$508M a year to buy traffic that's disappearing

Today, I'm digging into Angi Inc. (ANGI).

They run the biggest home services marketplace in the US, the business formerly known as Angie's List and HomeAdvisor. Homeowners describe a project, Angi matches them with plumbers, electricians, and roofers, and the pros pay for the leads and advertising.

A few things that stood out to me:

• They've been purposely shrinking their top line. Revenue went from $1.36B in 2023 to $1.03B in 2025, and management chose that decline. They cut low-quality leads and third-party demand, and operating income swung from a $26M loss to a $65M profit.

• They spent $149M on buybacks in 2025 alone, 62% of today's entire market cap.

• The reason why the point above is even possible is the stock being down 64% in the last year, including a 35% one-day drop after Q1 2026 earnings showed demand softening again.

• I've been spending more time thinking about funnels and conversion rate in my business, so it's interesting to look at Angi's funnel: out of every 10 jobs homeowners submit, about 7 get done, and Angi's pros win only 2. Management says doubling that 2 is within reach. If so, this looks incredibly undervalued, but it's hard to outsmart Mr. Market.

• Their marketing engine is at risk. When homeowners ask ChatGPT/Claude/Gemini/Grok who to hire for a particular service, a lead marketplace like Angi gets ignored.

Angi is a prime example of AI disruption. On one hand, this business seems incredibly cheap, but I keep coming back to my own usage patterns. I am doing more and more of my product and service discovery via tools like ChatGPT or Claude.

If I were them, I would look to exchange some of their treasure trove of data with AI companies for a built-in plugin or some other mechanism for the workflow to flow through the pipes that Angi has already built.

My key takeaway here echoes something I've said plenty of times before: there's real power in owning your own distribution. Give your potential customers something genuinely useful in exchange for an email or phone numberso you earn the ability to reach them directly.

With that, I'll see you in a couple of days with another Nerd Out on Business report.

Nick

TL;DR

• Angi runs the largest US home services marketplace, matching homeowners with pros and charging the pros for leads and ads.

• Revenue is down roughly 24% over two years by design, while operating income went from a $26M loss in 2023 to $65M of profit in 2025.

• Q1 2026 rattled the market: the company swung back to an operating loss, demand softened, and the stock fell 35% in a day.

• The stock is down 64% over the past year while peer Yelp fell 21% and Frontdoor gained 30%; over 5 years ANGI has lost 95% of its value.

• The balance sheet is manageable ($227M net debt against $105M of 2025 operating cash flow), and buybacks have been aggressive.

• Operator lesson: cutting your worst revenue can mint profit while you shrink, but eventually you have to prove you can grow again.

The 30,000-Foot View

Angi sits between two groups who badly need each other and don't trust each other: homeowners who need work done and tradespeople who need customers. Homeowners use it free. Pros pay, either through advertising subscriptions (Angi Ads) or by buying individual project leads (Angi Leads). A smaller international segment runs similar marketplaces in Europe under brands like MyHammer and Travaux.

The economics look like software: 95% gross margins, because the product is a match and matches are cheap to serve. The hard part is everything around the match. Angi has to buy homeowner traffic (marketing was $508M in 2025, about half of revenue), retain pros, and police quality on both sides. For years the model leaned on volume: sell each lead to multiple pros and let them fight for the job. Pros hated it and churned. Since 2024 the company has run the opposite play, chasing win rates over volume.

Revenue mix (FY2025):

• United States: ~88%

• International: ~12%

Key Stats

• Market cap: $0.24B

• TTM revenue: $1.02B

• TTM gross margin: 95.7%

• 1Y total return: -64%

• Employees: ~2,300 (down from 3,800 at the end of 2023)

• Industry: Local services marketplace / internet content

Company History

• 1995: Angie Hicks starts Angie's List, a paid-membership directory of contractor reviews, going door to door in Columbus, Ohio for her first members.

• 1998: ServiceMagic, later HomeAdvisor, is founded in Denver.

• 2004: IAC acquires ServiceMagic; it's rebranded HomeAdvisor in 2012.

• 2011: Angie's List IPOs on Nasdaq.

• 2016: Angie's List drops its membership paywall to chase scale, giving away the trust asset members had paid for.

• 2017: IAC merges HomeAdvisor with Angie's List to form ANGI Homeservices.

• 2018: Acquires Handy, a fixed-price home services platform.

• 2021: Rebrands to Angi and pushes into fixed-price services, including buying a roofing company.

• 2022 to 2023: The fixed-price experiment unwinds. CEO Oisin Hanrahan departs, the roofing business is sold, and the focus returns to the core marketplace.

• 2024: Jeff Kip, who ran Angi's international business, becomes CEO and starts the lead-quality reset.

• March 2025: IAC spins Angi off as a fully independent company, paired with a 1-for-10 reverse stock split.

• January 2026: Angi cuts roughly 12% of its workforce as part of a rebuild into what Kip calls an "AI native" platform.

• May 2026: Q1 results show renewed revenue softness; the stock drops 35% in a day.

