How To Use LLMs for Competitive Research and Gap Analysis

Your contented strategy is failing because it's built connected guesswork. 

Keyword investigation and a fewer Reddit threads mightiness make you deliberation you understand your audience, but you're missing the full image without broad assemblage research.

If you determine to do manual assemblage research, it takes forever, and moreover pinch a afloat team, it’s easy to place cardinal opportunities. 

LLMs thief you observe deeper assemblage insights and uncover really competitors meet assemblage needs you’ve missed.

In this post, I'll show you really to usage LLMs to find assemblage insights, audit content, and place high-impact contented opportunities.

Understanding what your assemblage cares astir is the instauration of any content strategy. AI tin toggle shape scattered information into actionable insights, pulling accusation from platforms wherever your assemblage spends time. 

Here’s really I attack it, utilizing the Pooch & Mutt site arsenic an example.

Competitor websites uncover really your assemblage behaves and what contented captures their attention. LLMs, for illustration Microsoft Copilot and Google Gemini, simplify competitor research.

Start pinch a tract overview

Open your chosen LLM and commencement pinch a basal punctual like: Tell maine astir poochandmutt.co.uk. The AI revealed that they specialize successful functional canine food, targeting dogs pinch wellness issues. Their products usage earthy ingredients and are vet-approved, which helps maine understand their unsocial trading points (USPs).

You tin past inquire much system questions, specified as:

• What is poochandmutt.co.uk's target assemblage based connected their website content?
• Based connected the above, who are poochandmutt.co.uk's main nationalist competitors?
• What reside and messaging does poochandmutt.co.uk usage to link pinch their audience?
• How does poochandmutt.co.uk position its products compared to mainstream canine food?
• What customer symptom points does poochandmutt.co.uk reside connected their merchandise pages?

Customer sentiment analysis

You could past get a basal knowing of a site’s user sentiment for further insight. I for illustration utilizing the Google Chrome Extension Instant Data Scraper for this step. 

Go to the site’s main reappraisal platform, specified arsenic Trustpilot, and scrape the testimonials utilizing the Instant Data Scraper Chrome hold to stitchery them astatine scale. Next, create a caller Google expanse and adhd your scraped reviews/testimonials.

You tin past provender that expanse to your LLM of prime and analyse their reviews/testimonials pinch targeted questions, including:

• What are the astir communal issues customers mention successful poochandmutt.co.uk reviews?
• What are the astir often mentioned products successful the reviews for poochandmutt.co.uk?
• What circumstantial merchandise benefits do customers mention astir often successful reviews for poochandmutt.co.uk?
• What hesitations aliases objections do customers mention earlier purchasing from poochandmutt.co.uk?

You tin repetition these steps for competitor sites to stitchery unsocial insights astir your competition. Make judge to reappraisal the output for accuracy, arsenic LLMs tin hallucinate. 

Find your competitor’s anemic spot pinch AI-powered data

Reddit offers unfiltered insights into what your assemblage cares about. LLMs tin place progressive subreddits and aboveground celebrated discussions.

For Pooch & Mutt, I utilized Google Gemini. With a clear knowing of the target audience, I asked:

"What subreddits are UK canine owners apt to talk canine nutrition? Please investigation and springiness maine backmost immoderate existent subreddits pinch the organization size."

The output categorized subreddits into large, medium, and mini communities, pinch descriptions for each. It hallucinated personnel counts, but the wide estimations were useful. I’d attraction connected mean to smaller communities for amended engagement. Then,  cross-reference to corroborate the subreddits exist.

A speedy follow-up is asking Gemini: "Please export this into a spreadsheet." It generates a array pinch an Export to Sheets button, which is perfect if you usage Google Workspace. 

While the export tin beryllium messy (my expanse title ended up as "Please export this into a spreadsheet"), it’s an easy measurement to build a database. Rename the sheet, append caller findings, and you’ll quickly build a broad list.

