A new CEO has been announced today. Their announcement included the following for the outgoing CEO.

Mr Freudenstein acknowledged Mr Ahmad’s tenure at Appen:

“Armughanhas been CEO during a period of great change at Appen. Having overseen a new strategic direction and re-sizing of the business, Armughan has decided to pursue new challenges. We wish him well for the future.”

I like the marketing spin. "re-sizing" in reality meant complete destruction of shareholder value .

The market has been brutal with APX today - price down 25% to $6.50

Here are key comments.

• Revenue was up only 8% but 32% in second half.
• Non-ad related projects now represent 76% of total revenue from Global customers, up from 71% in FY20.New Markets revenue grew 21% to $102.5 million, driven mainly by growth in China.  
• Ad-related revenue increased in second half by 18%. This was the big ticket item in previous report where it was down. Management said at the time it was because of other priorities in big tech companies. Market conjecture was that the companies had self-training models.
• The big tech companies still account for 75% of total revenue - so there is still risk there.
• Underlying NPAT was down 10.4% to $40.6 million, due to the increased amortisation associated with investment in product development.
• Appen had no debt and a cash balance of $47.9 million as of 31 December 2021. 
• They won't provide yearly short-term EBITDA guidance. Focused on growing doubling revenue and lifting EBITDA margin to 20% by 2026 (5 years)
• they have 11 autonomous vehicles customers for mage, video, LiDAR, speech

Importantly there wasn't any evidence that the AI market is dwindling - as many have suggested with AI models that train themselves. There is certainly a change in the market though with many more requirements for AI and a broadening of the customer base. It's unrealistic to expect that it will return to previous growth but there is still plenty of opportunity.

With the drop in price market cap @$6.50 is $800M. So a P/E of 20. Not very demanding. As much as it pains me to say, maybe this is reasonable given where the business is. Maybe it's now a mature business and not a growth stock any more.

(disclosure - I hold shares)

I've seen a two threads of comment about the drop.

1. The negativety of the presentation at the Macquarie conference
2. Facebook's breakthrough for self-supervised learning

From the Macquarie presentation transcript
- there were headwinds for big customers last year as they adapted to COVID and 'regulatory pressures such as anti-trust and data privacy'.
- "The competitive environment for relevance is unchanged with us and Lionbridge AI the key providers."
- 'There’s no systemic change to the demand for relevance data. Our relevance team delivered 90% of our revenue in 2020.'
- "There’s no change in the need for training data. More companies are investing in AI and they all need training data. The high growth in the number of customers we’re winning, including in China, supports that."
- 'We don’t see unusual pressure on pricing.'

My take
- Appen is a project based business. Their customers had to shift priorities last year. That's cyclical not fundamental. More broadly, there is going to be a lot more investment in AI from all sorts of organisations. Governments for sure are slow to adopt but will use it.
- 90% of their business is unchanged with no unusual pressure on pricing
- Facebook's breakthough was for image recognition - ie less than 10% of Appen's revenue. From their last annual report 'Relevant' delivered $538M revenue. 'Speech and Image' delivered $61M.
- China is growing 60% quater on quarter. They have an indepdent setuo of their systems installed in China and another for US Government.  

You'll see from my portfolio that I unfortunately bought some at $22 and then some at $15. I bought some more today in my real portfolio.