AI-Powered Gap Analysis: 5 Ways to Find What’s Missing at Every Stage in the Buyer Journey

Trending 2 weeks ago
Source

In-person tickets are 95% sold out; drawback yours now while you can! Save $300 pinch our Night Owl tickets, ending connected May 17.

1. What topics are missing from nan large blogs?

AI tin really spot holes successful nan net itself. Some topics are missing from nan large blogs and media sites that rank high. They are skipped complete aliases seldom covered but could beryllium very absorbing to your audience. They are bluish water and differentiated, waiting to beryllium discovered.

It’s fundamentally intolerable for a quality to publication everything connected each nan blogs successful a vertical and summarize what’s not there. But this task comes people to an LLM.

Here are a fewer speedy prompts to person AI quickly thief show america nan topics that are seldom covered successful your industry.

Use these prompts to observe serious, think-piece topics:

What questions are group successful [industry] acrophobic to answer?
What counter-narrative opinions astir [topic] are slightest apt to beryllium discussed by bloggers and thought leaders?


Use this punctual to observe lighthearted, op-ed topics:

What are immoderate comparatively mundane, almost trivial [industry] topics that professionals person very beardown opinions about?


Use this punctual to observe myth-busting topics:

What mendacious things do group successful [industry] judge to beryllium true? And vice versa?


Use these prompts to find “missing statistics” for original investigation ideas:

What are nan astir communal assertions successful [industry] that are nan slightest apt to beryllium supported pinch evidence?

What caller original statistic could beryllium created done investigation connected a [industry] blog that would support nan claims made connected different blogs?


Several of nan specifications successful nan prompts tin beryllium very powerful: “counter-narrative opinions,” “common assertions,” aliases “common misconceptions.” Any of these could animate topics that are conscionable waiting to beryllium published crossed channels successful various formats.

We showed examples of nan prompts and responses successful a station astir AI-driven contented strategy, wherever we utilized a spaceport institution arsenic an example. What outer motorboat topics are missing from nan large blogs? AI nailed it pinch immoderate fascinating answers connected nan first try:

2. What keyphrases are missing from this blog post?

A batch of search-optimized articles rank, but not very high, for each kinds of phrases. Often, these phrases don’t moreover look successful nan article.

Adding these already-ranking but not-present phrases to nan contented connected a URL will thief amended nan relevance of that building and apt amended nan rankings. You whitethorn besides amended nan rankings for nan superior phrases by improving nan wide semantic relevance of nan URL.

You know. It’s an ontological thing. And what is an SEO if not a reverse accusation retrieval specialist? I digress.

How tin we find each nan phrases for which a URL ranks but does not rank high? Paid SEO devices do this, including nan Moz Pro Site Overview report. But let’s usage nan free instrumentality that has ranking information from nan superior root itself. We’ll export nan information from Google Search Console (GSC).

Give nan CSV from GSC to a GPT for SEO tips ASAP. That was rather nan alphabet soup, but you astir apt cognize precisely what I meant. Nerd.

Here’s nan process: I’ll usage my ain Search Console relationship and do a keyword spread study connected an article astir societal media videos.

  1. From nan Search Results report, make judge each 4 metrics are turned on: Total clicks, Total impressions, Average CTR, and Average position.

  2. Create a Query select to exclude branded (company name) queries.

  3. Create a Page select to show conscionable nan URL that you’re optimizing.

  4. Export to a CSV file. It will really springiness you a ZIP record containing respective files. You conscionable want Queries.csv. Open it up to corroborate there’s thing weird. Rename it gsc_queries.csv.

Using nan information from our existent rankings lets america skip keyword research. We already rank.

3. What articles are missing from my blog?

If you’ve been publishing for a while, you tin usage nan behaviour of your past readers to pass nan early of your contented strategy. The information is successful GA4 and nan study tin beryllium done by an LLM. If you usage ChatGPT, you’ll request a paid relationship to upload files.

We’ll support it elemental and conscionable export our blog station titles and immoderate basal metrics. For much elaborate analysis, you tin adhd “session root / medium” and person AI analyse capacity crossed channels, which tin pass your contented strategy. But we’ll support it elemental for now.

  1. Go to nan Pages and screens report.

  2. Set nan superior magnitude (the dropdown supra nan first column) to “Page title and surface class.”

  3. To spot conscionable nan blog articles, click “Create select +” and group nan select truthful “Page way and surface class” contains “blog” (or immoderate files your contented is within).

It should look for illustration this:

4. What topics are missing from this guide?

Next, let’s do a spread study successful nan mediate of nan funnel, wherever large contented lives. Long-form formats for illustration gated guides tin show heavy expertise and turn email lists. Downloadable formats tin beryllium saved and shared, helping to support nan marque top-of-mind.

But what’s missing from that long-form portion you made? Are location gaps? Can AI thief you spot them?

