Wednesday, November 9, 2011


If, as Sun Tzu famously said, war is about deception then negotiation is about information.  I advise people to begin every negotiation by asking questions about the information they have at hand, such as:
  1. What information is critical to this negotiation?
  2. What parts of this information am I missing and how can I acquire them?
  3. How confident am I of the information I do have, and how can I test and/or improve it?
  4. What information do my counterparts have?  How confident am I about my assessment of their information?
  5. What information do I want to share with my counterparts?  What is the best way to do so?
  6. What information might my counterparts want to share with me?  What misinformation might they want to share?
Guhan Subramanian tells the following story to illustrate the gray area of ethics and law that negotiators can find themselves in.  To me, however, it's an even better example of the importance of asking questions about information.

A large division was being sold at auction by an investment bank.  It was understood that the number of serious potential bidders would likely be small (probably as few as three), and the bankers approached each of them to discuss the terms of the deal.  Due diligence would take place over a weekend at the bank's main offices, and bidders would get half-day slots, beginning on Saturday morning.  One interested party asked for the Saturday morning slot, but it was taken.  Same with Saturday evening and Sunday evening.  Sunday morning, starting at 8:00am was available.  When they came in on Sunday the trash cans were overflowing with Chinese food takeout boxes and the investment bankers had stubble and looked a bit rough.

You can probably guess the rest.  When the bankers approached the likely bidders only one was interested so they created the perception of strong interest in order to force the buyer to bid as though they were in a competitive auction.  The buyer paid hundreds of millions of dollars more than necessary.

The legal side of this behavior is interesting.  (Subramanian says that most American lawyers think the bankers acted legally; most European lawyers disagree.)  But let's take a step back and think about what asking the above questions about information might have made possible.

If you're bidding on something that only two other parties are likely to be interested in, what information is most critical to your negotiation?  Surely their level of interest.  And how might you learn something about that?  A rather obvious solution would be to hire someone to photograph (from a distance) everyone entering the building on Saturday and Sunday.  That wouldn't work in all cases, but a typical investment bank has extremely low foot traffic on weekends.  Given that the potential buyers would all be familiar to each other, it's quite likely that those pictures would have been worth many millions of dollars.

Wednesday, November 2, 2011

Relationship Building Before Negotiation

Back when I was a kid, my Dad decided to repaint the outside of our garage.  The garage was at the end of a driveway that bordered the property line with our neighbor, so one of the side walls was visible from the neighbor's house (and not really from anywhere else).  Dad approached the neighbors and asked what color they would like the wall painted.

It was an unusual request, and if I recall the story correctly they thought at first that Dad was asking them to help pay for the project.  No.  He just recognized that since the wall was only visible from their property they were the ones with an interest in its color and it didn't hurt us to have the color be different from the rest of the garage.  They chose something that suited them and Dad bought the paint.

I don't think we ever had any serious negotiations with our neighbors, but suppose we had.  Dad's simple act of generosity would likely have earned him their trust and goodwill, greatly improving the chances that any difficult discussion went well.

One of the most important lessons of negotiation success is that your most profitable work is done away from the table -- sometimes when there isn't anything to negotiate!  We don't always know where we'll find ourselves needing to negotiate, so even cynics should look for ways to strengthen their relationships and in particular their reputation for taking the other party's interests seriously.

Opportunities to generate goodwill, at low or no cost, are all around us.  The challenge is to be mindful of the opportunities so that the next time we're painting a garage or have some other chance to help someone out, we take it.

Tuesday, October 25, 2011

Courting bad behavior

The other night I was watching Storage Wars and saw an unhappy but predictable bit of value destruction play out.

For those not familiar with the show, kudos on having better taste than I do the basic setup is as follows.  Property that has been in private storage gets auctioned off when the owners fail to pay rent on their unit.  Storage Wars focuses on four men who try to make a living bidding on the contents of the units.  The bidding is complicated by the fact that the storage managers only open the front door to the unit; bidders can't go inside or open boxes, so they have to garner what information they can from a very limited viewing.

In this episode, nine lots from a single owner were being auctioned.  The most experienced bidder (Dave, nicknamed "The Mogul" on the show) looked in the first unit and realized that the semi-random pieces of metal it contained were part of an expensive stage lighting system.  He guessed (correctly) that the other units each contained part of the overall system; the whole was worth much more than the sum of the parts.

Dave decided to try to buy all nine units up for auction.  Most likely he reasoned that if no one else realized what was in them they wouldn't see the value and he could outbid them and still make money.  Moreover, if someone did realize what was in them, then once he bought the first unit he'd still have an advantage, since he could complete the system while a rival bidder's best hope would be to get part of it.

The snag in his plan soon became apparent.  One other bidder, Darrell, understood what was going on and waited until Dave had invested a lot of money buying several lots.  He then began to bid aggressively, starting with a lot he suspected had a particularly key component -- the control system.  Dave, likely realizing that if he forced the issue on this lot he'd have to keep doing so on the rest, let the lot go after the bidding had gone high enough.

After the dust settled, it seemed that Darrell's gambit had worked.  He did indeed have a central component to the system.  He'd bought it for around $1200 and it would cost Dave $5,000 to replace.  He offered to sell it to Dave for $3,000.

Dave refused, saying, "I'd rather die than see you make a buck."  Instead, Dave sold what he bought as scrap metal, letting him make a profit albeit a much smaller one than he could have made.  A lot of value was destroyed.

