Email users now regularly use the email platform for mission critical (time sensitive and reliable) operations due to its robustness and ubiquity. However email latencies is something they must watch out for.
What are we doing today with email? Why has it become so mission critical?
Besides the original purpose of being a carrier of documents and alerts (which is now official), the email platform has been used as a digital carrier service to connect business applications, trigger business functions, program work flows, support customers etc. – essentially it’s being used to connect transaction systems.
How did this happen?
- Reliability of the email platform -that the message will simply reach the recipient or you (sender) will be informed about the non delivery of the message.
- Unique Identification – The email address soon became the unique address/identifier of any individual on the planet.
- Messages meant for the User – Unlike the Timelines on social media apps, the Inbox gets the attention of an individual because it carries the messages specifically intended for him (even when he is marked in CC).
This made the email alert a natural choice for bringing things to the user’s attention.
Developers and system designers further innovated and created “non-human” email ids (popularly called system ids), which have plugs, hooks to intercept inbound mail, vet the senders of these mail for security checks and take some coded action, making it easy to program workflows and business applications, by loading the email with code snippets or instructions for delivery and action across business systems.
Some examples of the mission critical use of Email
Developers and system designers further innovated and created “non-human” email ids (popularly called system ids), which have plugs, hooks to intercept inbound mail, vet the senders of these mail for security checks and take some coded action, making it easy to program workflows and business applications, by loading the email with code snippets or instructions for delivery and action across business systems.Publishing email ids to receive lead or a purchase order:
Many small business owners publish their email id on their website since it’s the easiest and quickest way to get started with a web site and enable the outside world to connect with them. Any delay or drop in delivery would directly translate into a business loss.Seeking Approvals:
Many a times, operators are waiting on site and expecting an email approval from their managers to continue their work. Any delay here will directly translate to productivity loss.Sending one time tokens during a registration process:
Many sign up processes, involve verifying the authenticity of the user, by sending a message to his email id, midway through the process, with a unique token or link, which is valid for a very brief period. The user is expected to switch to his email system, retrieve this token, enter into the registration form to continue. A delay or drop here will ensure that the user can never sign up.
Requesting for information from a system (automation):
Companies may publish an email id where their customers can send mails from their registered email ids with predefined subjects such as BAL for getting a response with the account balance, MODIFY ADDRESS and provide the new address in the body of the mail to update their address in the vendors systems, etc. Essentially email as an automation tool, much like phone banking, which primarily used sms and now uses Apps.
Intercepting specific mail and retrieving attachments from them:
A popular use case with one of our customers, FFreedom, is to intercept mail, which contain daily financial transaction reports into a central attachment vault (FTP server) for further processing by another application. Here the Email engine is glue between two disjoint systems to complete a work flow.
What does all this mean for requirements from the Email service?
A dial tone reliability, low latencies (less than a minute) and 100% accuracy on email delivery, without security breaches are now default expectations.
Is there a flip side to this?
The email delivery system is a complex system dependant on a lot of elements such as – the network quality/latencies, client software behaviour, security systems, spam detection systems, filter configurations, etc. The whole flow or transaction from sender to recipient is not in the control of a single system.
Email providers normally guarantee mail delivery latencies for the flow within their system. For example: If an inbound mail enters the mail system for delivery to a local user, they will guarantee that it will be delivered within ‘20’ seconds for example. However this doesn’t account for any delays, due to which the mail was held up in the sender’s mail system itself.
Similarly if a user sends a mail (outbound), they guarantee that it will reach the outbound queue within ‘10’ seconds (example). Here if the recipient system is unavailable, the message will be held in the queue and the system will retry periodically to deliver it. However once an email is delivered into the recipient mail system, there is no visibility on whether the email was blocked by the recipient’s security system, whether it was marked as a spam, whether the user actually received it (in good time).
In such a vastly distributed system, which is not controlled at any single point, it is hard to guarantee mail delivery times “end to end”.
However since each email provider makes a promise of delivery times, it all adds up and overall the global email system performs to the expectations and email has successfully become a mission critical platform.
The email is a solid platform, good enough to carry mission critical work. But while choosing a mail service provider, users must insist more on the performance parameters than bells and whistles.