Yes. For this example, I’ll usage nan biggest format we have: our book, Content Chemistry. It’s nan “Illustrated Handbook for Content Marketing,” and it’s expected to beryllium comprehensive. Of course, it doesn’t screen everything, but possibly AI tin show america what we missed.

For this example, we had to return each nan matter retired and put it into a elemental .txt file. But for a emblematic PDF guide, that astir apt isn’t necessary. Either way, return nan record and upload it to your favourite LLM on pinch this elemental prompt:

What are nan astir important topics successful [industry / vertical / niche] that are not covered successful this [e-book / guide]?


The AI will quickly spot nan topical holes. Like magic, it quickly lists everything you grounded to cover.

You’ll astir apt already cognize astir about of them. They whitethorn beryllium unimportant aliases boring to you, but are they important to your reader? If so, see a revision.

5. What answers are missing from this webpage?

We now get astatine nan bottommost of nan funnel.

Really, we should person started present because improvements astatine nan bottommost person a bigger effect connected business outcomes. Visitors who commencement their acquisition connected a work page are mostly 10 times much apt to person into a lead than visitors who commencement connected a blog post, astatine slightest for B2B work brands for illustration ours.

Want to cheque for yourself? Head complete to GA4 and create a fewer comparisons to nan Landing page report. Then, usage nan dropdown nether Session conversion complaint to prime your lead gen conversion. The quality is huge.

Why do visitors who commencement connected work pages person into leads? Because they person commercialized intent. Every SEO learned this connected time 1 of keyword investigation class. But let’s inquire a mobility little communal successful SEO school…

Why DON’T visitors who commencement connected work pages person into leads? It’s usually because nan page grounded to reply a mobility (a deficiency of clarity). Or it grounded to reside an objection (a deficiency of confidence). Or it grounded to support claims pinch grounds (a deficiency of trust). In different words, location are gaps.

So, let’s occurrence up our gap-finding AI and inquire it to show america what’s missing. But we’ll person to usage a multi-prompt process because nan AI doesn’t yet cognize our visitors. Our gaps are circumstantial to our target audience, truthful we request to statesman pinch a persona.

If you person a battle-tested perfect customer profile, you tin simply upload it to nan AI. If you don’t, you tin usage nan AI to create 1 for you. Use this useful persona prompt:

Build maine a persona of a [job title] pinch [roles/skills/responsibility] astatine [industry/company size/geography]. This personification is looking for thief pinch [challenge/problem/task] and is considering [product/service].

List their hopes/dreams, fears/concerns, affectional triggers, and determination criteria for hiring/contacting a [industry/category/service/product].


If you usage ChatGPT, it will astir apt sanction your persona Alex. I person nary thought why. And naturally, location will beryllium accuracy issues. So, return clip to publication it closely. What’s incorrect? What’s missing? Fix it first. If you don’t, nan consequent AI responses will beryllium bad.

You capable these gaps yourself by telling nan AI to make changes. Here’s your persona-repair prompt:

Add nan pursuing to nan determination criteria for nan persona: [insert further purchaser concerns]


It won’t beryllium perfect, and it doesn’t person to be. But it has to beryllium bully capable to usage successful an information of your work pages. The much meticulous it is, nan amended nan responses will be.

Next, we’ll springiness it nan page we want to optimize. In nan aforesaid chat pinch nan persona, transcript and paste nan transcript from a cardinal work page (or a full-page screenshot) into nan AI pinch nan pursuing prompt:

You are a conversion optimization master skilled successful evaluating pages for their expertise to some pass and persuade. The astir compelling, highest converting web pages stock communal traits. The pursuing are champion practices for B2B work pages.

1. The header intelligibly indicates nan taxable of nan page, quickly letting nan visitant cognize they’re successful nan correct place.
2. The transcript intelligibly answers nan visitor’s mobility and addresses their objections
3. The bid of nan messages mostly aligns pinch nan visitor’s prioritized accusation needs.
4. Subheads for each conception are meaningful and specific
5. The page connects pinch nan visitant connected a individual level utilizing quality elements (faces, quotes, stories)
6. The transcript leverages cognitive biases successful subtle ways erstwhile applicable (loss aversion, urgency, etc.)
7. The page provides compelling calls to action

Create a database showing nan ways successful which nan pursuing webpage does and does not meet nan accusation needs of nan persona.

[paste successful nan page, aliases connect a afloat page screenshot]


Does that look for illustration a agelong prompt? In my experience, it is not. Some of nan champion prompts springiness nan AI a small database of champion practices truthful it knows what a successful deliverable will be. The punctual is for illustration a micro-blog station connected champion practices. Use your champion practices and what you cognize to activity good based connected your experience, your analytics, and your prospect.

Here’s a screenshot of a ChatGPT response. It starts by listing nan ways successful which you satisfied your persona’s accusation needs and ends pinch a database of nan persona's accusation needs that were near unmet.

More