So what happened?

First, a disclaimer.  Storage Wars is a TV show.  The incentives are almost certainly distorted by the fact that the players are paid and drama is encouraged.  That aside, this sort of dynamic plays out in competitive negotiations so let's treat it as real and unscripted.

The math on Darrell's gambit was flawed in one key respect -- it looked at this deal in isolation.  He was correct in assuming that he could force Dave to let him have one of the units (in order to prevent this, Dave would have had to pay a premium for every single unit remaining), but his assumption that Dave would then buy that unit back from him only makes sense if this is the only time these two are competing.  Since Dave knew there would be plenty of other opportunities for Darrell (and others) to try to hold him up in the future it could be rational for him to take a loss today in order to signal that he would retaliate against anyone trying to do so.

Darrell also magnified the likelihood that Dave would fight back by trumpeting what he was doing during the auction.  He crowed about how he'd blocked Dave's plan and was going to make Dave pay him.  Not only might this have made Dave angry it increased the signal value (or cost) of Dave's decision.  If he paid Darrell then all the other bidders would have been on the lookout for a chance to do something similar.

Darrell's chance of success would have been much higher if he took a different approach.  He should have bid on that unit without giving any indication that he knew what Dave was up to, instead indicating that he thought he saw something collectible in it.  Then, when the dust settled, he could have approached Dave and said, "That unit I bought had a control panel that's a key component to the system you put together from the other units.  It retails for five grand, but I wasn't buying it to screw you so I'll sell it to you for three."  Dave might still have refused, but he would have had better incentive to negotiate.  Instead, Darrell took completely unnecessary effort to make Dave angry and to create incentives for Dave to refuse to deal with him.

Yesterday I was in a meeting for a multi-party negotiation that has become quite contentious.  One of the parties accused another party of lying on a key issue for a period of several months.  I think the first party had some grounds for complaint and even for suspicion but it's very rare that calling someone a liar is constructive.  In this case it was particularly self-destructive because the latter party has nearly all of the power in the negotiation.  They'd like to reach a mutually-acceptable agreement but if none can be reached they are in a position to go ahead with what they want.  Calling them liars encourages them to be defensive about legitimate complaints and encourages them to shut off potentially constructive paths to dialogue.

In each case, the lesson is largely the same.  Negotiators need to think about how counterparts will react to their words and actions.  If you've pulled off a maneuver that benefits you at someone else's expense, the last thing you want to do is trumpet it!  If you want to voice criticism of the other party's behavior, think carefully about how you express it.  There's a big difference between, "There have been some important pieces of information that weren't communicated clearly; how can we improve this going forward?" and "You've been hiding information you knew we needed."

Negotiation is about getting other parties to do things that are in your interest.  Like a doctor's first rule is to do no harm, a negotiator's first rule should be never to cause others to act against our interests.

Tuesday, October 18, 2011

Leverage from Being Small

It seems absurd to think that being a relatively small client gives you leverage but in many cases it does -- provided what you want is reasonably easy to grant.  Companies generally develop a low cost approach to managing small clients.  The person (or service department) managing your account likely has a large number of similar clients -- more than they can possibly spend much time with.  The model works because most of your peers don't ask for service and (in many cases) because the customer service reps are empowered to meet small requests quickly so that the client is retained and expense is kept low.

This comes up quite often in personal business, such as banking, cable, etc.  You may not be able to avoid paying standard fees at your bank, but if you incur a one-time fee you can often negotiate it away just by asking.  Similarly, if there are two cable providers in your area and the one you don't use is offering a sweet deal, you can likely get your own provider to match it or to provide you with something similar.  Your leverage comes from providing the company with a choice -- make you happy at little cost or risk you switching to the competition.  Not only don't they want you to switch, they don't want to invest a lot of expensive staff time retaining you.  (In these negotiations it helps if you say that you're considering switching as a result of whatever it is you're not happy about or whatever special the other company is offering.  Agents sometimes only have the ability to grant your request if you indicate that you might switch.)

This type of leverage isn't limited to getting six months free HBO.  Consider my negotiation on behalf of a small healthcare non-profit we discussed under Negotiating with a bad BATNA.  My client represented well under 1% of revenue for Blue Cross Blue Shield of Massachusetts.  I wasn't even negotiating with the decision-maker there, but rather with two of his subordinates.  So how could my (implicit) threat to end our contract with them have any leverage at all?

Consider the situation the Vice President in charge of reimbursement was in.  He was engaged in a massive struggle with Massachusetts hospitals to halt and even reverse the long trend of rate increases.  He didn't want to lose small partners like us, particularly if we were profitable partners, but he also didn't want to spend any bandwidth on us.  My goal, therefore, had to be to make sure that when our case got to him (and we were asking too much to think it wouldn't), two things would be apparent:

  1. Giving us what we wanted was strongly preferable to losing us.
  2. Losing us was a real possibility if he said no.
  3. Trying to keep us while giving us less than what we wanted would be costly, in his time.

I didn't just make sure that his subordinates understood that using us was saving them money.  I prepared a powerpoint slide (designed so that they could go over it with their VP) that showed how much we were saving them and then compared that with operating losses our organization was incurring.  The message was clear -- we weren't just asking for money, we needed it to make our business sustainable.

Next, I provided a strong narrative for my counterparts to use in selling the increase.  Internal negotiations amongst the various people in an organization are critical, so I helped my counterparts with their task of selling the increase to senior management.

