To quote the author:

> Then automate it. If you could automate the collection of the data in the first place then you definitely can automate the rest of the life cycle. There is no technical hurdle companies won’t jump through if it gets them juicy bits of data but as soon as the data needs to be removed we’re suddenly back in the stone age and some artisan with a chisel and hammer will have to jump into action to delete the records and this will take decades for even a small website. Such arguments are not made in good faith and in general make the person making them look pretty silly after all nobody ever complained about collecting data, in fact there are whole armies of programmers working hard to scrape data from public websites which is a lot more work than properly dealing with the life cycle of that data after it has been collected. So yes, it is a burden, no, the burden isn’t huge unless you expressly make it so but that’s your problem.