Show Me the Money

Standout financial features:

• The headline story: revenue fell 24% from 2023 to 2025 while operating margin climbed from -2.0% to +6.3%. Almost every cut was aimed at revenue that had a relatively higher cost of service than other lines of business

• Gross margin sits at an impressive 95%.

• publicly traded companies like this are under insane amounts of pressure to maintain their top line. That pressure is evident through their insane marketing expense of $508M.

• Headcount went from 3,800 to 2,300 in two years, a 39% cut against a 24% revenue decline. With that said, why the hell do they need 2,000+ employees?

• They repurchased $149M of stock in 2025 and the entire company now trades for $240M. Enterprise value is around $467M, roughly 5x TTM EBITDA of about $88M.

• Q1 2026 is the wrinkle: an operating loss of $9.5M against a $20.0M profit a year earlier. Restructuring explains $14.9M of the swing, with $27.7M across the last two quarters tied to the workforce cut and the AI rebuild. Strip it out and the quarter is roughly break-even. Either way, TTM operating profit has fallen to $36M.

Financial Data

All four columns come from Angi's as-reported filings, with restructuring inside operating profit.

Stock Performance

The 10-year figure includes the pre-merger Angie's List era, so treat it as directional. Either way, a dollar invested five years ago is a nickel today.

1-year head-to-head:

Frontdoor sells home warranty plans into the same housing backdrop and gained 30%, so the market's issue is with Angi's lead-generation model specifically. The profit turn bought no benefit of the doubt.

The N.O.O.B. Nine — Competitive Powers

The Nerd Out on Business Nine is made up of Hamliton Helmer's famous "7 Powers" of competitive advantage (Scale Economies, Network Economies, Counter-Positioning, Switching Costs, Branding, Cornered Resource, and Process Power) combined with two of my own (Data Flywheel and Distribution Advantage).

Average Score: 2.4/5 - A weak moat: real scale and data, undermined by multi-homing and rented distribution. One of the lowest scores I've handed out.

Memorable Marketing

Angi's marketing history is really a story about trust: building it, giving it away, renaming it, and now trying to buy it back.

Notable campaigns and tactics:

• The membership paywall (1995 to 2016): Angie's List reviews were credible because members paid to contribute. The paywall was the trust mechanism, and "reviews you can trust" radio ads made Angie a household name.

• The Angi rebrand (2021): A national TV push that unified three brands and walked away from decades of "Angie's List" search equity. Marketers still cite it as a cautionary tale.

• "Anyone Can Be Anything on the Internet" (2026): TV campaign contrasting the unvetted internet with Angi's screened pros. Smart positioning against sketchy contractors and AI slop alike.

• Instant booking push (2024 to 2025): Creative built around same-day availability, selling a booked appointment because certainty converts better than a list of options.

Tactical takeaways:

1. If your trust mechanism (a paywall, a vetting step, a guarantee) is what customers actually buy, treat removing it as a product decision with revenue consequences.

2. Before a rebrand, price the search equity you're abandoning. It's often the most valuable thing you own.

3. Sell the outcome state ("it's booked for Tuesday") over the option set.

4. Aim campaigns at the category's core anxiety. In home services that's hiring a stranger.

AI Uses & Opportunities

Current exposure:

• An LLM-powered "AI Helper" walks homeowners through describing their project. Management says about half now touch it, and better-described jobs mean higher-intent leads for pros.

• The back-end is being rebuilt into an AI-native platform, and the January 2026 workforce cut was framed as part of that shift.

Future opportunities:

• An AI booking agent that answers calls and schedules jobs for pros, who are famously bad at answering the phone while on a roof.

• A pro-side sales coach: record the in-home visit, prompt the pro, and generate the quote before they're back in the truck.

• Photo and video scoping that produces instant estimates, collapsing the quote cycle from days to minutes.

• The flip side is existential: if AI assistants book jobs directly, Angi has to become the pipes those agents use or get bypassed.

Bumps in the Road

• Revenue is still falling, and TTM operating profit has slid from $65M to $36M. The margin story only works as a prelude to stabilization, and Q1 2026 pushed that proof point out again.

• Distribution is rented. Paid search plus SEO built this company, and Google's AI answers are shrinking the real estate both depend on.

• Years of aggressive lead selling left scar tissue with pros, and trust rebuilt slowly is easily lost to a Thumbtack ad promising better economics.

• A soft home-improvement cycle means fewer discretionary projects, pressuring lead volume and pros' willingness to pay.

• A $240M market cap against $471M of debt, plus a post-spin shareholder base still sorting itself out, keeps the stock violent in both directions.

Your Swipe File

• Score each revenue stream by the churn it causes and the cost to replace it. Angi found a quarter of its revenue was likely making the business worse so they axed it.

• Measure your funnel. Angi's "7 jobs done, 2 won by our pros" framing is an interesting way to describe the business and the upside in one simple phrase.

• Double-sided marketplaces are incredibly difficult, as we talked about a lot in these reports. Not many businesses have a trust mechanism as a moat, but Angi does. That needs to be guarded with the utmost priority

• Owned demand compounds while rented demand re-prices every hour (via ad auctions). Build an asset (brand, community, direct relationships) that Google/Meta can't repossess.