BCBS came back with an offer of 3%.  This was better than most providers were getting that year (most hospitals were having rate reductions) but we turned it down and said that we felt we needed to escalate to the VP.

(As a side note, I'd been very open with my BCBS counterparts both about what we needed and my awareness of their constraints, so that when I escalated it was expected and came across as a natural next step in the process rather than me going over their heads.  Now they were in a key way my allies -- an escalation like this is much better for them if it goes quickly and seems justified, so from their perspective it would be much worse if the VP turned us down than if he concludes that this is a worthy exception.  In the former case he's doing their jobs; in the latter they did their job correctly and now he's doing his.)

I revised the powerpoint presentation with feedback from my BCBS counterparts and they presented it to the VP, along with a list of reasons (ranging from financial to our non-profit status and work with poor communities) why we should be an exception to BCBS's determination not to give out rate increases that year.  They came back with a substantial increase, no doubt in large part because the VP saw that as a better option than continuing to negotiate or risking losing us as a partner.

Being small often means that your counterpart doesn't need you it also means that granting you what you're asking for probably doesn't cost them very much.  If you can frame the choice so that it feels reasonable and present your counterpart with the ability to make you happy quickly and easily, there's a good chance that doing so will be their best option.

Thursday, October 13, 2011

NBA Lockout: a mechanism for avoiding value destruction in disputes

The recently-announced NBA labor dispute will eliminate at least the first two weeks of the season.  On the face of it, such disputes should be rare.  The total pie (i.e. gross profits to be shared) is huge, both parties have expert negotiators to represent them and both parties incur substantial costs when they are not able to reach agreement. Today's post will look at how disputes of this kind arise generally (and not just in sports) and how they are made more likely by the structure of payments typical in sports.  Finally we'll look at a mechanism that could reduce the risk of destructive escalation going forward.

Let's start with a basic problem that arises when two parties must agree how to divide a large pile of money.  This seems like a wonderful "problem" to have, but it's surprisingly easy for it to go wrong.  To see why, let's play a game called "Ultimatum."

There is $1,000,000 on the table.  You get to propose a split.  I can then either accept the split, in which case the money is divided as you proposed, or I can decline it, in which case neither of us gets anything.  This structure puts you into a very powerful position.  If, for example, you propose that you get $900,000 and I get $100,000 (a very slanted split) my choice is between $100,000 and spite.  Unless I'm a very spiteful person, I'm going to take the $100,000.  If you know I really need cash you could propose a split of $990,000 to you and $10,000 to me, capturing 99% of the value on the table by leveraging your position.

Now imagine a more "normal" negotiation over how to split a million dollars.  We'll call this game, Even Steven.  We have an hour to negotiate; if we haven't struck a deal by then neither of us gets anything.  Neither of us has any advantage; if you suggest a split that favors you I can say "no" without losing anything.  Most likely (and this is supported by practical experience and academic studies) we'll end up splitting the pie evenly and each taking home $500,000.

Now, if you're choosing between a game where you win $900,000 and one where you win $500,000, which would you rather play?  Most of us would rather play Ultimatum, which is where commitment dynamic into play.

Suppose we're playing Even Steven and you come up with a way to convince me that you'll refuse any deal that doesn't give you at least $900,000.  Perhaps you sign a contract binding you to lose $1,000,000 if you agree to any worse deal.  Perhaps you have a reputation for doing anything to keep your word and you swear on your honor that you don't take a penny less.  Whatever the specifics, if you can make a credible commitment to take nothing unless you get $X, then you've changed the game from Even Steven to Ultimatum with you in the driver's seat.

The same dynamic happens in real-world negotiations.  When the pile of money to be divided is big or the consequences of failing to reach agreement are disastrous there is an incentive to claim a big share of the pie by committing yourself to an aggressive bottom line.  We saw this play out in the recent debt ceiling negotiations -- Republicans made very public commitments not to accept any tax increases precisely because saying it privately wouldn't have committed them.  That put Obama and the Democrats in a position where if they insisted on tax increases they risked finding out that they were playing Ultimatum rather than Even Steven.

Even if you don't have a mechanism to create commitment, you can try to play Ultimatum by appearing committed.  Using this approach, a negotiator stakes out a very aggressive position and either sticks to it or makes tiny concessions (which is generally a signal that someone is near their bottom line).  "OK, you won't accept me getting $900,000 to your $100,000?  I can go $880,000 to $120,000."  If both parties do this, of course, the negotiation can quickly break down into acrimony -- neither side trusts the other, the parties are far from agreement, and progress looks remote.

Things get worse when we add ego (and I think it's fair to say that professional athletes and the owners of NBA teams all have healthy egos) and different senses of what's fair to the mix.  If it's possible for people to fail to divide a pile of money on the table (and it most certainly is), imagine how much harder it gets if each person think that they deserve 70% of that pile.  People commonly overestimate their contribution to any shared effort; it's safe to assume that the players and team owners have very different views of how much each of them "ought" to get from the revenues generated by the NBA enterprise.

All that would be bad enough, but sports leagues often add one more structural problem to the mix. Players typically earn salaries that are paid throughout the season, while owners earn a relatively high percentage of their income during post-season television.  This means that each side has a window in time where the pain of a fight is greater for the other side than for themselves.  For this reason, lockouts are more likely to happen at the start of the season (more painful to players) and strikes more likely later in the season.

