Today’s is the third in a series of posts on contractual terms that clients want to try to get, but usually will be unable to get due to the harsh impact of reality. Thus far, we have considered [acp author=”Mickey Chandler” id=”Chandler-02″ media=”blog” url=”http://www.spamtacular.com/2010/05/11/asking-for-the-impossible-slas/” title=”Asking for the impossible: SLAs” year=”2010″ month=”May” day=”11″]Delivery Service Level Agreements[/acp] and [acp author=”Mickey Chandler” id=”Chandler-03″ media=”blog” url=”http://www.spamtacular.com/2010/05/12/asking-for-the-impossible-inbox-guarantees/” title=”Asking for the impossible: Inbox guarantees” year=”2010″ month=”May” day=”12″]Inbox Guarantees[/acp]. Today, we turn our attention to send rate guarantees.
Occasionally a company will try to get an ESP to agree to send rate guarantee. The terms of the contract amendment will usually look something like this:
The ESP will insure that all mail is delivered to the Client’s recipient list within 60 seconds of the time that the Client queues it.
Now, that seems to make some sense. The client obviously thinks that the mail they are sending is important, both to them and perhaps to the recipient. Because of this, they want to see it delivered in a timely manner.
For some mail, this even makes sense. For instance, Woot.com has some time-sensitive deals on their site. It makes sense that they would want to see their mail delivered well before the deal expires.
So, why do terms like this not fly? Again, it has to do with the reality of the email space. Email is not an instantaneous medium. Things can happen fast, but they don’t have to. Most mail transfer agents (mail servers) are setup to stop trying to send mail after three or four days. There is a reason for this default behavior: the reality of the Internet is such that things may take some time to happen.
As we have been mentioning on previous days, there are lots of things that can go wrong. Not everything will mean that a message can never be delivered, but it may mean that it cannot be delivered in the next minute, hour, or even day. Things along this line would include:
- Under sea cables being cut. You might remember that in early 2008, this happened to several cables.
- Mail servers being down. From time to time, the connection cannot go through because the mail server which is receiving the email is not receiving any email.
- Poor reputation. If a sender has a poor reputation, their messages may be blocked, but they could also get deferred for a period of time, perhaps even until off-peak hours for the receiving server.
- Technical issues. This post from the [acp author=”Margot Romery” publisher=”AOL Postmaster Blog” media=”blog” url=”http://web.archive.org/web/20121031084912/http://postmaster-blog.aol.com/2010/04/27/internet-inbound-mail-delay/” title=”Internet Inbound Mail Delay” year=”2010″ month=”April” day=”27″]{publisher}[/acp]points out that a server may not be down, but still may not be receiving mail at expected rates.
- Rate limiting. Some ISPs use rate limits to help make certain that things do not get backed up too much on their systems. This may be tied to poor reputation, but does not have to be.
There are, of course, other things that could contribute to not being able to send all of the mail you want to send as fast as you want to send it. But, all of these things combined mean that it is unrealistic to expect an ESP to agree to send mail with anything other than “best effort.”
[acp add publisher=”Wired: Threat Level” author=”Ryan Singel” media=”blog” url=”http://www.wired.com/threatlevel/2008/02/un-official-fee/” year=”2008″ month=”February” day=”19″ title=”UN Official Feeds Cable Cut Sabotage Speculation”]
