What boring online businesses still make money?

Are boring online businesses still making money today?

Boring online businesses are still making serious money today, especially when they handle accounting, payroll, compliance, websites, hiring, payments or another job customers cannot simply stop doing.

The latest numbers are surprisingly strong. Intuit's latest annual filing shows QuickBooks Online Accounting revenue reaching $5.05 billion, up 23% in one year. Its broader Online Ecosystem, which includes payroll and payments, reached $9.92 billion and grew 19%. GoDaddy recently reported $4.42 billion in annualized recurring revenue from domains, websites, commerce and related tools. LegalZoom's subscription revenue was up 12% in its latest reported first quarter.

These are mature categories. Nobody discovered bookkeeping last year, and businesses have been buying domain names for three decades. Yet billions of dollars of additional revenue are still moving into them.

The pattern gets clearer once we stop grouping every internet business together. Companies will experiment with how they create content, design ads or write code. They have much less freedom around paying employees, keeping accounts, filing documents, maintaining a website or finding customers.

AI has made that distinction more important because optional digital output has become dramatically cheaper while many boring obligations are still sitting there, waiting to be dealt with.

Boring online business	Recent evidence	What customers keep paying for
Online accounting	QuickBooks Online revenue +23%	Accounting, reporting, tax readiness
Domains and web tools	GoDaddy ARR $4.42B	Domains, websites, email, commerce
Legal and compliance	LegalZoom subscription revenue +12%	Filing, compliance, ongoing support
CRM software	HubSpot revenue $3.1B, +19%	Customer records, sales and marketing workflows

Why are boring online businesses suddenly more interesting again?

Boring online businesses look more attractive now because AI is crushing the cost of producing digital work much faster than it is removing recurring business problems.

Creating a decent landing page once required a designer, copywriter and developer. Drafting a basic contract required either a lawyer or a template. Producing ten articles could occupy a small editorial team. Much of that work can now be produced in minutes.

That hurts some online businesses far more than others.

A customer can replace a copywriting tool with ChatGPT. Replacing the system containing several years of accounting records is more painful. A company can generate a new logo almost for free. It still needs its domain to resolve tomorrow morning. AI can explain how an annual filing works, while the filing itself still has to be completed correctly and on time.

We can already see companies leaning into this. GoDaddy is adding AI tools inside an existing relationship covering domains, websites and commerce. LegalZoom is combining automation with human-assisted services and compliance products. Intuit is using AI inside accounting, payments and payroll rather than asking customers to abandon those workflows.

The safer boring businesses tend to own part of an ongoing process. AI can make that process cheaper to deliver, which can actually improve the owner's economics.

Is online bookkeeping and payroll still a great business?

Online bookkeeping and payroll remain two of the strongest boring businesses because every month creates another batch of transactions, reconciliations, reports and paychecks.

Intuit's latest annual numbers make the scale clear. QuickBooks Online Accounting revenue climbed from $3.38 billion to $4.12 billion and then to $5.05 billion over two years. That is almost $1.7 billion of additional annual accounting revenue in a category that was already enormous.

The surrounding products are growing too. Intuit's Online Services revenue reached $4.87 billion, up 16% in the latest fiscal year. The company said payments added $257 million of revenue growth and payroll another $266 million.

The more interesting figure is customer growth. QuickBooks' online paying customer count increased only 3%, while average revenue per customer rose 15%. Intuit generated much of its growth by becoming more valuable inside existing businesses rather than relying on a huge influx of new customers.

A small founder obviously cannot recreate QuickBooks. The opportunity is narrower: bookkeeping for dental groups, ecommerce sellers, property managers, agencies, restaurants or another customer type where the workflow repeats and the accounting quirks are predictable.

AI will automate more categorization, reconciliation and document handling. For a specialized bookkeeping company, that could mean serving more customers with the same staff. Customers still need usable books at the end of the month.

Do domains, hosting and website maintenance still make money?

Domains, hosting and website maintenance still make money today, although competing on cheap generic hosting alone looks increasingly unattractive.

GoDaddy's latest quarter is useful because this is about as mature as an internet business gets. The company reported $1.30 billion in quarterly revenue, $4.42 billion in annualized recurring revenue and $443.5 million in free cash flow. It had around 20.5 million customers and almost 82 million domains under management.

Revenue per customer also rose. GoDaddy's trailing ARPU increased from $230 to $250 in one year, while the customer count barely moved. Its old customer base is producing more money per account even though building a basic website has never been easier.

Domains have unusually sticky economics. Once a business has printed its URL on packaging, created email accounts, accumulated backlinks and trained customers to recognize the address, saving a few dollars by changing domains is pointless. Website infrastructure has similar inertia when DNS, email, databases, analytics and other services depend on it.

