What is threat intelligence? First of all, a threat is something that or someone who could exploit your vulnerabilities. It is not the vulnerability itself. A threat could be a person launching an attack, a backhoe near your network cables, a squirrel that likes to chew on insulation, or even a hurricane. It could be the malware that infects your system, the developer who thought it was a good idea to send one email alert for each log event through your Exchange server, or the auditor who may tarnish your good name with a material finding.
Intelligence, on the other hand, varies widely. I have seen the term used to describe the correlated logs in a SIEM (as if simply knowing what was going on in your network was intelligent enough). It is sometimes used to refer to the collection of event data from more than one enterprise; many vendors do this when they receive this data from their products installed at customer sites. Taking this a step further, many security companies have researchers on staff who try to find out what the newest attacks are, which systems they're targeting, and what vectors they're using. All of these types are "intelligence" based on knowing what's happening out there, beyond your own perimeter.
Besides the "what," "where," and "when," there's the "how." Threat intelligence can include the techniques and technology being used in attacks, the critical factors that mean success or failure, or the timing, location, and sources. This is more than just describing what happened; it's information on how to identify it and how to respond to it.
The "who" and the "why" are in the last tier of threat intelligence, and finding out motivation is often easier than attribution. Not many companies are willing to connect attacks and specific actors with a lot of confidence (unless, of course, you have a smoking gun in the form of an insider). But motivation and attribution are very important for organizations that need to prioritize their activities to deal with the most likely threats, particularly if they're concerned about targeted attacks; they're also crucial for law enforcement if they're to investigate these crimes.
What really brings this data to the level of intelligence is not just describing what you get out of customer logs, honeypots, sinkholes, and mailing lists. It's putting together the disparate sources of data and adding more valuable information: which families of malware could be stopped most effectively with limited countermeasures; which kinds of targets are favored by different attacker groups; how to tell who really attacked you (not just who claimed the credit); and how reliable the information is that you've received.
This is where the rubber meets the road, and this is what you should be expecting from anyone who is offering "threat intelligence." It's not enough just to tell you in detail what has already happened. If it's not helping you make decisions, or be proactive, then it's not worth paying extra for it.
Wendy Nather is Research Director of the Enterprise Security Practice at the independent analyst firm 451 Research. You can find her on Twitter as @451wendy.