Have you ever ever learn one thing within the press or on a media website after which searched Google to study extra in regards to the matter?
Virtually everybody has.
After studying an article a few product or concept, many individuals use a search engine to seek out deeper particulars in regards to the matter or its components.
This workflow can drive short- or long-term model or non-brand key phrase search patterns.
By analyzing these patterns, you possibly can reply very strategic digital marketing, PR and SEO questions:
- What kinds of media placements drove these search patterns?
- What product messaging and positioning is driving viewers curiosity?
- Ought to we constantly spend money on PR or SEO as a substitute of short-term campaigns?
- How do I steal my competitor’s finest placements with our distinctive positioning?
- What publications ought to I take advantage of to check product positioning for a brand new go-to-market technique?
- How do I enhance the brand search for a brand new product launch?
These questions simply scratch the floor of the way to use this evaluation to strategic search advertising and marketing, new class design or normal product administration.
First, let’s outline model search habits within the context of digital media and search.
What’s model search habits?
How an viewers searches out model particulars on account of both short-term or long-term engagement with model messaging throughout different platforms.
Model search habits is a powerful indicator of brand name engagement.
Because the viewers learns about your model or facets of it, they may naturally search Google for extra particulars.
Model search habits can work like this:
- Drawback consciousness: Search begins after the viewers identifies a necessity, concern or potential answer from press, social or ads, usually utilizing non-branded key phrases to study extra (e.g., “finest health tracker”).
- Model-specific search: After researching or studying about potential options, the viewers might seek for particular manufacturers utilizing branded key phrases (e.g., “Fitbit” or “Fitbit knowledge”).
- Deeper seek for validation: Shoppers evaluate manufacturers and search detailed data utilizing key phrases like “vs.” or “opinions” to guage choices.
- Buy resolution: The viewers may use search phrases targeted on discovering the perfect worth, the way to purchase or possibly contact (e.g., “purchase Apple Watch” or “Apple Watch worth”).
- Publish-purchase search: After shopping for, customers might seek for assist, suggestions or group engagement (e.g., “Fitbit cellphone”).
This course of will present deep perception into what drove this model search habits.
Evaluation course of overview
This course of identifies the messaging, sources and folks driving an viewers to seek for a model.
I take advantage of this device stack to investigate:
- Google Traits.
- Glimpse.
- Ahrefs.
- Grok on the X platform.
The steps are pretty easy:
- Choose competitor phasing: That is how folks seek for a competitor or your model title.
- Uncover search patterns: Discover short-term or long-term patterns for progress in model search patterns.
- Discover the supply of affect: Discover main indicators of these patterns. What’s driving model search?
Beginning by figuring out the phrasing used to go looking the competitor (e.g., model title) ensures you’re discovering the fitting key phrase phrases that an viewers makes use of to seek out the model web site.
The main indicators, or the supply of affect, can determine patterns for:
- Progress.
- Decline.
- Random adjustments.
An viewers can seek for a model title after simply studying one or a number of strategic placements or after fixed publicity to a model’s messaging.
I’ll use Lectric eBike for example to indicate how this course of works. They’re an ebike model that has gone from 37,000 to over 210,000 organic clicks monthly, with over 150,000 clicks monthly from model searches.
Their model search has grown within the short- and long-term over the previous few years.
1. Choose competitor and phrasing
Use Semrush to determine the highest variations of brand name key phrases. Go to Challenge > Overview > Natural analysis > Positions.
Choose the highest variation or run stories for all variations of the model or product names.
On this instance, the most well-liked model variation is “Lectric eBike,” however folks additionally use “lectric.”
On this case, we see that search phrases get an estimated 33.1K searches monthly.
However folks additionally seek for one of many ebike product traces, just like the “lectric XP 3.0.”
These phrases are a place to begin for a way the model drives these searches.

Use these key phrases to seek for patterns in Google Traits.
2. Determine patterns with Google Traits
Earlier than beginning your search, use the Glimpse plugin (for Chrome) to replace the Google Traits UI and add some very helpful options.
This device provides a trendline that reveals seasonality or year-over-year (YoY) traits in an easy-to-understand format. This has saved me a lot time over exporting and analyzing the info.
Determine progress patterns (e.g., spikes or upward traits) like seasonality, year-over-year progress or random spikes.
The random spikes are my favourite as a result of they’ll present short-term impression on account of a particular occasion. Then, export the info to be analyzed later.
On this case, the expansion in model search curiosity began in the summertime of 2022 however grew considerably in 2024.
Main indicator questions:
- What causes the preliminary progress in the summertime of 2022?
- What media protection induced YoY progress in model search in 2024 from 2023 and 2022?