For a solo operator, the interesting business sits above commodity hosting. Managed WordPress for law firms, Shopify maintenance for a particular merchant category, website operations for restaurants or hosting plus security for local professional firms can combine recurring revenue with work customers dislike handling themselves.

The server space itself may get cheaper. The customer relationship can still be valuable.

Can online legal and compliance businesses survive AI?

Online legal and compliance businesses can survive AI very well when customers are paying to get something completed rather than merely paying to read legal information.

LegalZoom is already providing a useful live test. Its latest first-quarter revenue reached $206.8 million, up 13%, while subscription revenue increased 12% to $130.2 million. Subscription revenue represented almost two-thirds of quarterly sales.

The company attributed that growth partly to higher-value human-assisted services and improved compliance products. Average revenue per subscription unit also increased even though the number of subscription units was roughly flat.

That would be difficult to explain if cheap AI legal answers were eliminating the category.

The actual customer need is broader than drafting text. Businesses need registered agents, filings, annual reports, licenses, signatures, corporate records, deadlines and, sometimes, a human willing to take responsibility for the final step. AI can make each provider much more efficient without removing those obligations.

There are dozens of smaller versions of the same model: permit renewal, insurance documentation, privacy compliance, vendor credentialing, professional licensing, employment paperwork and industry-specific reporting.

Generic legal templates are becoming abundant. Making sure something gets filed correctly is still worth money.

Can niche online directories still make money?

Niche online directories can still make money, but the directory needs to help someone make a valuable decision rather than simply collect information that Google or an AI assistant can reproduce.

Zillow's rental business shows the difference at enormous scale. Rental revenue recently grew 31% year over year to $209 million in a single quarter, with multifamily revenue up 42%. Zillow had around 79,000 multifamily properties participating in its rentals ecosystem.

Calling Zillow a directory obviously understates the business, and that is precisely the point. Successful directories tend to evolve beyond a list. Zillow connects searchers with listings, property managers, agents, mortgage products and transaction tools.

A smaller directory can follow the same logic. Imagine a database of wedding venues in Provence. Names and addresses are easy to scrape. Current pricing, minimum spend, guest capacity, accommodation options, planner reviews, available dates and a direct inquiry system are much harder to replace.

The stronger directory also operates where one lead is valuable. A commercial property tenant, cosmetic-surgery patient, wedding client or industrial buyer can justify much more monetization than someone looking for a nice coffee shop.

We would be cautious about starting another giant collection of SEO pages. A narrow database that helps people complete an expensive decision is a different business.

Does online lead generation still work after Google's AI search changes?

Online lead generation still works, but depending on free Google traffic for nearly every lead has become a dangerous way to build it.

The demand side remains healthy. Zillow recently grew mortgage revenue 75% year over year as its purchase-loan origination volume almost doubled. NerdWallet still generated $197.3 million of quarterly revenue. Companies in mortgages, insurance, financial services, legal work, home services and other expensive categories continue paying for qualified customers because one conversion can be worth hundreds or thousands of dollars.

Distribution is where the economics are changing.

NerdWallet recently said continued pressure on organic-search traffic reduced its credit-card and small-business revenue. This has persisted for multiple quarters rather than appearing as a one-off fluctuation.

Independent research points in the same direction. Pew Research Center analyzed almost 69,000 Google searches from U.S. participants. When an AI summary appeared, users clicked a normal search result on 8% of visits. Without an AI summary, the click rate was 15%. Links directly inside AI summaries received clicks only 1% of the time.

For years, a boring lead-generation site could rank for “best personal loan,” send a visitor elsewhere and collect a fee. That model has become shakier.

Lead generation itself remains attractive when founders can build direct traffic, a recognizable brand, an email audience, useful calculators, proprietary data, repeat usage or paid acquisition with proven economics. Owning the demand is becoming much more important than owning 500 pages Google happens to rank.

Do niche job boards still make money?

Niche job boards can still make money because companies continue spending heavily on hiring, but a new board needs a reason to exist beyond reposting vacancies.

Recruit Holdings provides the clearest scale reference. Its HR Technology division, which includes Indeed and Glassdoor, produced roughly $9.7 billion in annual revenue in its latest fiscal year. Segment profit was about ¥550 billion, and management expects double-digit revenue growth in the current fiscal year.

Hiring remains an enormous commercial problem even after years of LinkedIn, Indeed and recruiting software.