Two Different Types of Value Destruction

If labor and management, players and owners or whatever parties share in creating a pool of value cannot agree on how to divide it, they typically stop creating that value -- temporarily.  Strike, lockout, "taking a break" in the relationship, the basic dynamic is the same.  Each side stops receiving its current share of value creation until the dispute is resolved.

While in some sense any value destruction is unfortunate, a good case can be made that at least some of it is necessary, too.  If the current division of value favors one party over the other, that party may need leverage to bring the other party to the negotiating table.  Even in the game Even Steven it's useful to have a deadline, since otherwise negotiations could drag on indefinitely.

It's worth distinguishing, however, between two different forms of value destruction.  The first is having to forego current value that would have been created if no dispute takes place.  In a strike/lockout, for example, workers don't get paid* and owners don't get the profits the business would normally earn.  Ideally agreements would be resolved before a strike is necessary but having that stick available is healthy and probably unavoidable.

Another source of value destruction is damage done to the enterprise itself by the strike.  If a company loses key customers permanently because a strike disrupted its service to them, then when labor and management agree on how to divide the pie they are sharing a smaller pie -- potentially much smaller.  As Hall of Fame goalie Ken Dryden put it (discussing the 2004-5 NHL strike), "You never want to give a fan a chance to find out whether it was passion or habit."  If fans get turned off watching players and owners (each group being far richer than the typical fan), future revenues can be significantly damaged.

A Mechanism to Reduce Value Destruction

So now we have a conflict.  The ability to inflict economic pain is a useful tool, without which a division that favors one side may be impossible to resolve because that side won't come to the table.  At the same time, both sides have an incentive to limit the pain to lost revenues during the negotiation itself rather than killing the golden goose.

In theory, this can be avoided by replacing an actual strike with an artificial one.  Players and owners contractually would commit to pay money into a pool in the event that they are unable to agree on contract terms -- with the payments designed to simulate the revenues that would have been lost in the event of a strike.  Instead of an actual strike/lockout, however, play continues -- and fans, instead of being disappointed and potentially putting their entertainment loyalty elsewhere, get to watch the teams they love compete.  Best of all, instead of the losses being "real" losses they could be donated to charity.  Imagine the positive press if as a result of a contract disagreement the NBA were announcing that it would be donating a large sum of money to improve inner-city parks or to provide sports equipment to low-income public schools...instead of canceling games.

Such a solution would be unusual, but far from impossible to implement.  A trusted third party (e.g. McKinsey & Co.) could be brought in to do the analysis on how much each side would pay and to provide a mechanism for re-evaluating the system as revenues change in future years.  All it would really take is a willingness to adopt an unusual agreement on how to resolve future disagreements.  The end result would be superior for all parties -- owners, players and fans.

* Workers typically get paid by their union, but not at full wages and in any case this is money come out of their collective pockets rather than from the company.

Thursday, September 22, 2011

How does your Reservation Value relate to your BATNA?

Your reservation value is the lowest value you will accept in a deal.   It goes by many names -- your bottom line, your walkaway point, but whatever you call it it is the worst possible deal you would be willing to accept.  We always hope to do better than our reservation values, but it's important to know what yours is, both to avoid accepting a deal you shouldn't have and as a reference point for how much a current deal is worth to you.

If you read a typical textbook on negotiation, you might think that the question of how reservation value  relates to BATNA is trivial.  Most books define Reservation Value in terms of the value of one's BATNA and assume that they are equal.  This makes sense -- as long as the deal is better than your BATNA you'd rather take it than walk away; if it's of equal value then you're indifferent and if it's worse than you'd rather follow your BATNA than take the deal.

For example, imagine you're selling a used car and your BATNA is to sell to someone who has offered you $6,000 in cash.  Your BATNA is therefore worth $6,000 and any deal worth less than that will be refused. If you had a deal for $6,100 and were sure you couldn't increase it, you'd take the deal.  Right?

I'm not so sure.  I think there are good reasons to set your reservation price at a premium to your BATNA, coming from such diverse fields as psychology and poker.

Pot Odds

Experienced poker players will make bets they expect to lose because of "pot odds".  If the cost of losing is small relative to the gain of winning, the expected return can be positive even if the most likely outcome is a loss.

Suppose we're playing no-limit poker and I'm down to my last $1,000.  There's $10,000 in the pot and it's raised $1,000 to me.  Based on my read of my opponent I think I'm an underdog, with only a 25% chance to win.  Even though I expect to lose, I should call assuming my decision is based purely on maximizing my expected return on this hand.  The math works as follows:

Fold: 100% chance of having $1,000.

Call: 25% chance of having $12,000 ($10,000 plus $1,000 bet plus $1,000 call) and 75% chance of having $0 which is a weighted average return of $4,000.

Since the payoff was much larger than the investment the expected return on calling was positive even though I expect to lose three times out of four.

One professional negotiator I know defines reservation price as "the lowest price you'd accept if you were positive you couldn't get more" but this assumes knowledge we almost never have -- that the other person really won't do any better than their "final" offer.

In our car example, suppose you refuse the $6,100 offer and say you want $7,000.  What might happen?  It's possible that the other party will say, "Sorry, $6,100 really is the best I can do."  They might groan and say, "I can do $6,300 but that's really it."  Or they might actually value the car at $8,500 and accept your price.  Since $6,100 is really only worth $100 to you (relative to your alternative offer), it's quite likely that you have a positive return on holding out for more, even if you think it's likely that $6,100 really is the buyer's top price.

