Ann: Macquarie Australia Conference, page-9

I had a read and first thought was nothing too new that differ from what Mark spoke about during previous earnings call. Competition maturing outside of relevance (i.e. speech, image etc) was previously flagged already and is something that is much easier to enter and compete compared to 'relevance'. Relevance is Appen's competitive advantage and main revenue generator. Competition in that space remain unchanged if you believe in management. Regarding 'resource allocation' shifts of Appen's customers, the way I interpreted it is that:

- Development of new products goes through initial small-scale internal pilot of such AI models. Therefore you wouldn't want to engage externals like Appen to do that until that model has been tested and proven to be worthy of progressing to larger scale real world implementation. It's just like any other Agile development process to test/experiment through pilot with small amount of end users first before pouring decent money in it to scale it up (which is where Appen will come in).
- This means relatively less demand for Appen's data on large proven projects whilst their customers work on spinning up new products internally first.

One quote that I am unsure about is "Our major customers are also responding to data privacy and anti-trust concerns, as well as heightened competition amongst each other. These issues are also driving new AI product developments, ones that rely less on personal data and make them more independent."

My main question/concern surround the above would be does that mean once these new products are ready for scale-up, there will be relatively less demand for Appen's relevance data?