Hot take: I wouldn’t even bother publishing TOFU content right now
Reason 1: AI has already scraped TOFU content
AI has already scraped, synthesized, and repackaged an enormous amount of basic educational content. And when someone asks a broad question, they can often get a perfectly adequate answer without clicking through to a brand's website.
And frankly, it doesn't really matter who dominates those searches anymore. Ranking first for a broad informational query doesn't mean much if nobody clicks through to your site or remembers your brand for a deeper search later.
More importantly, many of these searches have incredibly low commercial value in the first place.
The people you actually want to sell to probably aren't searching "What is ecommerce?" "What is content marketing?" "What is CRM software?" or "What is influencer marketing?"
They already know. And the people who are searching those terms are often students, beginners, job seekers, or people looking for a quick definition, not buyers evaluating a solution.
So even if you manage to outrank a massive brand for one of these generic terms, what exactly have you won?
I’d argue traffic is still highly valuable. But not if the person behind the search was never likely to become a customer.
Reason 2: TOFU content is mostly noise
Generic TOFU content is like the international bazaar of search. To a searcher/consuemr, the AI and first-page search experience right now is a lot like walking through an international market.
I was in Thailand in June (me being weird on IG to prove it), so I'll use that as an example.
When you stroll through a Thai market, especially near the entrance, nearly every stall looks identical.
You've got elephant pants, Muay Thai boxing shorts and gear, Thai silk, Chang beer T-shirts, elephant refrigerator magnets, woven bags, and all the other classic souvenirs.
Walk another 20 feet, and you'll see almost exactly the same merchandise again. And again. And again.
As a shopper, it doesn't really matter which stall you choose. The products are largely the same. The prices are similar. Unless one seller gives you a compelling reason to stop, you could buy your elephant pants (which are v. cool) from stall #3 or stall #333 and walk away with essentially the same thing.
That's what the first page of search results looks like for a lot of generic TOFU topics.
Search "What is ecommerce?" and you'll find page after page of brands offering essentially the same definition, covering the same basic concepts, answering the same questions, and targeting the same keywords.
This is what I like to call SNOOZE-FEST content.
And AI has made the sameness problem even more pronounced. Instead of asking you to walk from stall to stall comparing essentially identical information, it can scan the market, synthesize what's there, and give you the answer directly.
The consumer gets what they came for without needing to care which stall originally sold it.
Quick note: There is some TOFU content still worth publishing
There are some instances when publishing TOFU content still makes sense. It’s if you have:
A unique take. An argument readers can't get from an AI summary because nobody else has made it. Make it contrarian, specific, and yours.
New data. Original research, proprietary benchmarks, and survey results. AI remixes what already exists. A dataset that didn't exist until you published it is the hand-painted ceramic of the content market. And original data—even presented in an AI search result—is typically worth a click.
Heavy SME insight. The things practitioners know, and generic writers can't fake: edge cases, hard-won opinions, what everyone gets wrong in the field.
Everything else, meaning the keyword-first, definitional explainer content we all built careers on, can probably go.
We published that content back in the day (like a year ago) because it inflated the traffic charts. Those charts are collapsing, and if we're honest, the traffic was never doing much. A reader who wants a definition and a reader who wants a product are two different people, and only one of them pays you.
Value, in the Content Connect 3V Framework sense, means your content does a specific job for your reader. And generic top-of-funnel explainers have no job left. The machines took it.
So, what do high-intent searchers want?
Here's where I put my consumer hat on.
Picture this. I have a budget to spend, a problem that needs solving now, and a shortlist of tools I want to check out.
I am as bottom-of-funnel as a human buyer can get, and I want detailed information. (And, TBH, I don't care how promotional the content gets. In fact, PLEASE, get promotional. I want to see exactly how your product works, and I want to know why you think it's better than your competitor.)
I write for several SaaS brands, and here's how this translates into a nice do's-and-don'ts list.
Do:
Verbally and visually invite me into your software. Show me how it works. I'm talking screenshots, GIFs, videos, product walkthroughs, and real examples. Don't tell me your platform "streamlines workflows." Show me the workflow.
Give me customer examples. What problem did another customer have that your solution solved? Tell me about it. Give me numbers. Give me specifics. Help me picture what my results could look like.
Tell me exactly who your product is for. I don't want to spend 45 minutes digging through your site to figure out whether you serve a company like mine. Tell me the company size, use case, industry, team, or problem you're built for.
Get specific about your differentiators. "Easy to use" isn't a differentiator. Neither is "AI-powered." Tell me what your product does that competing products don't, and explain why that difference matters.
Talk about your competitors. I'm already comparing you. Help me do it. Tell me where you're stronger, where you're different, and even where another solution might make more sense.
Answer the uncomfortable buying questions. Pricing. Implementation time. Integrations. Migration. Limitations. Contracts. Support. Security. Learning curve. These are the things that determine whether I actually buy.
Make the next step obvious. If I'm ready to see the product, don't make me hunt for a demo, free trial, pricing page, or sales contact.
Don't:
Give me a listicle with your company at the top and a generic rundown of features. I know why you ranked yourself #1. You know why you ranked yourself #1. Let's stop pretending this is independent research.
Give me vague marketing copy. "Transform your business." "Drive efficiency." "Supercharge productivity." Cool. How?
Hide the product. If I've read 1,500 words about your software and still haven't seen what the interface looks like, you've missed an enormous opportunity.
Make unsupported claims. If you're faster, show me how much faster. If customers save money, give me numbers. If you have better results, show me the results.
Make me leave your site to answer basic buying questions. If I have to go to Reddit to figure out what your product actually costs, whether implementation is a nightmare, or how you compare with your biggest competitor, that's a content gap.
Write for the keyword instead of the buyer. Yes, optimize the page for search and AI discovery. But once you've attracted someone who's actively evaluating solutions, give them what they came for: the information they need to make a decision.
So go ahead and write deeply promotional content
This is where your content can do the heavy lifting your sales team will thank you for. Skip the next listicle and go deep on your own product. Consider publishing:
How-it-works walkthroughs. Pick one job your product does and document the entire workflow, annotated screenshots included. Write it so a buyer finishes the post knowing whether your product fits, before she ever talks to sales.
Feature deep dives. One feature, one problem it solves, real examples of it in use. Searches for "[product] features" and "[product] for [use case]" come from people deciding, so give them a decision-grade answer.
Use-case content mapped to pain. Start from the problem ("our reporting takes four hours every Monday") and walk through your product solving that exact problem, step by step.
Honest comparisons. A "[You] vs [Competitor]" page that admits where the other tool wins earns more trust than any superlative. Buyers are cross-checking you in ChatGPT anyway.
Add screenshots, GIFs, and short videos. If a page about your product contains no pictures of your product, that page needs a rewrite.
Case studies. Show me what happened when a real customer used your product. Start with the problem, explain what they actually did with your software, and give me measurable results.
Bonus: this is also the content LLMs pull from when someone asks "what's the best tool for X and how does it work?" Specific, product-descriptive pages give the machines something concrete to cite. Vague brand poetry gives them nothing.
BOFU content is valuable content
The reader who deserves your best content is the one holding a credit card, and what they want from you is clear, generous, screenshot-heavy answers to questions like:
How does this work
What problems does it solve for me?
Why should I buy this?
Promotional content earned its bad reputation when it was vague. Deeply descriptive promotional content is a different animal. It respects the buyer enough to show them everything.
For more weekly hot takes, subscribe to Content Connect.
