Brain-Dead Xero Machine Learning, programmers misdirected, slow, incompetent. Part 1 of 5

Machine Learning can quickly and easily save lots of accounting and bookkeeping time (while minimizing errors) IF YOU AVOID BRAIN DEAD XERO MACHINE LEARNING.

I was once #3 in U.S. Xero clients, but the long-delayed Brain-Dead Xero Machine Learning was a small part of why I dropped Xero. Despite wildly misleading Xero Machine Learning publicity, here is why Xero will have Brain-Dead Machine Learning for years.
1. Horribly Brain-Dead Xero Machine Learning, programmers misdirected, slow, incompetent.)
2. Brain-Dead Xero Machine Learning will persist, Xero culture denies problems instead of fixing them
3. Brain-Dead Xero Machine Learning ridiculously incomplete
4. Brain-Dead Xero Machine Learning by design
5. Brain-Dead Xero Machine Learning because of rejected add-on
1. Brain Dead Xero Machine Learning, programmers misdirected, slow and incompetent
This is a big reason why Xero programmers will not soon fix the Brain Dead Xero Machine Learning. Xero programmers showed how misdirected, slow and incompetent they were in creating payroll about five years ago. They were misdirected in creating it from scratch, not using an industry standard interface link to a payroll product or buying a payroll company. The first release remained terribly buggy for many months. It lacked essential tax returns and withholding for a long time, for all but a few states. There were ridiculous delays between when payroll amounts were deducted from company bank accounts and when employees and taxing authorities got paid, which apparently never changed. Even in states like Florida, with no income tax, Xero had to refund state taxes and credit Xero payments for many months. Users then had to manually create state returns and directly pay tax.
Xero personnel actually begged key users to keep testing payroll, while deriding payroll programmers. However, it took a long time before we heard of firings. It also took years for Xero to buy a payroll company and add states for their limited, slow, and mispriced payroll. Mispricing related to a large change in initial pricing, which came before users could print year-end payroll returns. It also relates to the abnormal bundling of payroll and accounting charges. Xero still bases charges on the number of employees, not the number of transactions, in overly narrow brackets. Except for a $9 price (for a useless 20 bank transactions), Xero is $30/month for 5 employees or $70/month for 10 employees, with unlimited entries. A day with an extra employee makes charges jump for 5 employees for a month. I never of other bookkeeping – general ledger – financial statement programs that charged based on employees, in brackets, even if they were no bookkeeping users.
Xero programmers also were horribly misdirected, slow and incompetent in creating a good Xero Help Search. That is why I wrote (ALL CAPS mine), “Replace Help Search” in the Xero Business Community. “Xero has a TERRIBLE help search. If you search for several keywords, it gives you every page that has ANY of them. This always wastes time. It often makes it difficult to impossible to find the help you need, resulting in many extra Community posts and support emails. Changing this should be a top priority. Until then, add to a Google search, instead of using the terrible Xero help search.” The Official Xero Reply, by its leading Community Manager [wife of co-founder and CTO [lead programmer], included, “Fair comment – now that there’s so much content… we need to take another look at the search. FYI before we wrote our own search we did use Google Site Search but it had problems deciphering the 5 different versions of the content that exists behind the scenes… so the overload and duplication returned wasn’t ideal.”
This shows that the Xero programmers did not properly arrange help, so Google site search could work with it. Of course, simply automatically adding “Xero Business Community” (or other search additions) might fix many searches. This comment also was not updated, though made many years ago. We cannot know how old it is because Xero retroactively deleted dates from such comments, rather than show how many years it often takes to update feature requests. Search date parsing also does not let us determine this, though it means it is unlikely that Xero ever fixed this problem.”

search and replace (also known as “Find and Recode”). Search and Replace has 2,054 posts in Xero Business Community Feature Requests and Find and Recode has 4,749 more pots

Xero CTO (CTO@Xero) wrote a current Official Xero Reply to one of many related threads endlessly requesting this), “… released Xero Search (2015)… going to be a journey to get it right… tuning our indexing and our search algorithm… currently indexing amount fields… If you search for “100” what would you expect that to find (totals == 100? amount due == 100?)? We can do totals easily… derived amounts (like paid or due) are not as easy (and won’t happen quickly). Also investigating line item description search… we turned off because it seriously bloated our index size.” This means Xero has big problems with the description search, which is critical to Brain Dead Xero Machine Learning.”

Xero compounded this by needing five or so years to deliver (or not deliver) many other near-unanimous requests, with around 500 user posts each. Xero programmers are SO misdirected, slow and incompetent that will not soon fix the Brain Dead Xero Machine Learning.”

For daily updates on this, subscribe to the QuickBooks-Blog.”

About the Author ()

The fastest and easiest automatic accounting - accurate, economical, safe, and reliable, with low taxes.

Comments are closed.