A final consideration is that refusing someone's final offer isn't always irrevocable.  If you say "no" to $6,100 and the buyer walks away you can often (not always) call back and say you've changed your mind.

Underestimation of your BATNA

I find people frequently underestimate the value of their BATNA -- in particular the potential to improve it.  Is your BATNA regarding the used car really $6,000?  That's an offer you have, but it might be negotiable.  You might also find another buyer.  We tend to focus on the most tangible aspects of our BATNA, which can lead us to undervalue it.  Of course, in theory this means we may need to improve the evaluation of our BATNAs rather than that our reservation price should be set higher, but in practice it speaks to holding out for more.


Many negotiations you undertake are semi-public -- people you'll negotiate with in the future (including those you're negotiating with now!) will have some awareness of the results.  Given that, it's worth thinking about the reputation you'd like to craft and then weighing that in to the decisions you make.  An ideal reputation would include integrity and an ability and willingness to create value but I think it should also include being someone who expects significant value out of a deal and will walk away if that proves impossible to achieve.  This may shield you from wasting time on deals that don't offer you much and it encourages your counterparts to adopt a value-creation posture rather than trying to squeeze you.

Cognitive bias

There are at least two types of common cognitive bias that make a no-gain reservation value dangerous: small pie bias and focal point bias.

Small pie bias is pretty straightforward -- people tend to underestimate the size of the pie, which can lead to missed opportunities to grow it and to accepting too small a piece.  Focal point bias refers to research showing that in stressful, complex situations (e.g. negotiations) people tend to gravitate mentally to whatever focal point is the most significant to them.

Combined, these factors can easily lead us to think that our reservation value is the value to think about and that any gain above that is a win -- a prescription for leaving a lot of value on the table.

One strategy for fighting this is to set an aspirational target and try to focus on that, keeping the reservation value in your back pocket to be looked at only if the negotiation goes more poorly than expected.  This approach has some validity but I think it's risky for all but the most experienced negotiators.  Moreover, if you come to believe you've misunderstood the ZOPA and are faced with a deal that is better than your BATNA but worse than the reservation value you've set there's nothing stopping you from taking a step back, re-examining the new information you've received, and adjusting your numbers accordingly.

Putting it all together

Let's look at an example.  You've been tasked with buying the patent rights for a piece of technology that's worth $10 million more to your company than your best alternative.  Based on a classic BATNA analysis, you'd set your reservation price at $10 million.  Your company, however, is the only natural buyer for the technology and you're fairly confident from your analysis that the patent isn't worth anything to its owners and that the best rival offer they're likely to get (from a company in a different business) is in the $3-4 million range and probably less.

There are a lot of away-from-the-table steps you'd take if this was a real situation, but to keep it simple let's assume that they have an unknown but firm offer and are negotiating with you to see whether your company gets the patent.

Suppose they say, "We have a very good offer from (the firm you expected would pay at most $3-4 million), but we know this technology is very valuable to your firm.  We'll sell it to you for $15 million."

You take this in stride as an anchoring tactic, state clearly that they have substantially overestimated how much the patent is worth, and make a counter-offer of $2 million.  There is some back-and-forth but abruptly they say, "Our absolute final offer is $9.7 million.  If you won't pay that, we'll have to go with the other company."

You tell them you'll consider their offer, cautioning them that it's high, and go back to your desk.  You revisit your analysis with your team and bring in an outside expert for a fresh view.  You still can't believe that it's worth more than $4 million to the other company.  But what if it is?  You don't want to throw away money, even if it's "only" $300K.

In this situation, there are some compelling reasons not to accept their "final" offer.

First, let's be pessimistic and assume there's only a 20% chance that they're bluffing -- but that if they're bluffing you'll be in the driver's seat and can reasonably expect to pay $5 million.  That's an expected return of $1 million relative to your BATNA, which is greater than the $300K on offer.

Second, if your analysis is correct and the ZOPA is $6 million or $7 million in size, neither your firm nor your own career benefits from capturing so small a piece of it.

Next, is your BATNA really to do without the technology?  If the firm that buys it is in a different business you may be able to license the technology from them.

This is, of course, simplified.  You have a range of tactics available to gain information...but at some point in a negotiation you can be presented with a "final" offer that doesn't make sense given everything you can learn but that is slightly better than your BATNA.  Theory says you should take it -- I hope I've provided some reasons to question theory on this point.

Wednesday, September 14, 2011

Own Your Mistakes

I started this blog while at an executive negotiation class at Harvard Business School.  It was a fantastic opportunity to negotiate with professional investors from around the world, and I take real pleasure in the fact that in all of my negotiations I reached agreements that created the maximum value possible (in one case my group was the only one that did) and captured significantly more of that value than average.

All of them but one, that is.  In one simulation, I failed to reach a deal at all.

The negotiation involved three entrepreneurs who had formed a very loose ownership agreement when they started a new venture.  A few months in, one of the founders (me, in the exercise) concluded that he was working harder and contributing more that was reflected in the original agreement and that this should be reflected with a higher share of equity.  He initiated a renegotiation, sending a proposal to his co-founders which formed the starting point of the exercise.

