Home Services AI, ChatGPT 8%, SMB Lead Laggards

Private Equity, Home Services & AI

Home Services are growing faster than other local business categories, according to a recent Yelp economic report. Home Services – with 62 subcategories on Yelp – saw an 11% increase in new business openings in 2024. It leads all other categories with 217,000 new openings in 2024. The data are obviously not comprehensive but directionally show a very healthy and growing market.

Private equity has become increasingly attracted to Home Services over the past decade and dozens of local services brands are now private-equity owned. Yet it remains a fragmented market with hundreds of thousands of independent local businesses. Investors see still more opportunities for consolidation and greater efficiencies. Another draw is the perception that home services are recession-resistant.

Since 2014, many billions of dollars have flowed into local acquisitions, with a focus on HVAC, Plumbing, Electrical Services, Pest Control, Home Repair and Maintenance, Landscaping and several others. In fact, the Wall Street Journal just ran an upbeat article about private equity's impact on the lives of these business owners and their companies, which is not always what happens when private equity gets involved.

One of the beneficial things that private equity does, however, is upgrade technology and software tools. This means SaaS and, now, AI. Even though it's not private equity owned we got a strong sense of what this will increasingly look like when we spoke with Daisy, a new startup (and consolidator) in the "smart home installation" segment. The company is integrating AI deeply into the operational core of the business.

The SMB market is generally outpacing enterprises in AI adoption. That's likely to further accelerate in private equity owned Home Services operations (see, e.g., Neighborly). Unlike some legacy SaaS applications, AI-simplified software should be an easier "sell" to local operators, given the general AI-enthusiasm we've seen in our survey work. In the Construction, Home Services & Real Estate segments in particular we see lower current adoption but greater enthusiasm (85% positive) than in the full sample.

Franchise brands will have a similar impact on their local operators, by recommending specific AI applications and promoting adoption. That will extend to technical training as well. AI is emerging as a powerful tool for these local home service businesses, not only enhancing operations but also accelerating training and improving overall service delivery.

News & Noteworthy

• Evercore survey (n=1.3K US adults): 8% using ChatGPT as "go to" search engine, up from 1% just months earlier.
• SearchGPT coming to ChatGPT by the end of 2025; OpenAI losses could reach $14 billion by 2026.
• Google's new AI Organized Pages appear to mostly shut out organic search results in favor of GBP-filled local carousels.
• Why OpenAI’s deal with Hearst should scare Google: Hearst's publications will appear in ChatGPT output and search results.
• Reddit using AI to target user conversations for advertising.
• As pressure mounts to move beyond AI experimentation, enterprises are struggling with data issues, culture issues, bureaucracy and AI limitations.
• Amazon introducing AI shopping guides for 100 product categories.
• Apple researchers: no formal reasoning in LLMs; it's pattern matching, which is a problem for math in particular.

Data & Discovery

Lead response times for SMBs: Local businesses are notoriously bad at responding to inbound inquiries, which costs them business – though they may not fully understand that. There are numerous studies supporting this idea. A more recent one comes from Leadferno, which conducted a kind of "mystery shopper" lead-gen study earlier this year.

Leadferno contacted 225 US SMBs through forms on their websites in several verticals. Each businesses was chosen because they ranked well on Google and had at least 25 reviews (average 138). Leadferno posed as a would-be customer, submitting inquiries with clear buying signals: requests for appointments, estimates, next steps. What the company found however was poor response rates and times:

• 42.6% of contact form submissions received no response at all.
• The average first human reply took 17 hours and 49 minutes. Just 10.6% responded within an hour.
• 65.9% of leads that received a response only got one follow-up, reducing the chances of closing.
• And only 15.6% of businesses sent an auto-reply confirming the submission, creating doubt about whether the inquiry was received.

AI and automation are the obvious answer. AIs can automate responses and help close even before there's a human conversation. At a minimum prospects will feel acknowledged and that the business is taking their inquiry seriously. Often negative reviews on Google and Yelp reference poor engagement by businesses ("they never got back to me"). AI tools that can help are already in the market. It's simple: faster response times mean more customers.

Decline in job creation: Last year the Bureau of Labor Statistics (BLS) published an article about long-term SMB job creation statistics and revealed a paradox of sorts. The number of new businesses has increased roughly annually since the 1990s but the number of people employed per firm has been declining. The article says that new companies today employ roughly half as many people as those formed in the 1990s. The explanations are multi-faceted but boil down to digital technology and outsourcing/freelancing.

AI will only continue this trend. Business owners will use the efficiencies of AI to avoid hiring more employees, who may be in short supply anyway. Both SMBs and smaller agencies see reduced hiring needs as an AI benefit. In both surveys a roughly equivalent percentage of respondents cited it.

Dialog: ICYMI

Recent Dialog content and data you might have missed:

• Larger Agencies Seeing Revenue Growth from AI Adoption – Small and larger agencies have embraced AI; however larger firms are seeing a wider array of benefits, including revenue gains.
• What We've Learned So Far – Observations, lessons and recommendations from our research in 1H 2024.

Funny | Weird | Other

• AI assistants are ratting you out for badmouthing your coworkers – AI assistants transcribing meetings are exposing confidential and private information.
• Anthropic lays out mostly utopian view of AI societal impact, even as it rethinks its own hiring plans because of AI.
• AI pioneer Yann LeCun disputes that AI is ready to surpass human intelligence, calling BS on capabilities and doomsday scenarios.