The opportunity for a smaller operator comes from narrowing the market until the board actually solves filtering. Cybersecurity jobs requiring specific certifications, bilingual hospitality roles, climate-tech positions, veterinary jobs in one country or remote accounting positions can be more useful than another generic “remote jobs” website.

A good niche job business can gradually add salary data, candidate profiles, qualification checks, employer subscriptions, screening, alerts and recruiting services. At that point, companies pay because the site helps them hire somebody.

AI may improve matching considerably. That should favor job businesses with a real pool of candidates because better matching makes their inventory more useful.

Are online marketplaces still good businesses to start?

Online marketplaces can become fantastic businesses, but they are among the hardest boring online models to start from zero.

The mature economics are seductive. Recruit's hiring marketplaces generate billions. Zillow monetizes both sides of housing decisions. Upwork recently produced $191.7 million of quarterly revenue and a 33% adjusted EBITDA margin.

The problem appears when we look beneath the revenue.

Upwork's gross services volume recently fell 4% year over year, and active clients were down to 763,000. However, spending per active client reached a record $5,230, up 5%. Fiverr shows an even sharper version: annual active buyers fell 21.9% to 2.7 million, while spending per buyer increased 15.6% to $368.

The remaining customers are spending more, yet these platforms are finding it harder to keep a broad base of buyers growing. A new marketplace enters that environment without their brands, liquidity or installed supply.

The sensible small-founder version often starts as a service. Manually find ten buyers and ten suppliers in one narrow category. Facilitate the transactions yourself. Learn why matches fail. Only then does marketplace software become useful.

Starting with “we'll build a platform where X meets Y” still sounds easy. Getting both X and Y to show up at the same time remains brutally difficult.

Can a boring online service business be better than SaaS?

A boring online service can be a better small business than SaaS today because AI is letting service companies automate delivery without forcing customers to buy software.

Consider specialized bookkeeping, website maintenance, recruiting, CRM cleanup, ecommerce operations or compliance support. The customer wants books closed, a website maintained, a role filled or paperwork completed. Whether an employee, script or AI agent performed 60% of the hidden work is largely irrelevant.

That gives service owners an interesting advantage. They can adopt new automation internally while keeping the same offer externally.

The freelance-marketplace data also hints at a shift toward higher-value work. Fiverr recently lost more than one-fifth of its annual active buyers year over year, yet spending among remaining buyers rose 15.6%. Upwork's number of active clients has also declined, while spending per active client reached a record level. Upwork separately reported that AI-related work volume grew more than 22%, with AI strategy and consulting growing more than 50%.

Customers appear increasingly willing to concentrate spending on work that feels valuable while commodity tasks get automated or squeezed.

That makes “we make websites” a weak pitch. “We maintain and secure WordPress for independent accounting firms” is much clearer. The second business can build repeat processes, recurring revenue and expertise around one customer's problem without spending years building a software product first.

Is micro-SaaS still worth building now?

Micro-SaaS is still worth building when the software sits inside a recurring workflow; tiny standalone features are much easier for AI platforms to absorb.

Businesses clearly have not stopped paying for software. HubSpot generated $3.1 billion of revenue in its latest full year, up 19%, and customer count increased from roughly 248,000 to 289,000. Intuit's online ecosystem grew 19%. GoDaddy's recurring revenue increased even with a nearly flat customer base.

What customers expect from small software products has changed.

A tool whose entire job is summarizing PDFs, generating captions, rewriting emails or creating simple reports now competes with ChatGPT, Gemini, Claude and features bundled into larger software suites. Five years ago, a clever feature could sometimes support a whole subscription. These days, that is far harder.

The more resilient micro-SaaS products handle ugly, specific jobs. They synchronize data between two niche systems. They turn incoming documents into a format required by another platform. They track recurring inspections. They generate an industry-specific calculation. They manage approvals. They preserve an audit trail. They catch errors before a submission goes out.

A useful test is simple: imagine deleting the software on Monday morning. If someone at the customer's company suddenly has two hours of annoying work to do every week, there may be a business there.

Can newsletters, databases and paid information businesses still work?

Paid information businesses still work when the information helps customers make money, save money or avoid missing something expensive.

General information has become a rough business. AI can summarize public news, compare products and answer many basic research questions instantly. Search engines are also keeping more informational queries on their own pages.

Specialized information behaves differently.

A database tracking government tenders for construction contractors can reveal actual sales opportunities. A service monitoring permits can help property developers find projects earlier. A newsletter covering regulatory changes for one medical specialty can prevent costly mistakes. A database of private-company prices can help buyers negotiate. A feed tracking new restaurant openings can help food suppliers find prospects.

The value comes from finding, cleaning, organizing and monitoring information that affects a decision.

