Spam Traps are an important concept and useful tool in the email spam-fighting industry.
Spam Traps look like regular email addresses and are visually indistinguishable from regular mailboxes. Those mailboxes, though, do not belong to real people. Instead, they are programmatically monitored and controlled by spam-fighting agencies. Being not controlled by humans means that, if recommended mailing list collection practices are used, spam trap addresses would never be able to "subscribe" to any of the lists. This is because no human would ever enter these addresses in any signup form, not to mention that such addresses wouldn't be capable of passing through double opt-in mechanisms.
Instead, spam trap addresses are "seeded" around the web: they publicly appear on various websites, infiltrate circulating shared "mailing lists", etc.
As a consequence, when an email is delivered to a spam trap address it automatically means, with a high degree of confidence, that the party that sent that email did not adhere to permitted mailing list practices, using non-permitted techniques instead: scraping the web for email addresses, purchasing mailing lists, using outdated unverified lists, etc.
Given these mechanics, spam traps are a great indicator of a mailing list's (and, subsequently, email sender's) good standing and reputation: the more spam trap hits are observed for a sender, the higher is the probability that the sender is using email channel in an abusive manner.
Agencies controlling spam traps share spam trap hit details with companies running email/spam blacklists. As a result, if a certain sender (domain and/or IP address) crosses a certain threshold with regards to the ratio of spam trap hits per email sent, then such a sender gets penalized by being added to the mentioned blacklists. Most email services (such as Gmail): public or private, consult with large spam blacklist providers (e.g. sorbs, Spamhaus, etc.) and tend to reject or throttle, temporarily or permanently, emails from such senders (domain and/or IP addresses). As a result, overall email performance for those senders drops significantly: fewer emails land in inbox (vs. spam) folders, email delivery is delayed or even not accepted at all, etc.
As a consequence, all senders using affected sending domains/IP addresses get penalized. For example, if multiple senders share a sending IP address and one of those leads to a sending IP address getting into a blacklist, all other senders are equally penalized.
As Piano ESP in many cases uses shared sending IP addresses, we take this risk very seriously to make sure that our well-behaving customers are not affected by the "misbehavior" of others. For that, we continuously monitor spam trap hits for our IP addresses and senders and as soon as we notice signs of abnormal spam trap hit ratios, we immediately contact customers to mitigate the situation.
In general, "fixing" spam traps issues is much more difficult than avoiding them: as spam traps are visually indistinguishable from normal addresses, it is impossible to "clean them up". Instead, if the issue occurs, preceding mailing list activity has to be analyzed, all recently added subscribers have to be removed and subsequent mailing list subscription mechanisms have to reassure that no spam traps would be able to enter the lists in the future. Additionally, in most cases, mailing lists have to be also cleaned off all "passive" subscribers (those that never click any email links), which is a good practice anyway.
Taking into account all mentioned above, it is highly recommended to adhere to only accepted mailing list collection practices to avoid even the slightest chance of facing spam trap issues.