Our discussion didn't go well.  One of the other co-founders was keen to make a deal but the other wasn't.  He made demands that we couldn't meet, said things that were frankly insulting (always interesting in a simulation!) and ignored our responses.  As an example, one of the topics for discussion was his salary (the other co-founders were working just for equity) and neither of us could agree to anything above $110K as we didn't think the venture's cash flows could handle it.  Even after we'd made that clear he was proposing deals in which he got paid $150K.

Near the end of the exercise it looked like we might be able to reach a deal, but at the last moment he said he couldn't agree to what was on the table and we "agreed to disagree".  Upon returning to the class

My initial reaction was to blame our difficult partner.  This was reinforced as we learned more about the other roles and it became clear that salary -- what he had made the largest demands about -- was actually his least important interest and he'd failed even to bring up his most important one.  Small wonder we couldn't create enough value to reach agreement when a key player was giving poor information.

Unfortunately, life is rarely so simple.  Negotiations is a complicated endeavor and one where we rarely, if ever, get a precise read on how well we did.  For this reason it is vital that we review our negotiations candidly, gathering new information if possible and evaluate our performance as objectively as possible.  Most of all, we have to own our mistakes.  That's why I always review and analyze my negotiations to see what went well and, in particular, what I should have done differently.

In this case I did at least a few things wrong.  Foremost, I was in too aggressive a mindset.  This exercise, the second in the class, came immediately after a review of a solo exercise in which I had paid $70K for an asset on which the median purchase price was over $200K.  Our negotiation had been discussed in class and I judged (correctly) that in the next exercise my counterparts were going to be aggressive in order to avoid a similar result against "the tough negotiator".

Because I was gearing up for a fight, I didn't put as much thought as I normally would into understanding where my counterparts were coming from, something that is key to successful negotiation.  Afterwards I would blame our difficult partner for not raising his key issue but normally I would have figured it out ahead of time and been able to raise it myself.

In negotiations it is generally preferable to take an aikido approach.  If someone is being aggressive, don't meet that aggression head-on.  Doing so makes irrational escalation more likely and makes it harder  to work together to create value.  Instead, disarm him with a lateral move.  In this case if I'd put myself in his position I would have realized that he was feeling like an employee rather than a partner and that "my" initial proposal emphasized that.

Imagine that you're feeling unappreciated and are gearing up for a tough negotiation against a hard-nosed bargainer.  Imagine that he begins by saying that he tried looking at the situation from your side and realized that if he was in your position he would feel taken for granted.  He wants to start the discussion by asking if that's the case for you, apologizing for it, and then asking what factors are most important to making you feel good going forward?

Another mistake I made was in my orientation to any final deal.  As we discussed here, the value you capture in negotiations that involve ongoing relationships are a combination of the tangible factors covered in the deal itself and the intangible consequences of how your counterparts behave in the future based on their own sense of the deal.

In this exercise, three entrepreneurs were agreeing on principles of fairness and value distribution...but what value was being distributed?  The equity of the company!  This is far from a fixed pie; if we reached a deal but ended up not wanting to work together, we'd each end up owning a slice of nothing.  Even in a class exercise this should have affected my approach.

One of the most common forms of human bias is thinking we've performed better than we have.  A famous example is car drivers.  In one survey, motorists were asked to rate their own driving skill on a scale of 1 to 5 as well as the average skill of other drivers.  Other drivers came in at 2.7, while drivers rated themselves 3.9.  Unfortunately, this self-delusion can be extremely costly when it prevents us from improving.  A tough self-assessment in which we explicitly look for our own errors is one tool to help us resist it.

Thursday, September 8, 2011

Adding Issues to Create Value

As we’ve discussed in prior posts, one of the most straightforward and effective ways to create value in a negotiation is to identify and implement value-creating trades.  If getting your way on an issue is worth a million dollars to you while getting my way is worth half a million to me, then my goal shouldn’t be to get “my way” but rather to let you get your way…in exchange for more than half a million dollars in value.

In some deals this is relatively easy.  The discussions begin with a variety of interests identified, the parties are sophisticated enough to be able to evaluate trade-offs efficiently and discussions move effectively towards the value-maximizing configuration of terms.

Frequently, however, life isn’t so simple.  Today’s post looks at introducing issues to a single-issue negotiation using a real-world negotiation I’ve just begun with a software developer.

Some years ago I published a card game called The Battle for Hill 218.  It’s fun, quick and deep and has been my most successful single product (still selling quite well today).  A few different players have asked me to consider making a version for the iPhone or iPad and recently one of them introduced me to a developer who was interested in creating an iOS version on a revenue share basis (i.e. instead of me paying for his time, he’d be paid out of a share of sales for the app).

The developer’s initial offer was a 70/30 revenue split in his favor.  Alternately, I could pay him a flat fee which would be somewhere in the $10K to $20K range.

If the revenue split is all we’re negotiating there’s not much we can do to create value in the deal.  Naturally there’s some value in each of us being happy with our agreement and being motivated to promote the app, but for the most part we’d be haggling on price and trying to capture as much of the ZOPA as possible.  If there’s only one issue, there’s no room for value-creating trades. 

Price, however, is rarely the only important issue worth discussing and in this case there are a number of other possibilities worth considering besides the split of sales.