This also explains why some apparently tiny B2B publications can charge hundreds or thousands of dollars while mass-market newsletters struggle to charge $10 a month. Ten useful procurement leads can easily justify a four-figure annual subscription for the right company.

Generic AI-written newsletters face the opposite economics. Production is almost free for everyone, so publishing more words rarely creates much of an advantage.

Which boring online businesses are getting worse right now?

Generic affiliate sites, thin SEO directories, commodity freelance services, basic digital templates and tiny single-feature AI tools look much weaker today than boring businesses tied to recurring operations.

Search-dependent publishing faces the clearest pressure. Pew's research found almost a halving of normal-result click behavior when Google displayed an AI summary. NerdWallet, despite having a major consumer brand, has now reported organic-search pressure across multiple quarters.

Commodity freelancing also deserves caution. Fiverr's annual active buyer base recently dropped from about 3.4 million to 2.7 million in one year. Marketplace revenue fell from $74.7 million to $63.1 million in the latest quarter. Higher spending per remaining buyer softened the blow, but the direction of buyer volume is difficult to ignore.

Digital templates have a different problem: supply. AI can generate planners, résumé layouts, social-media packs, illustrations, ebooks and worksheets almost instantly. Good sellers can still make money through distribution, reputation or exceptional niche knowledge, but producing the asset itself offers much less protection than it used to.

Single-feature AI SaaS has the same weakness. Every improvement in general-purpose AI raises the baseline product customers receive elsewhere.

Business model	How it looks today	Main problem
Generic SEO affiliate site	Weak	Google sends fewer clicks on many informational searches
Thin niche directory	Weak	Easy for search and AI to reproduce
Generic template store	Weakening	Supply can grow almost without limit
Commodity freelance agency	Tougher	Automation and global price competition
Single-feature AI SaaS	Risky	Larger platforms can absorb the feature
Niche workflow SaaS	Stronger	Repeated operational job
Compliance service	Strong	Mandatory or costly-to-ignore work
High-intent lead generation	Strong	A qualified customer can be worth a lot

What boring online business would we actually start today?

For a small founder, we would start with a narrow B2B service, workflow tool, compliance product or high-intent lead business before building a generic content site, broad marketplace or another disposable AI feature.

The economics work better at small scale.

Ten companies paying $500 a month create $5,000 of monthly recurring revenue. Reaching the same number with a $5 consumer subscription requires 1,000 paying customers. That gap changes almost everything about acquisition, support and how quickly a founder can learn what customers really want.

Specialization makes the offer stronger too. “Bookkeeping for small businesses” enters an enormous competitive market. “Monthly bookkeeping and reporting for five-to-20-location dental groups” immediately narrows the software integrations, chart of accounts, reporting needs and customer-acquisition channels.

The same idea works elsewhere. Compliance monitoring for one regulated profession. Website maintenance for one type of local business. A tender database for one contractor category. CRM operations for one industry. Lead generation for one expensive service. Recruiting for one scarce profession. A small piece of software that automates one repetitive workflow inside any of those businesses.

We particularly like businesses that can begin manually and become more automated over time. AI then lowers fulfillment costs while the founder keeps learning from paying customers.

Starting with a painful recurring job is much safer than starting with a clever piece of technology and searching for somebody who needs it.

So what boring online businesses still make money?

The strongest boring online businesses today are bookkeeping and payroll, compliance services, domain and website operations, specialized B2B services, niche workflow SaaS, high-intent lead generation, focused recruiting businesses, transactional directories and paid data products.

The latest evidence makes that conclusion fairly strong. QuickBooks Online Accounting is growing above 20% despite its size. GoDaddy is producing more than $4 billion of recurring revenue from some of the oldest products on the web. LegalZoom's subscription business is still growing in the middle of an AI boom. Zillow's rental revenue recently jumped 31%. Recruit's Indeed-led HR technology business still generates close to $10 billion annually.

At the same time, several models that once looked like easy online money are clearly under more pressure. Organic-search dependence is hurting even established comparison companies. Fiverr has lost more than one-fifth of its active buyers in a year. AI has flooded the market with cheap content, designs, templates and basic software features.

The divide is becoming clearer.

Businesses keep spending on problems that return every week, month, quarter or year: getting paid, keeping records, filing paperwork, renewing infrastructure, maintaining systems, hiring people, finding customers and monitoring information that affects money.

A founder does not need to predict the next big consumer trend to build around those problems. In many cases, the attractive opportunity is sitting inside work companies already hate doing.

That is why boring businesses still make money. The technology used to solve them keeps changing, while the billable problem keeps coming back.