IP Reputation

laura
Industry
September 8, 2014

A throwback post from a few years ago on IP reputation.

Why IP addresses?

ISPs built reputation around IP addresses because it was one bit of data that malicious senders / spammers couldn’t forge. The connecting IP is a fundamental part of the network transaction and if you forge an IP then SMTP can’t work. Because that was the reliable data they had to work with, that’s what they used. Even now, when there are other kinds of data, the IP address is still the first thing the receiving MTA sees.

What is IP reputation?

IP reputation can best be summed up as “past performance is an indicator of future results.” In other words if recipients responded well to mail from an IP address in the past, then they’re likely to respond well to new mail from that IP address.

How is IP reputation measured?

While each spam filtering company and ISP have their own ways of calculating the reputation of an IP address, there are some similarities in what they measure.

How many non-existent email addresses is this IP attempting to deliver to?
How many abandoned email addresses is this IP attempting to deliver to?
How many “known bad” email addresses (spamtraps) is this IP attempting to deliver to?
How many recipients complain about receiving this mail?
How many recipients complain about not receiving this mail?
How respectful of my resources is this IP?
Does this IP keep connections open for long periods of time?
Does this IP retry deliveries too aggressively?
Does this IP stop mailing addresses after receiving a “user unknown” message?
Is this IP address configured as if the associated machine was infected by a virus?
Is this IP address listed on blocklists we use?

That is by no means an exhaustive list of what ISPs measure. If they can measure it they’ve tried. If the measurement helps them separate spam mail from not-spam mail then they’re using it.

How fast does IP reputation change?

IP reputation is often measured over multiple time periods. ISPs can look at a 1 day, 7 day, 30 day and 90 day reputation. A good analogy is stock prices. Prices can be very volatile in the short term, but more consistent over the long term. A single bad day, where one or more reputation measurements go bad, may affect delivery that day or the next day but won’t damage an overall good reputation. Likewise, a few days of improved mail may not be sufficient to counter months of poor reputation.

How is IP reputation used?

Mail from IPs with a high reputation is accepted faster and at a higher rate than mail from IPs with a lower or unknown reputation. IP reputation can also influence whether mail is delivered to the inbox or the bulk folder.

Key IP Reputation takeaways

IP reputation is about how recipients react to mail from that IP. Happy, content recipients turn into good delivery.
Brief changes (for good or bad) don’t necessarily ruin delivery over the long term.
Steady improvements will result in improved reputation.
It may takes as much time to change a reputation in one direction or another as it took to establish the reputation in the first place.

Reputation is more complex than a single number

I checked our SenderScore earlier this month, as quite a few people mentioned that they’d seen SenderScore changes – likely due to changed algorithms and new data sources.

It sure looks like something changed. Our SenderScore was, for a while, zero out of a hundred. That’s as bad as it’s possible to get. I didn’t get a screenshot of the zero score, but I grabbed this a couple of days later:

Are ReturnPath wrong? No. Given what I know about the traffic from our server (very low traffic, particularly to major consumer domains, and a negligible amount of unavoidable backscatter due to our forwarding role addresses for a non-profit to final recipients on AOL) that’s not an unreasonable rating. And I’m fairly sure that as they get their new algorithms dialed in, and get more history, it’ll get closer. (Though I’m a bit surprised that less than 60 mails a day is considered a moderate volume.)
But all our mail is delivered fine. I’ve seen none of my mail bounce. It’s very rare someone mentions that our mail has ended up in a bulk folder. I’ve received the replies I’ve expected from all the mail I’ve sent. Recipient ISPs don’t seem to see any problems with our mail stream.
A low reputation number doesn’t mean you actually have a problem, it’s just one data point. And a metric that’s geared to model one particular sort of sender (very high-volume senders, for example) isn’t going to be quite as useful in modeling very different senders. You need to understand where a particular measure is coming from, and use it in combination with all the other information you have rather than focusing solely on one particular number.

Hunting the Human Representative

Yesterday’s post was inspired by a number of questions I’ve fielded recently from people in the email industry. Some were clients, some were colleagues on mailing lists, but in most cases they’d found a delivery issue that they couldn’t solve and were looking for the elusive Human Representative of an ISP.
There was a time when having a contact inside an ISP was almost required to have good delivery. ISPs didn’t have very transparent systems and SMTP rejection messages weren’t very helpful to a sender. Only a very few ISPs even had postmaster pages, and the information there wasn’t always helpful.
More recently that’s changed. It’s no longer required to have a good relationship at the ISPs to get inbox delivery. I can point to a number of reasons this is the case.
ISPs have figured out that providing postmaster pages and more information in rejection messages lowers the cost of dealing with senders. As the economy has struggled ISPs have had to cut back on staff, much like every other business out there. Supporting senders turned into a money and personnel sink that they just couldn’t afford any longer.
Another big issue is the improvement in filters and processing power. Filters that relied on IP addresses and IP reputation did so for mostly technical reasons. IP addresses are the one thing that spammers couldn’t forge (mostly) and checking them could be done quickly so as not to bottleneck mail delivery. But modern fast processors allow more complex information analysis in short periods of time. Not only does this mean more granular filters, but filters can also be more dynamic. Filters block mail, but also self resolve in some set period of time. People don’t need to babysit the filters because if sender behaviour improves, then the filters automatically notice and fall off.
Then we have authentication and the protocols now being layered on top of that. This is a technology that is benefiting everyone, but has been strongly influenced by the ISPs and employees of the ISPs. This permits ISPs to filter on more than just IP reputation, but to include specific domain reputations as well.
Another factor in the removal of the human is that there are a lot of dishonest people out there. Some of those dishonest people send mail. Some of them even found contacts inside the ISPs. Yes, there are some bad people who lied and cheated their way into filtering exceptions. These people were bad enough and caused enough problems for the ISPs and the ISP employees who were lied to that systems started to have fewer and fewer places a human could override the automatic decisions.
All of this contributes to the fact that the Human Representative is becoming a more and more elusive target. In a way that’s good, though; it levels the playing field and doesn’t give con artists and scammers better access to the inbox than honest people. It means that smaller senders have a chance to get mail to the inbox, and it means that fewer people have to make judgement calls about the filters and what mail is worthy or not. All mail is subject to the same conditions.
The Human Representative is endangered. And I think this is a good thing for email.

Delivery challenges increasing

Return Path published their most recent Global Deliverability report this morning. (Get the Report) This shows that inbox placement of mail has decreased 6% in the second half of 2011. This decrease is the largest decrease Return Path has seen in their years of doing this report.
To be honest, I’m not surprised at the decrease. Filters are getting more sophisticated. This means they’re not relying on simply IP reputation for inbox delivery any longer. IP reputation gets mail through the SMTP transaction, but after that mail is subject to content filters. Those content filters are getting a lot better at sorting out “wanted” from “unwanted” mail.
I’m also hearing a lot of anecdotal reports that bulk folder placements at a couple large ISPs increased in the first quarter of 2012. This is after the RP study was finished, and tells me increased bulk folder placement is more likely to be a trend and not a blip.
One of the other interesting things from the RP study is that the differences are not across all mail streams, but are concentrated in certain streams and they vary across different regions.