Unpredictable Market Forces — Learnings from the AnyClass Failure.

There are many posts dissecting the reasons for startup failure but you don’t see many that talk about the impacts of competitive rivalry. As such, we thought we’d share some lessons from a recent failure of ours, AnyClass.

GETTING STARTED

In 2014, we launched AnyClass, an aggregator of boutique fitness classes. A local version of US-based ClassPass, it offered customers the ability to attend many classes at participating gyms and studios for a fixed monthly fee, without having to become a member at any particular studio. To enable this, we would pay studios for each class one of our members took.

A natural question might be “Wait…the model is to pay studios to send people to them? Shouldn’t studios be paying you as a lead-gen channel?” That would be a good question and highlights one of the core challenges we’d identified for the model — the less a customer uses the service, the more profit we make.

That inherent misalignment was something we thought we could fix over time. Local studios were surprisingly well informed about the model though, so we decided to launch despite the misalignment to quickly build the marketplace. Once we had sufficient market share we could fend off the incoming competition we suspected would soon appear, then look to make the model changes that aligned AnyClass and its customers health outcomes.

HERE THEY COME

Our hunch that we needed to move quickly in anticipation of inbound competitors turned out to be correct. Within six months of starting AnyClass there were 8–10 competitors, all focused on the same supplier side of the market.

Most were small players but a few were very well resourced. These included:

- BodyPass: Owned by Fairfax, the second-largest print and digital media company in Australia. The original BodyPass business was actually acquired by Fairfax in what was an unusual transaction for the media giant. Fairfax had traditionally acquired later stage startups that they then threw their marketing and distribution weight behind to accelerate growth. BodyPass, however, was run by first-time founders and acquired when it was still pre-revenue. Post acquisition the task of growing the startup was handed over to internal Fairfax teams brought in from other parts of the corporate entity.

- KFit: Sequoia-backed, KFit focused on the South East Asian market, and expanded into Australia after raising a US$15M round.

- ClassPass: The original. By the time they arrived in Australia ClassPass had already raised US$55M, were experienced in entering new markets rapidly, and would go on to raise an additional $30M later in the year.

THINGS START TO GO WRONG

Despite the strong competition, AnyClass initially did well. Our key numbers on both sides of the marketplace were strong and feedback from studios (the critical side of the marketplace for this model) was overwhelmingly positive.

Then something strange happened.

Fairfax started reducing the price of memberships and increasing payments to studios. However, lessons from ClassPass in the US were that as customers became more familiar with this type of service they attended more classes, meaning expenses increased as the customer base matured.

This was a big enough issue that ClassPass was actively seeking to change its model in its major markets, but that warning went unheeded in Australia. Instead, the increased pressure on both revenue and costs as a result of the Fairfax strategy accelerated their, and everyone who matched their pricing, movement towards unprofitability.

We understood the rationale at the time — it appeared to be a classic loss-leader strategy to capture market share — but with at least two other competitors with deep pockets willing to go with Fairfax, price competition took over and led to its natural conclusion … even at bare minimum usage, the unit economics became unsustainable.

Eventually ClassPass, who had been working on a more sustainable model in their core markets, pulled out of Australia (two years later, they’ve now returned with a refined model). KFit moved to become a booking marketplace for fitness classes. BodyPass went through several management changes, until eventually Fairfax pulled the plug.

Our focus on providing a customer-led (on both sides) service, and our unwillingness to compete on price, should have meant that AnyClass was well positioned to clean up the mess, but suppliers had been conditioned to expect high payments per attendee, and consumers to expect low prices. Rather than wait for the market to reset, we pulled the plug on AnyClass too.

THE END

There were lots of lessons from the AnyClass experience. A few stood out for us.

- In two-sided marketplaces, when you’ve identified the primary target side (in our case studio owners) there are often a small number of targets who have a disproportionate effect on acquisition of the alternative side. In the AnyClass example our ability to bring on a handful of well known yoga studios drove much more customer acquisition than the mountains of smaller studios very few had heard of. To be clear, the perception of variety that the long tail provided was important, but big names were even more so.

- Not all players will act in a way that you perceive to be rational. Organisations, where there is an imbalance between the rewards of success and the risks of failure, are more likely to make decisions that would otherwise be irrational.

- If you’re operating in Australia against international competitors, this is often not an important market for them. They’re rarely subject to local capital-raising restraints, and sometimes have decisions made by people at head office without regard for local conditions.

- Once market expectations have been set, especially around payments or pricing, it’s incredibly difficult to change in the short term.

- Where there is no alignment between customer (in this case studio) and company, you do not have a sustainable business. Studios saw this model as a quick way to make what were, in some instances, large gains from unused capacity. When things changed they were happy to acknowledge the free ride was over and move back to business as usual.