Price of the app.  Even if we agree on what price will maximize revenue of the iOS version of Hill 218, either of us might want to set a lower price in order to attract more users to the app.  The developer might like wider exposure of his work and to be able to say (to other potential clients) that he produced an app that a lot of people downloaded.  More concretely, I have a monetary incentive to maximize users (especially new users) since I’m still selling the physical game.  I’m convinced that offering a free Java version of Hill 218 played an important role in sales of the physical game, and having an iOS version is likely to lead to more sales as well.

Let’s assume for illustration purposes that his ideal price is $3.99 and mine is $1.99.  Our next task is to determine how strongly we prefer those prices, as well as how we feel about prices in between.  If a price of $1.99 is worth $5,000 more to me than $3.99 and is worth $3,000 less to him, then pricing the app at $1.99 creates $2,000 in value between us.  Put another way, I can give up over $3,000 in value from app revenues (which makes him better off) and as long as I give up less than $5,000 I’m better off as well.

Marketing budget.  Presumably we can increase sales somewhat by advertising, but I get a double benefit since advertising should boost sales of the physical game as well.  To keep things simple, let’s assume that we each think that each dollar (up to $1,000) of advertising would lead to only eighty cents in additional app revenue (forty cents to him and forty to me if we’re splitting things evenly) and that I also think it will be worth forty cents in additional revenue from physical sales.  The total value of advertising is $1.20 per dollar, so it’s clearly attractive.  If, however, we don’t negotiate advertising then it won’t happen; he gets only forty cents on the dollar and I get eighty (forty from the app and forty from the physical game).  Neither of us gets enough to justify doing it alone.  If, however, the developer “pays” me more than twenty cents and less than forty cents for each dollar of advertising I do, we’re both better off.  (Payment could be done directly (i.e. we share the actual ad expense) or indirectly (i.e. factored into the terms of the deal).

Split based on results.  Hill 218 is a new game for the developer.  He’s played it online and likes it, but he’s naturally cautious about how much it will sell.  I’m more optimistic, given its sales numbers and how popular the Java app has been.

This difference will naturally factor into our discussion over what a split should be.  He’ll point to the number of game apps that sell less than a thousand copies and argue that he’s unlikely to make a normal developer fee even with 70%.  I’ll talk about how many unique players the Java version has, how the physical game has nearly sold out its second print run and the posts on BoardGameGeek asking for an iOS version.

The key to moving that discussion past arguments for value capture is to recognize that different expectations create room for a compensation split that both parties prefer to any flat fee they might agree.

To keep things simple, let’s assume a price of $3.99 (before Apple’s 30% commission) and volume forecasts as follows:

Developer’s Guess
Chad’s Guess
Total Sales (unit)
Total Revenue ($)
Each party’s share

(Total sales is calculated by multiplying each percentage estimate by that amount of sales, i.e. the Developer’s total sales is 70% of 1,000 plus 15% of 4,000 plus 15% of 10,000.)  N.B. These are not my assumptions and I have no idea what assumptions the developer may have -- I put them up here purely for the sake of discussion!

The developer’s average estimate puts total dollar sales at less than $10K and he thinks there’s a 70% chance that sales will only be about 1,000 units, or less than $3,000.    I think sales will be nearly $20K and my most likely scenario is a “hit” with 10K unit sales and about $28,000 for us to split.

Given this variance, the last thing we should try to do is agree a single fixed split.  That’s could break the deal altogether and is likely to lead to resentment down the road when one of us turns out to be right.  If I persuade the developer to take 50% and he ends up getting paid $1,400 it’s a turkey for him; similarly if sales are close to $30,000 and he got 70% of it I’d wish I’d never agreed to a revenue share deal.

A simple alternative is a split that changes as sales grow.  Suppose the developer gets 100% of the first 1,000 sales, 75% of the next 1,000 and 25% of sales beyond that.  The developer would expect to receive $4,480 (on average), an increase of 15%, and his risk is much lower.  If he’s right that the most likely outcome is only 1,000 units at least he gets 100% of the sales.  I also expect to receive about 16% more this way, albeit with a higher risk of getting nothing.

Some would argue against calling this value creation since eventually one of us will be proved wrong.  Both parties might like this deal better up front but in the future one will be better off and one worse off by exactly the same amount.  I would counter with two points.  First, being able to “bet” according to one’s beliefs is generally regarded as valuable.  Investors buy and sell stocks, companies invest in markets and technologies, collectors buy stamps, coins and other items based in part on their estimation of what the future will bring.  Enabling two parties to act according to their expectations of the future is a valid goal of negotiation.

More fundamentally, while one of us will be worse off we’re more likely to be satisfied with the variable commission because the result will be much closer to what we would have agreed if we knew the sales up front.  If unit sales are just 1,000 I won’t feel bad that the developer gets all the revenue…and if I expected sales to be that modest I would view the app as a nice promotional item rather than a source of revenue.  Similarly, if app sales are 10,000 units can the developer really feel badly about only getting around $10K?  Not only do sales of that level imply that the game (my contribution) was highly valuable, but if we knew that sales would be that high I could contract to have the app done by a professional firm for that amount.

How to Introduce Issues

Introducing issues is generally trivial with experienced, sophisticated negotiators.  They’re eager to include as many important issues as possible for the same reason you are – they know that’s a powerful way to maximize the deal’s final value.  The challenge arises when your counterpart doesn’t have this perspective.  He may see an attempt to introduce issues as some sort of gamesmanship or a waste of time.  Perhaps worse, he may see it as a request by you that merits a reciprocal concession – that is, your effort to introduce value creation may be met by a value capture effort by the other side.