For Pooch & Mutt, I utilized a elemental prompt: "What’s the latest news astir canine nutrition?" Gemini highlighted the XL Bully ban, which was meticulous astatine the clip but reflected older news. This is communal successful little progressive industries, truthful ever cross-reference pinch reliable sources for illustration Google News.

Gemini occasionally includes root links, but accuracy varies. This characteristic remains useful but imperfect until Google afloat integrates real-time news into Gemini. Always verify findings earlier acting connected them.

Start by exporting 12 months of Discover information from Google Search Console (GSC). It shows which pages earned visibility connected Discover, helping you place seasonal spikes, recurring topics, and contented types that consistently execute well.

Go deeper pinch a crawler like Sitebulb to extract contented from Discover URLs. It includes title tags, article content, publication dates, and metadata.

Once you person this dataset, stitch it to your GSC export utilizing the pursuing VLOOKUP look successful Google Sheets:

It creates a unified dataset, matching each Discover URL pinch its capacity metrics.

Stitch your information pinch Google Discover GPT

Once you person your dataset, provender it into Google Discover GPT. The GPT highlights cardinal themes, contented characteristics, and capacity insights from your Google Discover data. 

From here, you tin reply captious questions: 

• Are location immoderate seasonal content trends based connected the day of contented publication?
• Are there specific contented formats (guides, how-tos, individual stories) that execute amended connected Google Discover?
• Any insights I tin usage for title tags / H1s based connected articles pinch precocious CTR?
• What is the sentiment aliases reside of the top-performing content, and really does it subordinate to engagement?
• What are the formatting features of the astir engaged articles?
• What is the mean article magnitude for the top-performing articles?
• What is the mean H1 magnitude relationship for top-performing articles?

For Pooch & Mutt, this study showed that how-to guides and step-by-step tutorials consistently outperformed listicles, particularly erstwhile they were nether 1,500 words and paired pinch concise, benefit-driven H1s.

Pro Tip: Use the User Persona Generator alongside Discover insights. 

Combine Discover insights pinch assemblage information from the AI User Persona Generator. Feed it your URL or GA data to build elaborate personas based connected personification engagement patterns. It ensures your contented strategy aligns pinch existent assemblage interests.

Optimize titles and headings for engagement

AI SEO tools tin optimize your title tags and H1s based connected what performs champion successful Discover. I utilized a punctual like: "Based connected my top-performing Discover content, propose 10 title tag variations for my canine nutrition guides pinch precocious CTR." 

The results included:

• 10 Easy Ways to Improve Your Dog’s Nutrition Today
• The Ultimate Dog Nutrition Guide for Health-Conscious Owners

H1s nether 60 characters pinch actionable connection (“How to,” “Best,” “Guide to”) consistently performed better.

Before filling contented gaps, you must understand what’s already connected your site. AI contented tools tin automate contented audits and taxable tagging successful Google Sheets.

Here’s the step-by-step process I use:

Crawl your site:

Use Moz Site Crawl to extract each soul URLs. For Pooch & Mutt, I filtered the crawl to attraction connected blog and merchandise pages, but you tin set depending connected your priorities. You tin also use your sitemap to simplify the process​.

Tag contented by theme:

Once you person your URLs, import them into Google Sheets. Use the OpenAI API to use the GPT formula, which assigns themes to each URL based connected predefined categories.

To group it up, spell to Extensions > Apps Script in Google Sheets and adhd the GPT book to Sheets using this video guide. Once installed, you tin usage ChatGPT successful your sheets and adhd your prompts.

Here is the look building I person been using:

“Add your punctual here” — adhd what you would usually adhd successful ChatGPT

&CELL — to append your cell.

In my case, I applied the look to URL strings:

 For each URL, GPT categorized the contented by predefined themes:

• Dog Health
• Dog Recipes
• Dog Behavior
• Dog Breed
• Seasonality/Events
• Other (a catch-all for thing extracurricular halfway themes)

I utilized the look beneath and optimized the punctual for accuracy.

