The problem
The keyword an advertiser targets and the search term a person actually enters in Google can differ. Reviewing the search terms report helps a team identify which queries relate to the business. Across multiple brands, that review can involve thousands of rows.
The same search term may be a sales opportunity for one brand and irrelevant traffic for another. Terms therefore need to be assessed against the products and services each business offers.
This application, whose platform is not named, aimed to classify search terms across multiple brands and prepare them promptly for review. JEV was integrated into this recurring task within a data analysis platform.
Analysis objectives
- Number of search terms evaluated per brand
- Analysis time from receiving the term list to having classifications ready
- Total analysis cost for the run
- Separation of uncertain terms and detection of unsuitable negative keyword recommendations during human review
How the system works
- 01 Google Ads search terms for the selected brands are brought together in a data analysis platform.
- 02 JEV evaluates search intent and relevance to each brand’s products or services.
- 03 Results are classified, with uncertain terms separated for further review.
- 04 An advertising specialist reviews the results alongside campaign objectives and conversion data.
- 05 The team decides which changes to apply.
Method
- TypeSafe AI’s JEV model was used to classify a large set of search terms. The data analysis platform brings that evaluation into the advertising team’s workflow.
- Both relevance to the brand and the person’s search intent are considered. Uncertain results receive separate attention during specialist review.
Where people stay involved
Intent and relevance scores alone are insufficient grounds for changing an advertising account. An advertising specialist reviews conversions, campaign objectives, the services offered and the scope of the recommendation. Competitor queries and informational searches can be valuable depending on the strategy; a category name alone is not grounds for excluding them.
Results
The implementation team reported an eight-second analysis covering thousands of search terms across 15 brands in the initial run.
Team report dated 22 September 2026. The exact term count and distribution across repeated runs have not been published for this case. This is not a turnaround guarantee for every dataset.
The reported total analysis cost for this 15-brand run was approximately US$0.20, or 20 cents.
This is the implementation team’s reported analysis cost for the initial use, not a platform subscription price or a quote for another deployment.
Data and security
The data analysis platform, brand names and actual search term lists are not disclosed in this case. Data access permissions and the scope of sharing are agreed before applying the workflow to another account.
Who this suits
A good fit
- Agencies managing Google Ads accounts for multiple brands
- Performance marketing teams classifying thousands of terms in each advertiser’s context
- Advertising specialists who want to prioritise human review of uncertain records
Not a good fit
- Teams expecting a sales or advertising return guarantee from classification alone
- Teams seeking to exclude all terms automatically without business context or conversion data
Project and technical sources
- TypeSafe AI: Introducing System One Models & Jev
Diogo Almeida, 15 September 2026. The official model announcement. The duration and cost of this case are based on the implementation team’s report.
- TypeSafe AI: JEV documentation
Official explanation of the JEV model’s capabilities and uses.
Published:2026-09-22 · Updated:2026-09-22
Frequently asked questions
What does JEV do in this data analysis application?
JEV is an AI model developed by TypeSafe AI. In this case, it was integrated into an unnamed data analysis platform to assess Google Ads search intent and relevance to each brand. Results were classified for the advertising team to review.
How does this help the advertising team?
It presents thousands of queries as a classified list. The team can identify searches that may be relevant to the brand and terms requiring further review, then consider conversion data when deciding what to change.
What do 15 brands, eight seconds and 20 cents refer to?
The implementation team reported that the initial analysis covered thousands of search terms across 15 brands in eight seconds. Approximately US$0.20, or 20 cents, was the total analysis cost for that run. It is not the price of a service package. Duration and cost can vary with data volume in subsequent runs.
Does the analysis automatically change my advertising account?
The JEV step described in this case classifies and evaluates search terms. Adding negative keywords or changing targeting is a separate decision. An advertising specialist reviews the recommendation against campaign objectives and conversion results.
Could this work for my brand?
It can be considered for brands that regularly review Google Ads search terms and teams managing multiple accounts. Term volume, products or services and the current review process need to be assessed together to determine the fit.
What information is needed for implementation?
The number of accounts, approximate search term volume and the most time-consuming part of analysis are identified first. These inform the scope, required data access and responsibilities of the reviewing team.
✦ PI ASSISTANT · ARTIFICIAL INTELLIGENCE
onlineAsk the DijitalPi AI Assistant your question
No need to fill in a form and wait for a reply — write your question and let Pi, trained on DijitalPi's 20 years of know-how, answer within seconds.
- Instant answers, 24/7
- Service-specific context — no generic replies
- Can connect you to a free consultation if you wish
Pi Assistant answers are for information purposes; for a proposal get in touch.
ASK AI ABOUT DIJITALPI
Let an AI explain what DijitalPi does.
Opens your chosen assistant with a ready research prompt. It reads the site live and answers.
Cookie preferences We use cookies to improve your experience and to analyse site traffic anonymously. Details: Cookie Policy · Privacy