ensuring the integration of high-quality keywords directly into your Google Ad groups.
TAI Playground follows a step-by-step workflow to guide you through the keyword recommendation process. To begin with, navigate to the TAI playground from the Pixis platform:
1. From the Pixis Dashboard, search for your brand on the left side panel.
2. Go to Playground under Targeting AI.
Once you are on the Playground, you have to go through the following steps to generate keyword recommendations:
AI Initiation
Calibration
Keyword Tuning
Keyword Explorer
To begin receiving keyword recommendations, first select the AI Group for which you wish to receive recommendations and provide your brand details. This enables the TAI Playground to analyze the corresponding campaigns and associated ad groups. Targeting AI will suggest new, positive, and negative keywords tailored to your campaign objectives by utilizing historical data, your inputs, and sources such as Google keyword planner.
Select AI Group: Choose an AI group containing one or more Google Search campaigns.
Select Platform: Here, you can select Google Search.
Select Brand Website: Link the website associated with the Brand.
Select Ad Account: Choose the Google Ads account name.
Note: The country option is disabled for Google Search keyword generation as it is only required for Facebook Audience Cohort generation.
The system will analyze the part performance in the background while you can proceed to the calibration step by clicking on "Run" and then "Next".
AI Group Segmentation: Segregate your AI Groups into brand, non-brand, and competitor terms. This segmentation allows for more targeted campaign management and optimization strategies. For instance, let's say you're managing Google Ads campaigns for an electronics retailer. You create separate AI Groups for branded products like "TechZone," generic electronics like "smartphones" and "laptops," and competitor terms like "GadgetGuru." This segmentation allows for focused campaign management and optimization strategies tailored to each category.
In the Calibration step, you can fine-tune your inputs to ensure that the AI suggests the most effective keywords that meet your campaign objectives. This involves making minor adjustments to the configuration of sources, lookback period, and other parameters to identify the most relevant keywords using the available data sources.
This option leverages your existing campaign data and search term reports to suggest relevant keywords, i.e., positive and negative keywords. Scenario: You sell hiking boots and use keywords like "hiking boots," "trail shoes," "outdoor footwear," and "backpacking boots" in your Google Search campaigns.
How "Exploit" Helps:
Analyzes your data: The AI reviews how people search to find your ads (e.g., "best waterproof hiking boots").
Suggests positive keywords: It recommends adding high-performing search terms like "waterproof hiking boots" directly to your campaigns.
Identifies negative keywords: It highlights terms that aren't converting well (e.g., "cheap hiking boots") and suggests excluding them.
Explore helps you find new keyword opportunities. This is useful for new campaigns or when you don't have enough data on existing campaigns. It works by checking external sources like Google Keyword Planner to find relevant keywords for your campaigns.
For example, if you're advertising hiking boots and need more keyword ideas, enable Explore. It searches sources like Google Keyword Planner for relevant keywords such as 'durable hiking boots' or 'waterproof hiking shoes,' helping you target a broader audience interested in hiking gear.
You can enable the Explore feature during the initial learning phase of your campaign. This feature is particularly useful when launching new campaigns or experimenting with different ideas.
Data Recency allows you to select how far back the AI analyzes your campaign data, such as 7, 14, 30, or 90 days. This enables the AI to identify trends, such as seasonal variations in purchasing behavior or shifts in customer preferences over time, and generate recommendations accordingly.
Expand your keyword reach by integrating Google Keyword Planner. If you disable it, you'll only receive positive and negative keyword suggestions based on your ads manager.
If you disable Google Keyword Planner, you may not receive many new keyword suggestions. Therefore, it is recommended that you keep it enabled to generate new keywords.
Expand your reach with custom keywords. Manually add terms you want to target. Include keywords that share a similar intent to your successful ones.
Once you have configured the calibration options, you can click on Next and move to the Keyword Tuning process.
In this step, you can fine-tune several parameters to generate keyword recommendations tailored to your campaign goals. By adjusting parameters such as keyword cost, impressions, competition index, and search volume, you can refine your keyword selection process for optimal campaign performance. For example, you can prioritise keywords with low competition and high search volume to maximise visibility and engagement.
Keyword Cost: The Keyword Cost setting in the TAI Playground lets you control how much weightage the AI places on the cost-per-click (CPC) or Cost-per-acquisition (CPA) of keywords when generating recommendations. It's like a sliding scale where you balance cost with your other campaign goals. Moving the slider towards higher values prioritize high-cost options. Prioritizing keyword cost is helpful when you have strict budget limitations or are operating in a market with high average CPCs.
Impressions Significance: Choosing "quantum" prioritizes quantity over quality in keyword selection, potentially providing a larger pool of keywords but with lower conversion rates. Opting for "quality" ensures the best-performing keywords but limits the quantity. Adjusting the slider towards "quantum" expands the keyword pool, but the default setting is optimized for the best performance.
Competition Index: This is a metric that indicates how much competition there is for a particular keyword among other advertisers. It is often shown as "low," "medium," or "high," similar to Google Keywords Planner. For example, a high index for "hiking boots" means many advertisers are bidding on it, indicating strong market competition, while a low index implies fewer advertisers are targeting it.
When configuring keyword competition settings, select all types of competition index to gain a comprehensive understanding of keyword competitiveness. This allows for informed decision-making when selecting and prioritizing keywords for your campaigns.