- Tip: Constant media protection can present YoY progress, particularly throughout peak shopping for seasons. When analyzing progress, take a look at media protection from the previous 12 months, not simply the month when the expansion occurred.
Discovering main indicators will be difficult. Typically, a number of media mentions can result in a big spike in curiosity. At different instances, ongoing advertising and marketing efforts over a number of months or years might end in gradual YoY progress.
With the questions prepared, use the device stack to investigate the info and determine the media driving the search.
3. Determine the supply of affect
Use the device stack to determine the supply that influenced the expansion patterns.
This doesn’t usually embody evaluation of YouTube movies, Instagram/TikTok or types like Reddit, however it’ll present quick insights into the messaging and trusted media sources.
For every sample you need to perceive, determine potential main indicators.
Right here’s the method:
- Grok for Twitter search: Ask Grok for media protection of the model title. Use this to determine potential traits in messaging or product launches.
- Ahrefs model mentions: Run stories in Content material Explorer to determine model mentions over time. I haven’t discovered a greater device to do that course of.
- ChartGPT evaluation: Analyze all three Ahrefs stories to determine model mentions to investigate.
Search X with Grok
Grok has real-time entry to X’s content material, so you will get current details about X’s content material and have Grok summarize the findings.
You will be artistic with the prompts, however I prefer to ask for a abstract of brand name mentions.
When Grok summarized mentions of “Lectric eBikes” on X, it supplied extra particulars that may require in depth analysis.
Grok shortly reveals the brand new product launch protection. The messaging of “high quality at a low worth” is constant across the model.




Grok has some insights, however I take advantage of Ahrefs to investigate the media protection extra deeply and validate Grok’s output.
Ahrefs model mentions
Use the Content material Explorer report in Ahrefs to determine historic traits in protection of the model on press, blogs and a few message boards. Export the total listing of publications to be uploaded to ChatGPT for evaluation.
On this evaluation, the Ahrefs bar chart illustrates the media story.
Mentions in 2021 have been extra constant, whereas late 2022 noticed a concentrate on vacation promotions. Each traits deserve additional evaluation.


Tip:
- Check utilizing phrase (use ” ” for phrase) or precise (use [ ] for precise) match Boolean queries to seek out extra exact mentions. With “lectric,” Ahrefs returned a number of mentions for “electrical,” which was not a related model point out.
Analyzing this knowledge manually will be very laborious, so I take advantage of ChatGPT to take a lot of the work out of the evaluation and discover particular insights.
ChatGPT evaluation
Add the Google Traits and Content material Explorer exports to ChatGPT (I used 4o for this text) to investigate patterns within the model mentions and model search traits.
For this instance, I shortly recognized some key insights from correlation in Google Traits and model mentions from Content material Explorer.
From ChatGPT evaluation, I discovered that the model targeted on some high-authority (e.g., Forbes) and in addition vertically focused publications (e.g., Electrek). But in addition their constant messaging about low price and top quality is interesting to the viewers on these publications.
I discovered these outputs from ChatGPT helpful for understanding the media impression on patterns or figuring out media to investigate additional.
- Get key insights into main indicators of random patterns/spikes or long-term traits


- Discover normal traits in media sorts that correlate with spikes or traits


- Determine what particular websites might have pushed spikes


A giant limitation is that this mannequin doesn’t embody bigger market traits that may affect. However you need to use Google Traits to assist determine these as nicely.
Bonus
You could need to use the PEST (political, financial, social, technological) mannequin to determine main market elements that would contribute to those progress patterns. I do know this doesn’t embody competitors, nevertheless it might fall below every of those classes.
The higher societal traits can have a big effect on demand for sure services or products.
Google Traits seek for “ebike” grew considerably after the pandemic began in 2020.
It is likely to be brought on by a rising variety of individuals are fascinated with a more healthy life-style, which helps to drive this demand.


- Tip: When importing, describe the info and the supply to ChatGPT and ask for the sector names. Then ChatGPT will possible have to scrub the info.
Analyzing model engagement with Google Traits
This text serves as a information for conducting your personal competitor search habits evaluation.
By making use of the insights shared right here, you possibly can uncover efficient methods to surpass your opponents in natural search and drive sustainable income progress.
Notes on knowledge and evaluation high quality
- Ahrefs: Whereas helpful, it doesn’t account for promoting, social media, or bodily publications.
- Grok: Supplies real-time entry to X knowledge, nevertheless it’s unclear what number of manufacturers are talked about past content material from the corporate or its group members.
- Google Traits: Doesn’t seize all searches and consists of some crawler knowledge. The “word” markers might affect progress, however there’s no robust proof of great distortion over time.
- ChatGPT: Lacks entry to real-time internet knowledge and should generate inaccurate insights (“hallucinations”). This may be mitigated by manually verifying media placements.
Contributing authors are invited to create content material for Search Engine Land and are chosen for his or her experience and contribution to the search group. Our contributors work below the oversight of the editorial staff and contributions are checked for high quality and relevance to our readers. The opinions they categorical are their very own.