Why does it take several days to unsubscribe?
One tweet asked why unsubscribing from a newsletter can "take several days." Fasten your seatbelt, I’ll tell you an incredible story about how this is done in enterprise development ™ ...
There is one bank. Surely you heard about him, and if you live in the UK - with a 10% probability this is your bank. I worked there as a “consultant” on an excellent salary.
The bank sends out marketing letters. In the basement of each letter there is a small “unsubscribe” link. People sometimes click on these links.
Clicking on the link brings up one prehistoric web server that spins somewhere in the bank. Honestly, it took me three weeks just to find him.
This service sends a letter to the internal mailbox every time a link is clicked. This happens several hundred times a day.
Previously, these letters were sent to a specific employee, but five years ago he quit.
Now the message is forwarded to the distribution group. They could not change the address of the recipient, because it is hardcoded, but they did not find the source for the service. The service is written in Java 6.
The letters in the mailing group are checked by two employees of the bank's offshore center in Hyderabad (in India). They work hard and carry out their tasks awesomely , but plaque-fly, this work is unbearable.
I talked to them through a video conference and they had all the signs of an enterprise-post-traumatic syndrome. They fought with this nonsense for years and during this time nothing has changed.
When the letter arrives, they must execute an SQL script that determines whether the unsubscribed address belongs to the client of the bank (then the protocol is one) or not (then the other).
If the recipient is a client, they need to execute another SQL script that updates the client record in the preliminary ETL environment. All changes are checked at 16:00 London time by a separate team in Scotland. If the changes passed the test, they will be applied to the real database after another day at 16:00.
If the recipient is not a client, they add it to the Excel spreadsheet and send it to the marketing team in Swindon before leaving home.
The marketing team, by divining on the coffee grounds and other occult practices, determines whether the client is “potentially significant” (which, according to the internal rules, is “up to 48 hours”). If not, then the address is added to another table and sent back to India to execute another SQL query.
If marketing has defined the customer as “significant,” a letter like “do you really really want to unsubscribe?” Is sent to him manually. It looks like it is automatically generated, but in fact it is not.
If they answer “yes” (initially it was necessary to write “YES” in capital letters), then the team from Swindon sends the third table to India and the next script is solemnly executed there.
If I remember correctly, it takes an average of four business days . On average, about 700 people unsubscribe per day, of which 70% are "potentially significant."
By the way, these two Indians transferred to our development team and became PM for a system that replaced all this nonsense. These were the kindest, most responsive and hardworking people of those with whom I had the pleasure of working. It is thanks to them that this nightmare corporate grief-process so "worked" so smoothly all these years. They later moved to England and one of them now runs a department with 40+ employees.
Translator's note: owl on KDPV - Yoll .
There is one bank. Surely you heard about him, and if you live in the UK - with a 10% probability this is your bank. I worked there as a “consultant” on an excellent salary.
The bank sends out marketing letters. In the basement of each letter there is a small “unsubscribe” link. People sometimes click on these links.
Clicking on the link brings up one prehistoric web server that spins somewhere in the bank. Honestly, it took me three weeks just to find him.
This service sends a letter to the internal mailbox every time a link is clicked. This happens several hundred times a day.
Previously, these letters were sent to a specific employee, but five years ago he quit.
Now the message is forwarded to the distribution group. They could not change the address of the recipient, because it is hardcoded, but they did not find the source for the service. The service is written in Java 6.
The letters in the mailing group are checked by two employees of the bank's offshore center in Hyderabad (in India). They work hard and carry out their tasks awesomely , but plaque-fly, this work is unbearable.
I talked to them through a video conference and they had all the signs of an enterprise-post-traumatic syndrome. They fought with this nonsense for years and during this time nothing has changed.
When the letter arrives, they must execute an SQL script that determines whether the unsubscribed address belongs to the client of the bank (then the protocol is one) or not (then the other).
If the recipient is a client, they need to execute another SQL script that updates the client record in the preliminary ETL environment. All changes are checked at 16:00 London time by a separate team in Scotland. If the changes passed the test, they will be applied to the real database after another day at 16:00.
If the recipient is not a client, they add it to the Excel spreadsheet and send it to the marketing team in Swindon before leaving home.
The marketing team, by divining on the coffee grounds and other occult practices, determines whether the client is “potentially significant” (which, according to the internal rules, is “up to 48 hours”). If not, then the address is added to another table and sent back to India to execute another SQL query.
If marketing has defined the customer as “significant,” a letter like “do you really really want to unsubscribe?” Is sent to him manually. It looks like it is automatically generated, but in fact it is not.
If they answer “yes” (initially it was necessary to write “YES” in capital letters), then the team from Swindon sends the third table to India and the next script is solemnly executed there.
If I remember correctly, it takes an average of four business days . On average, about 700 people unsubscribe per day, of which 70% are "potentially significant."
By the way, these two Indians transferred to our development team and became PM for a system that replaced all this nonsense. These were the kindest, most responsive and hardworking people of those with whom I had the pleasure of working. It is thanks to them that this nightmare corporate grief-process so "worked" so smoothly all these years. They later moved to England and one of them now runs a department with 40+ employees.
Translator's note: owl on KDPV - Yoll .
All Articles