Why N-Gram Finder?
N-gram analysis is especially important because Google Ads match types have become looser over the last decade, with this trend accelerating in the last four years (we confirmed this by analysing real Google Ads account data, find it here).
This change makes exclusions like negative keywords key to optimising performance. Our N-Gram Finder is designed for you to quickly and easily identify underperforming search terms and apply negative keywords to block them.
What is n-gram analysis?
N-gram is a methodology used by PPC marketers that groups up the search term report by individual words, pairs, or trios of words to identify themes of what’s performing particularly well, or particularly poorly.
For example, the search terms “free cinema tickets” ,“free cinema popcorn”, and “free cinema locations” all contain the 1-gram “free”, as well as the 2-gram “free cinema”.
Given the length of most search term reports it isn’t possible to perform an n-gram analysis simply by reading through the search term report and splitting out the themes. It’s also very time-consuming to perform this analysis in Excel. We need to download the search term report, split the search terms into individual words, then group each term by word. We then need to split the search term into pairs, trios, or longer streams of words, and group by those phrases too.
How the N-Gram Finder works
The N-Gram Finder analyses the performance of all n-grams in the account. It does this by sifting through the search terms report and filtering out irrelevant words like “and”, “in”, or “the” based on the account's main language. It then breaks down search terms into n-grams of one or several words.
Next, we evaluate the performance of each n-gram and assign it what we call an nScore. We do this by comparing the performance of each n-gram to the account average. This performance analysis will be based on CPA or ROAS, depending on your account settings.
To score the n-gram meaningfully, we also consider the percentage of spend it accounts for. Between two poorly performing n-grams, the one responsible for a larger percentage of spend is a bigger liability for the account and thus receives a worse score.
We also estimate the potential savings from adding an n-gram as a negative. This gives you an easy overview and the ability to filter by potential impact.
Key Features
Performance prediction: Some n-grams have a glowing red dot next to them. These are n-grams where we have additional data to help predict their performance. This can either be the bounce rate (imported from GA4) or industry-specific data. This enables you to make data-driven decisions much sooner.
Pmax compatibility: The N-Gram Finder supports Pmax, making it easier to reign in this campaign type. As Google does not release cost data for Pmax search terms, this section is based on conversion per thousand impressions instead of CPA/ROAS.
Download .csv: The N-Gram Finder has a download function that allows you to work with .csv files and quickly get an overview of large accounts.
Change History: The built-in change history allows you to track the negatives added through the tool.
N-gram scoring & cost saving estimate: Quickly and easily surface the n-grams which are the worst performers and the biggest spenders.
Change history: Keep track of which negative keywords were added, when, and by whom in the built-in change history.
Using the N-Gram Finder
Use the N-Gram Finder to identify and manage underperforming search terms in your Google Ads campaigns.
Customising your view
You can choose what you want to see in the N-Gram Finder to narrow down your analysis and ensure you are looking at campaigns with comparable performance. There are three menus for customising your view:
Use the campaign selector on the left to choose which campaigns you want to see. It is often useful to look at one campaign group at a time.
In the “filters” menu, you will by default have excluded stop words, low volume n-grams, and search terms blocked by negatives. You can change these preferences and choose to exclude n-grams of one, two, or three words in length. Additionally, you can toggle performance predictive insights on and off in “Insights Mode”.
Use the date range drop-down to look at performance for different date ranges.
Evaluating an n-gram
Once you have customized your view, you will have a list of n-grams that are good candidates for negative keywords. To evaluate if an n-gram should be added as a negative, open the “actions” drop-down. Here, you can preview the search terms you would block by adding the negative. You can also see the keywords which would be affected.
If you find that an underperforming n-gram is an important search term for your business, you can choose to find the worst performing search terms containing that n-gram and add those as exact match negatives instead of the original n-gram.
Keep an eye out for n-grams with a glowing red dot — these have additional data-driven insights that can help you identify n-grams likely to underperform, even before they have accumulated enough data to base a decision on.
Adding a negative keyword
You can add one or several n-grams as negatives at a time. To add an n-gram as a negative keyword, use the “Actions” menu or select one or more n-grams in the left-hand checkboxes and click “Add as Negative Keyword”.
On the following screen, you can review insights into blocked keywords and search terms, and toggle between different date ranges to compare performance.
When adding negative keywords, you can choose:
The match type for each keyword.
One or more destinations for all the keywords. You can choose from your existing shared negative lists, create a new list, or select account, campaign, or ad group level to add the negatives to. You can add the keywords to multiple destinations.
Adding negatives to Pmax
Pmax campaigns do not allow us to apply negatives at the asset group or campaign level, nor can we add a negative list to a campaign ourselves. To add negatives to Pmax, create a negative list and ask a Google representative to apply it to your campaign. You can do this through a Google contact or by using their support contact form here.
Once the list has been added to your Pmax campaign, you can continue adding negatives to it through the N-Gram Finder without needing to contact a Google rep. Another option is to add the negative at the account level, where it will also be valid for your Pmax campaigns.
Technical notes
Before scoring and n-gram, we perform a ‘Minimum Clicks’ calculation which looks at on average, for this keyword and keywords like it, how many clicks does it take to get to 3 conversions (minClicks = (1/AvgConvRate)*X). For example, if a group of campaigns has an average conversion rate of 10%, then minClicks = 30 clicks. We will only score an n-gram once it reaches the minClicks threshold.
For N-Gram Finder’s performance prediction to use Bounce Rate, you need to connect your GA4 account to your Opteo account.
Google actively restricts the speed and volume of access to n-gram data for Pmax. Because of this, it can take up to a couple of hours to load your Pmax n-grams for the first time.