Top Ten Investigation Blunders

While I would love it if you hired me to do every investigation your company needs, that is neither practical nor necessary.  Company HR professionals do the bulk of investigations in-house, as they should.  Here are my Top 10 blunders (some personally experienced and others just observed over 19 years of this work).  Avoid these for the best possible investigation outcomes!

(1)            Not investigating because the complainant “asked me not to”/subject quit/                        “it’s just gossip”

Once you are on notice that you may have a legal/policy violation that would trigger an investigation, you have to act.  Even if the complainant doesn’t want you to, even if it’s just a rumor, even if the bad actor quits.

(2)            Not responding quickly enough

Putting off an investigation to run things by the Board, or waiting until your compliance officer is back from vacation to respond . . . these things will get you into hot water with the EEOC.  An employer must respond with a “prompt” investigation.  This depends upon the circumstances, but particularly if you have an issue of potential violence or ongoing harassment, the investigation should start yesterday.

(3)            Not being thorough for the sake of being DONE

This happens a lot.  An investigator fails to interview everyone with direct knowledge.  She doesn’t talk to everyone identified by the complainant and subject.  She interviews by phone when in-person interviews would have been possible.  She doesn’t ask any questions about possible bias or undue influence.  Or she is just a slave to her interview outline.  An employer must do a “thorough” investigation, and this means gathering all of the relevant data.  If you miss a key witness, the rest of the work will have been for naught.

(4)            Not talking to comparator employees or past employees

It isn’t always necessary to interview employees who are in the same situation as the complaining employee, but sometimes it is essential.  If your complainant is racially diverse for example, and claims he was treated differently on the team, it is imperative to interview comparators and see what they say.  Similarly, sometimes the facts show you that you should talk to former employees.  Failures to cover these bases can come back to haunt you later when this information comes out in discovery.

(5)            Not keeping an open mind until all of the data is in – building a case

Workplace investigations require neutrality.  And this means you are neutral at the start – you don’t have a bone to pick with anyone involved and they don’t supervise you – and it means you remain neutral.  Just because the first 8 witnesses are saying the same thing, this doesn’t mean you should believe it.  The 9th witness could be the iceberg that sinks the Titanic and changes the whole story.  You must gather data on both sides, up until the end of the investigation.  Curiosity, not judgement, is the right approach.

(6)            Talking too much – listening is NOT just waiting for your chance to speak

Your job as an investigator is to gather information, not impart it.  If you find yourself talking more than 15% of the time, you are talking too much.  Ask questions, make them open-ended, give your witness the space to speak.  Don’t interrupt.  Don’t rush in to fill silences.  Sometimes silence can be a tool you use to get your witness talking, to fill the space you’re leaving behind.

(7)            Leading and / or accusing questions

This goes along with keeping an open mind.  Your job is to ask as many open ended questions as you can, and also to ask the tough questions, but your approach must always come from curiosity and not judgment.  Your job isn’t to suggest the right answer by your questions, or to accuse anyone of anything.

(8)            Not getting facts – letting the witness characterize events

The facts ma’am, only the facts.  Am I sounding old by that reference?  As investigators, hearing someone say, “He was drunk,” isn’t nearly as helpful as, “I came across him sleeping in the supply closet.  His mouth was open and he was drooling.  When I dropped a box of files he didn’t wake up.  He reeked of bourbon.”  What you want as an investigator is facts, not characterizations.

(9)            Promising confidentiality

The NLRB’s Banner decision has everyone in the investigative community talking about confidentiality.  How we reconcile Banner’s prohibition on blanket confidentiality advisements with the EEOC’s requirement that workplace investigations be conducted confidentially is a subject for another day’s blog, but suffice it to say that no investigator should be promising confidentiality to witnesses.  First of all, the company is investigating because they want to know what happened so that problems can be rooted out and fixed.  This may require people to be identified in some cases.  Additionally, an investigator will explore issues of bias and credibility and must allow an accused employee to know who is bringing the complaint.  Finally, if there is ever litigation, it’s highly likely that everything the witness says and every piece of data will come out.

(10)            Checking your other senses – including common sense – at the door

As an investigator, you don’t have to take a witness’s word for anything.  Your job is to test the statements you hear.  If the witness tells you that she overheard comments between two employees, see where she sits.  Is she right outside the office door in question, or across the room at a cubicle 50’ away.  If another witness tells you that the parking lot was brightly lit by moonlight in the night in question, check a moon phase calendar and weather – was it a new moon (essentially no moon), was it a full moon?  Was there a thunderstorm?  Workplace investigations are different from crime scene work or following a cheating spouse with a camera, but they are still investigations.  You are not limited to witness statements or emails – use your creativity and common sense to test the data you are gathering, and make your assessment.