It’s classified - Thoughts on REA and CAR

Digital Classifieds are a subset of Tech businesses caught up in the recent sell off.

They have some key differences to the rest and some differences to each other.

As a group they damned the once famous ‘rivers of gold’ that funded the newspapers.

As the internet developed, these businesses challenged, disrupted and took these rivers of gold from the incumbents who were too slow to see this happening to mount a successful challenge.

Arguably Domain was a semi-successful attempt to somewhat counter this.

REA & CAR have become dominant in Australia and hugely profitable over a long period of time to the point where they are now the incumbents and seemingly ripe for disruption from the biggest tech revolution since the internet that spawned them.

Thinking on paper here

I’ve followed these businesses from a safe distance for a while and owned them periodically but am far from an expert.

They seem to share a few things in common – dominant in their niche, B2C model, maturing with an eye on M&A for growth.

I probably understand REA a bit better than CAR.

98%+ of homes are sold through a Real Estate (RE) agent who encourages the vendor to also advertise on REA (and ideally Domain – why not?) at the vendors expense – helping the RE agent to get the sale, nice guys.

Most would use REA as it’s the biggest - has the most buyers so attracts the most sellers and so on – aka a network effect in full flight.

Having eaten the newspapers lunch, REA has been encroaching on the RE industry for a while by upping their prices steadily and trying to capture more of the agents work and fees.

Still RE agents don’t seem to be a dying breed.

The hunter becomes the hunted?

In the spirit of Bezos’ “your margin is my opportunity”, AI is allegedly coming for REA the way the internet came for newspapers ad revenues.

If REA was to be disrupted by growing AI capabilities (including AI agents) what would this look like?

In theory an AI agent could scrape all the listing on all RE agents websites (pretty easy) and aggregate these.

AI Agent would need to commonise all the data (like REA have done) from agents sites to allow filtering, etc, etc.

To make this work well, RE agents could collaborate to make their listing details compliant with what the AI Agents are looking for.

RE agents could even band together as a consortium to own / build an AI agent to replace REA.

This is becoming increasingly cheap to build and would be relatively cheap for vendors vs the expensive REA prices currently on offer.

This would give agents collectively more control over their business / industry that REA has been chipping away at.

Could they / would they actually do this? If RE agents see the threat from REA as significant enough to them they could perceive REA as their longer term competition more than each other - "The enemy of my enemy is my friend".

REA would likely see this as an act of war and there are weapons available to REA (mainly investing in their own AI capability), but they would not be in a position to disrupt their own business model in order to win it.

Key differences between REA & CAR

While 98%+ of homes are sold through a RE agent, 60%+ of second hand cars are sold privately, so these will only be listed on CAR or a similar local website for an AI Agent to scrape.

In theory dealerships could also set up a consortium to fund / buy an AI agent to challenge CAR but I don’t see them having the same incentive as RE agents.

More importantly, there isn’t the latent data set to be scraped off the internet and aggregated as 60%+ of cars are not digitally listed until they get paid ads on CAR (or a smaller, local site).

Moats

Network effect is the big one for REA and this has kept Domain from becoming a serious threat to margins so far. I expect this is a big incumbency benefit for REA - having a large ‘share of mind’ as they have built a strong brand in the Australian Market.

Buyers are unlikely to use AI agents for a house purchase any time soon I expect, especially as the cost is with the vendors. Sellers need to go where the buyers are to maximise their market and potential price.

This could be their best defence, the question is how strong is it in the face of emerging AI capabilities (whatever these might look like in a couple of years) and how long can it hold. I need to think about this a bit more.

Other intangible benefits (not everything that counts can be counted) like having developed this market over decades, built up a large historical data library, understanding the industry probably better than anyone, etc, etc. -

Scale benefits are definitely there but this is something AI agents can in theory undermine pretty quickly, especially as these develop more in the direction of AGI.

Switching costs are likely low as paying customers / vendors are not frequent / recurring users as house sales are for most people a rare event.

Some scenarios

If a credible AI agent threat does start to emerge, this could be detrimental to REA’s margins in several ways.

There are a couple of broad scenarios (and plenty in between) for how AI might unfold in the coming years

Game changer - AI Agents, etc proliferate and take over jobs, industries, etc, etc.

Fizzer - AI is way over-hyped and the anticipated revolution doesn’t show up in such a meaningful way.

If the game changer scenario emerges, REA might be reluctant to push through price increases as this would make alternatives more attractive to spin up or for sellers to try. They would need to mount a defence and quickly – likely via expensive AI investment / over-investment, also a marketing / disinformation campaign.

In the Fizzer scenario, REA (and many other) will still likely spend a lot of Capex on AI capabilities to keep up with / fend off competition (real or imagined) which may or may not generate a return for them.

Hard to believe AI will be a fizzer for long.

In any case, rising Capex costs and reluctance to hike prices will likely be a source of margin pressure for a while.

What have I missed or got wrong?

Plenty of people on here understand these businesses and likely AI impacts on them and more generally better than me.

What say you?