In 2024 I had a lot of discussions about content sites. If everybody’s abandoning them and the multiples are dropping from 4x in 2021 to 2x now, did that make them attractive again?
The best time to buy a business is when other buyers are scared.
That’s a basic tenant of how value investing works. When a category falls out of favor, multiples compress, sellers get fewer offers, and the buyers who understand the fundamentals get better deals.
For content sites, I ultimately concluded the pain was not yet over and decided to keep away from the asset class, but the exercise showed me there are likely real opportunities out there.
AI is creating exactly the right kind of dynamic for finding hidden value in online businesses right now.
The fear is real, but it’s not evenly distributed
Ask anyone in the online business brokerage world and they’ll tell you the same thing: buyers are spooked by AI. They’re asking harder questions. They’re discounting more aggressively. In some categories, they’re walking away entirely.
This makes sense for certain business models. Chegg lost 99% of its stock value. Commodity content agencies are getting undercut. Content based businesses are not the fun ride they used to be.
The fear has data behind it.
But the fear is bleeding into categories where it doesn’t belong.
I’m seeing buyers apply AI-risk discounts to businesses where AI is actually a tailwind.
Strategic agencies getting questioned about whether “clients will just use AI instead.”
Course businesses getting asked if “people will just learn from ChatGPT.”
SaaS tools with deep integrations being lumped in with simple features that large platforms will absorb. Branded e-commerce being discounted because “AI will change online shopping.” Marketplaces with real network effects getting painted with the same brush as information-dependent businesses.
The result: some genuinely strong businesses are trading at lower multiples than they deserve because the AI anxiety is being applied with a broad brush rather than a scalpel.
What we look for in the AI era
Our approach hasn’t fundamentally changed because of AI, but the filter has become sharper.
We’re looking for businesses where AI is a tool, not a threat. Specifically:
Businesses where AI improves delivery but doesn’t replace the product. An SEO agency that uses AI to write first drafts faster is getting more efficient. An SEO agency whose entire value proposition is “we write blog posts for you” is getting replaced. Same industry, completely different AI exposure.
Businesses with trust-based client relationships. When a client has worked with an agency for three years, knows the team by name, and trusts their judgment on strategic decisions, that relationship doesn’t disappear because AI tools exist. The switching cost isn’t about the deliverable. It’s about the relationship.
Businesses where margins improve with AI adoption. This is the clearest buy signal. If a business can adopt AI tools and expand its margins without changing its revenue model, AI is making it more valuable, not less. That’s the opposite of what the market is pricing.
Businesses with physical, social, or data moats. Branded e-commerce with real supply chains, marketplaces with network effects, communities with genuine engagement, vertical SaaS with deep integrations. AI can’t replicate a loyal customer base, a two-sided network, or a trusted community.
Businesses with recurring revenue that isn’t tied to a single platform. Platform risk and AI risk compound. An agency with a diversified client base on annual retainers has minimal exposure to both.
Why disruption creates opportunity for acquirers
Every major technology shift creates a window where asset prices disconnect from fundamentals.
When COVID hit in early 2020, online business multiples dropped fast. Buyers froze, sellers panicked, and then Amazon piled on by slashing affiliate commission rates across most categories — some by more than half.
Suddenly every ecommerce and affiliate business looked like damaged goods. But the smart buyers recognized what was actually happening: the underlying businesses were fine, and in many cases about to boom as consumer behavior shifted online.
The market painted everything with the same brush, and the people who moved quickly picked up strong businesses at a real discount.
The same thing is happening now with AI, but at a larger scale. The headline is “AI is killing online businesses.” The reality is more nuanced: AI is killing some business models while making others more profitable. Buyers who can tell the difference have a significant advantage.
For us, the practical impact is straightforward. Sellers of strong businesses are getting fewer competing offers than they would have two years ago. The buyers who would have been bidding against us are sitting on the sidelines asking “but what about AI?” That means better terms, more creative deal structures, and sellers who are more open to stock-based consideration because the cash-offer pipeline has thinned.
I’m not celebrating other people’s fear. I’m pointing out that disciplined buyers with a clear framework for evaluating AI risk are in the best position they’ve been in since we started acquiring businesses.
The framework in practice
Here’s how the AI filter changes a real deal evaluation.
A content writing agency approaches us. $400K in annual revenue, $150K in profit, 3.8x asking price. Two years ago, this is a straightforward deal. Good margins, reasonable multiple, recurring client base.
Today, the first question is: what do the clients actually pay for? If the answer is “500 blog posts a month at $50 each,” the business has a problem. Clients are already discovering they can produce adequate blog content with AI tools at a fraction of the cost. That revenue is going to erode, if it hasn’t already.
But if the answer is “content strategy, editorial calendar management, brand voice development, and content production as part of a larger retainer,” the AI risk is much lower. The production piece gets cheaper with AI tools (margin expansion), and the strategic piece stays valuable because clients are paying for judgment, not words.
Same agency category. Completely different analysis. The difference is whether the business sells a commodity or a capability.
We’ve passed on deals in the first category and remain interested in the second. The multiples for both should be different, but the market hasn’t fully sorted this out yet. That’s where the opportunity lives.
A note for business owners thinking about selling
If you run an online business and AI has you thinking about your exit options, the timing dynamics are worth understanding.
Waiting isn’t necessarily safer. AI capabilities are improving every quarter. A business that looks moderately exposed today might look heavily exposed in eighteen months. If AI is a headwind for your business model, the best exit price is probably sooner rather than later.
On the other hand, if AI is improving your margins and you can demonstrate that to a buyer with data, waiting might increase your valuation. A year of improved margins attributed to AI tool adoption is a powerful selling point.
The worst position: a business where AI is clearly a threat, but the owner hasn’t adopted any AI tools and can’t articulate how the business survives the transition. That combination of vulnerability and inaction is what spooks buyers most.
If you’re thinking about selling an online business or want to understand how we evaluate deals, I’m always open to conversations. Reach out directly or subscribe to the newsletter for more from the buy side at onfolio.com/subscribe.
Disclaimer: This post reflects my personal observations and opinions. It is not financial advice and should not be construed as a solicitation to buy or sell any security. For complete financial information about Onfolio Holdings (ONFO), refer to our SEC filings.