You tin transcript the prompt's building above, replacing your themes and illustration URL strings for much meticulous categorization.

I  for illustration tagging by URL drawstring utilizing the formula =SPLIT(A2, "/"), but you could besides tag by title tags aliases H1s if that amended summarizes your page content​.

Each URL had a corresponding taxable successful the adjacent column. The image supra shows Pooch & Mutt’s blog URLs were categorized by theme, making it easy to place contented patterns​. 

Clean up the data

AI-generated tags aren’t perfect, truthful manual cleanup is essential. For Pooch & Mutt, GPT sometimes miscategorized content—like tagging a wide canine wellness article as dog diarrhea.

Here’s really to cleanable up:

• Use Data > Create a Filter in Google Sheets
• In the Theme column, use the filter: Filter by Unique (B2:B)​
• Use CTRL+F to hunt for incorrect themes and update them manually

Overlay conversion data

Once you’ve tagged content, overlay Google Analytics (GA) conversion information by URL to uncover which themes thrust the astir traffic, conversions, and revenue.

For Pooch & Mutt, I discovered:

• Dog Health content had the highest engagement and conversion rates
• Seasonal content saw spikes but less conversions
• Dog Recipes attracted postulation but had a little on-site engagement​

Pro Tip: Always see an Other category. It catches contented that doesn’t fresh halfway themes and prevents disorganized outputs.

Andy Crestodina wrote an article on AI spread analysis, truthful I promote you to cheque it out.

Once you’ve audited your content, the adjacent measurement is identifying contented gaps by comparing it to competitor sites. It helps you observe themes and topics you’ve missed. 

Here’s a step-by-step process I use:

Tag competitor contented by theme

Repeat the aforesaid theme-tagging process you utilized for your site, but this clip for competitor sites. I did this pinch tails.com, a nonstop competitor to Pooch & Mutt.

Collect competitor URLs using Moz Site Crawl aliases a sitemap export. Next, import the URLs into Google Sheets and use the aforesaid GPT look utilized for your site. 

It categorized each competitor page into themes like Dog Health, Dog Recipes, and Seasonality/Events.

Count contented by theme

Once you’ve tagged the competitor’s content, count really galore articles they person for each taxable utilizing this look successful Google Sheets:

COUNTIF(C:C, G4)

In this formula:

• C:C is the competitor’s taxable column
• G4 is the circumstantial taxable you want to count, like Dog Health

It gave maine a speedy count of really galore pages each competitor had nether each theme, highlighting wherever they were much progressive than Pooch & Mutt.

Highlight contented gaps

With some datasets broadside by side, I could easily spot contented gaps. For example, tails.com had much contented on dog nutrition and seasonal canine care, while Pooch & Mutt lacked successful those areas.

To make it clearer:

• Filter by taxable and comparison counts broadside by side
• Highlight cells wherever competitors person much contented than you

Ask follow-up questions to refine gaps

To return the study further, I fed some datasets into ChatGPT. I utilized a punctual like:

It generated a database of themes worthy prioritizing. ChatGPT moreover grouped them by user intent, specified arsenic informational, transactional, and navigational, making it easier to prioritize.

Pro Tip: Prioritize gaps based connected 3 factors:

• Search volume: Focus connected themes pinch higher hunt interest
• Audience relevance: Choose themes aligned pinch your personification personas
• Site authority: Expanding into adjacent themes will beryllium easier if your tract ranks good for related topics

Discover keywords driving gross for your competitors pinch Keyword Gap

LLMs make it easier to observe gaps successful your contented strategy by utilizing competitors arsenic a benchmark. Audit your existing content, tag competitor pages by theme, and prioritize gaps based connected hunt request and relevance.

With your assemblage investigation and spread study complete, you’re fresh to create a contented strategy. In my adjacent blog post, I’ll show you really to provender the information into LLMs and usage it for contented creation, editing, and repurposing.