Monthly Average Search Volume: Set the desired search volume range for suggested keywords. This metric reflects the number of searches for a specific keyword within a month. For instance, a high search volume for "hiking boots" suggests a large number of people are searching for this term, indicating strong interest in hiking boots. Conversely, a low search volume implies fewer searches for that keyword.
Monthly Average Search Volume: Set the desired range for keyword search volume.
High Search Volume: Indicates strong interest and higher competition, which may lead to higher advertising costs.
Low Search Volume: Shows less interest and competition, potentially resulting in lower costs but also a smaller audience.
This metric helps gauge both keyword popularity and potential advertising spend.
Keyword Tuning Configuration
Stick to the default tuning configuration for keyword tuning. This ensures that your campaign leverages the most effective settings for keyword optimization, minimizing manual adjustments and streamlining your campaign management process.
Based on the chosen settings, once click on Generate Keywords, the tool generates keyword suggestions for your Ad groups. You can find them in the next step, Keyword Explorer.
In the Keyword Explorer, you can see the keywords that Pixis AI recommends for your ad groups. These recommendations include New, Positive and Negative Keywords based on configurations from previous steps.
Note: New Keywords are generated if you’ve enabled “Explore” in step 2. Positive and Negative keywords are the result of “Exploit”, which suggests keywords based on the historical data of your campaigns.
You will get the following types of Keyword recommendations:
New keywords: These are recommendations sourced from Google Search Term and Google Keyword Planner, suggesting additional terms to include in your ad campaigns. For example, if you're advertising hiking boots and notice that "waterproof hiking boots" is a popular search term with high search volume, Pixis AI may suggest adding it as a new keyword to target users interested in this specific feature.
Positive Keywords: Keywords suggested based on your historical data and search term reports. These keywords can be added to your ad groups.
Allows bulk editing of match types and assigning keywords to different ad groups
Negative Keywords: Manage a list of keywords you want to exclude from your campaigns, such as irrelevant searches or those leading to low conversions.
Allows bulk editing of match types and assigning keywords to different ad groups.
The keywords are recommended based on the following two views:
In this view, keywords are grouped according to their recommended ad group, showing how they fit into your campaign's structure. This helps you see which keywords are relevant to specific ad groups, making it easier to create targeted ads.
Always analyze keyword performance at the Ad Group level to understand the effectiveness of your targeting and messaging strategies. Publishing and optimizing ads at the Ad Group level ensures that your campaigns are optimized for maximum performance and relevance.
In this view, keywords are grouped by similar meanings rather than by ad groups. This gives you a broader understanding of keyword themes and trends, helping you spot overarching patterns in user intent.
Cluster Name: Cluster names are labels assigned to groups of keywords that share a common theme, intent, or focus.
You can filter keywords based on quality. High k-sensitivity results in fewer but higher-quality keywords, while low k-sensitivity generates a larger number of keywords, potentially including those with lower quality.
Bulk actions help you apply the below setting to selected keywords all at once.
Change Match Types
Switch a group of keywords between exact, phrase, or broad match types to refine how your ads match with search queries.
Consider using both broad and phrase match types to capture diverse search terms and user intents. These match types offer insights into user search behavior and aid in identifying relevant campaign keywords.
Exact match type is default for new keyword ideas. For positive and negative, match type will depend on search terms.
Behavior (Only for Positive/Negative Keywords)
You can move selected keywords from negative to positive and vice versa. For example, If several keywords are performing poorly or are irrelevant, you can add them to your negative keyword list in bulk to prevent your ads from showing for those terms.
Move Keywords
Transfer selected keywords from one ad group to another, to organize keywords or create better campaign structure. You can also create a new ad group to move selected keywords. You can also create a new AI Group and move the selected keywords to it.
Duplicate Keywords
Create copies of keywords and place them in different ad groups to test variations or expand your targeting.
Delete Keywords
Remove a batch of unwanted keywords from your list.
To select all keywords listed under the tab, simply click on the Check Mark icon next to the Bulk Action button. Conversely, to deselect keywords, click on the Cross icon.
You can mark the keywords you want to move to another Ad Group. To move the Keywords follow the below steps:
Click on the star button to mark the keywords you want to move
Click on Bulk Action
Click on Move
Select the campaign
Click on + New Ad Group
Name the New Ad Group
Click Apply
Filter Keywords
You can filter the new recommended keywords based on the following:
Monthly Avg Searches: You can set the maximum and minimum keywords search volume of search a keyword should have.
Competition Index: You can filter the keywords based on their competition level index.
Three Month Change: The percentage change in the average monthly search volume of a keyword over the past three months. This metric helps you identify keywords that are gaining or losing popularity in searches.
Impressions: Filter keywords based on the number of searches the keyword received.
Avg CPC: Sort keywords on the basis of their Average cost-per-click(CPC).
Cost/Conversions: Filter keywords that have higher or lower cost/conversions.
Filter Positive & Negative Keywords
Download a selection of keywords and their associated data (search volume, competition index, etc.) into a spreadsheet for further offline analysis or editing.
Here’s a demo keyword recommendation sheet.
You can save specific keywords under existing Ad groups on Pixis Audience Manager by clicking the Star icon and then selecting "Save Selected" from the Bulk Actions menu.
Once you have selected the keywords you want to publish to Google Ads Manager, you can click on Review > Publish.
Additionally, you can export recommendations as a CSV file for later use or save your keywords in Audience Manager for future reference.