One of the most boring things to do in any business decision is making reference calls.  Usually you’ve already made a selection of whatever type (hiring a new employee, deciding on a VC, choosing a new vendor), and then the next/final logical step in a “process” is to call a few people just to verify what you’ve already decided.  But particularly in a startup, where time is of the essence, those reference calls fall to the bottom of the to-do list, and often are never made.  Making reference calls is like buying insurance – it’s not sexy, you’re not satisfied immediately after you’ve done it, and only helps you avoid theoretical disaster down the road.

All of that being said, it surprises me how infrequently startups, and especially startup founders, make diligence calls to pressure-test their decisions.  Founders are used to going-with-their-gut in making choices and the rote process can seem arduous and tedious.

But reference checks don’t need to be a laborious time-sink chore.  A few lessons which I’ve learned about making reference calls which are focused on startups (but can be applied in almost situation):

1. Make them.  Assuming you’re at least somewhat plugged into the entrepreneurial ecosystem, there is real and valuable information out there within your extended network.  Leverage it.
2. One datapoint is just one datapoint.  One phone call that produces negative feedback, especially if it is mild, is a yellow flag, not a red one.  There are many reasonable explanations for why somebody had a not-so-positive interaction with an individual or company.  Often a specific situation can be more powerful than the individual, or at least unique to why someone has acted or references the way they did.  After all, there are three sides to every story.
3. Two datapoints create a straight line, three datapoints form a real trend.  Even though one negative review can be taken in strides and understood within context, once there are two people who have a the same negative thing to say about a person, there is usually more than a mere kernel of truth.  In fact, once three separate individuals have a similar story to tell, it’s fairly certain that the message should ring loud and clear.
4. Mixed feedback is mixed, but not necessarily bad.  Different people excel in different situations.  Some founders triumph when innovating on product, but not on managing people.  A CEO who is a forceful and perhaps a divisive change-agent may be the right executive for an organization which needs a dramatic modification of course, but not the right one to take the helm of a nurturing culture already energized by creativity.  It’s extremely important to take all comments about individuals within context, and extrapolate if that context is applicable to the situation/role you’re diligencing.
5. 80/20 is key.  The old adage that it’s only necessary to expend 20% of the effort to benefit from 80% of the results is absolutely true with diligence calls.  Especially in a startup where time is so precious, a few calls just to pressure-test perception to reality is all that’s needed.  Once a couple conversations start to paint the same story, it’s time to quit.  There are no points for complete thoroughness – no need to make ten calls just to ensure that no stone is left unturned.  There’s always the possibility of outstanding risk, and your job is to mitigate, not eliminate, that risk.
6. Off-list references are twice as good.  The path of least resistance is to just ask an individual or company “for references” and then call the person(s) at the top of that list.  It’s obvious that those are the individuals who are going to say the best things and are the most prepped for a call.  But it’s too easy to make those calls, so don’t make them – consider those people green flags, then dig deeper.  You’re looking for unprepped honest feedback, and you’re more likely going to find that off of a “reference list.”  I’ve found that the best source is just to go to LinkedIn and see who else they know and have worked with intimately in the past.  Even if you don’t have a mutual contact in common, you can proactively reach out to someone who you don’t know with a basic inquiry that you’re doing a quick reference check on someone.  I’ve found that people will usually respond, and that’s a signal in and of itself how they respond to a request.
7. Ask questions that allow people to signal not say bad things about individuals.  Even if they don’t like a person or don’t think that they’ve excelled in their roles, people don’t like saying negative things about others (especially to callers who they don’t know).  It’s best to ask questions that allow people to respond positively but still communicate their point of view.  One person I know used to intentionally call references after-hours so that he could leave a voicemail saying, “Only call me back if you think he’s the absolute best person you’ve ever had in this role.”  That test will surely provide clear feedback without anyone ever having to say not-so-nice things.

At the end of the day, diligence calls come in two varieties: either check-the-box affirmative of what you already know, or tougher ones which come down to interpretation.  Rarely (though I’ve definitely encountered it) will a diligence check be disastrous that it’s immediately and unambiguously obvious that you need to completely reverse your decision.  Even if you heard “mixed” things and decide to proceed anyway, the diligence conversation(s) empower you to go into a new situation eyes-wide-open to the risks associated with any new relationship.