I’ve found the best approach is to be up front and open about why you want to add issues and then to share at least top-level thoughts about where your interests lie (and where you think your counterpart’s may lie) on each topic.  For example, here’s an excerpt from the email I sent to the developer on the subject:

If we can put haggling aside for a moment, however, I'd like to make sure that whatever agreement we reach is as valuable to us in total as possible.  As you may guess from my signature, I'm a negotiations geek in addition to being a game geek.  This doesn't mean that I'm a particularly hardball negotiator but rather that I'd like us to make sure that we aren't forgetting any "levers" in our agreement that might make the final deal more valuable to each of us.  In addition to royalty split, I can think of some other terms we should discuss and I'll share my thoughts about my own interests on them.

I start off by establishing this as separate from haggling over price.  This helps avoid the “OK, we can discuss these things but you’ll have to do better on price” trap and makes it clear that adding issues is a mutual and collaborative process of value creation.  I finish by offering my thoughts, which then invites him to reciprocate with his.

Value Creation in Parallel with Value Capture

A key consideration through all this is that adding issues (along with other value-creation techniques) don’t replace value capture efforts.  They happen in parallel.  While raising other issues with the developer, I also expressed my view that his initial offer of a 70/30 split was too high.  I’m also addressing his implicit anchor and working on my BATNA by getting quotes from iOS development firms for a cash-only development deal.  Although I want my counterpart to be happy with our final deal, emphasizing value creation in no way means that I have to give anything away.  It just means that when we finally split up the figurative pile of money on the table, we’ve got as big a pile as possible.

Monday, August 8, 2011

Your Bad BATNA is Your Own Fault

Everyone, even professional negotiators, sometimes find themselves forced to negotiate with a bad BATNA.  In another post I looked at some of the ways you can still negotiate effectively with a bad BATNA but today I want to focus on the fact that our BATNAs, good and bad, are generally self-created. I'll go through some examples and then talk about how you can develop good BATNAs long before you even know you're going to be negotiating.

Walking the Dog

We recently got a puppy.  In order to keep the option of day and half-day trips to places Jordan can't come we found two local companies that offer dog-walking and "doggy day care" services.  One of the companies was well-established and the owner was a friend of a friend.  We knew several people who used her service.  The other was a relative newcomer.

I met first with the owner of the established company.  We had a good conversation about how her service worked and what they offered, and she a contract to fill out at my convenience if I wanted to hire them.  I always think you can learn a lot about a firm by the contracts they offer, so I skimmed through it as we talked and found several "interesting" terms.

As you can imagine, one problem a dog walking service has is preventing "poaching" by walkers.  The firm hires and trains young workers and then introduces them to clients, in exchange for which it keeps a significant cut of the walking fees.  The value provided by the firm is largely up front -- once the walker is trained and has met the client the client doesn't lose much by paying the walker directly so they split what the agency would normally get.  In order to avoid this problem, the owner had come up with a creative solution -- a $1,500 "referral" fee the client would pay if he or she hired one of the agency's walkers other than through the agency.

Another problem she must have had was collecting, and she had another solution -- a clause that said that the client was responsible for any legal fees the agency incurred in order to enforce the terms of the contract.

It's likely that if we hired that agency nothing would ever go wrong.  I have no interest in poaching and I pay my bills.  But disputes do happen, even in relationships where they seem unlikely.  When you sign a contract you need to ask how the terms of that contract will affect the resolution of disputes in the event that they do happen?

Let's say a couple of years from now I'm taking Jordan for a walk and we meet a college kid who mentions that he walks dogs to make some extra money.  Unbeknownst to me, he once worked for this agency -- or perhaps he actually walked Jordan once but I forgot.  The owner thinks that we're "poaching" and demands her fee.  Can I negotiate?  Not very effectively -- instead of a mutually unpleasant BATNA of going to court, the pain of court is all on my side because I'd have to pay her expenses.  She could even make my BATNA worse by mentioning that if we do go to court her attorney happens to be from a premium firm (i.e. very expensive for me)!

Inability to Tolerate a "Normal" BATNA

In discussing how to negotiate with a bad BATNA I mentioned my work with a healthcare non-profit that wanted to negotiate higher reimbursement rates.  One of our biggest challenges in those negotiations was that the organization's balance sheet was weak, meaning it couldn't easily afford the lost revenue that would come from temporarily refusing to accept any of the major insurers.  Put simply, their BATNA was weak and for purely internal reasons.

Sometimes it's not that your BATNA is all that bad but that you're unable to bear the costs of a moderately bad BATNA that is normal for the negotiation you face.  If a union can't afford to strike, a person can't afford to go to court or a client can't afford to change providers for a key service then their negotiation position is weak for potentially unnecessary reasons.

Creating Our Own Problem

All too often we create our own bad BATNA out of whole cloth.

For a vendor, two of the most beautiful words in the world are "switching costs".  Switching costs are expenses a customer needs to incur in order to change service providers, and companies work hard to build them in.  Some switching costs are natural.  If your company contracts out key services it's likely that any new provider would require at least a learning period before they provided as good a service.  In some areas, particularly involving software or services that are tightly intertwined with one's core business, switching costs can become prohibitive.  In extreme cases, a company can have no viable alternatives at any bearable cost.

Switching costs occur regularly in day-to-day life as well.  Loyalty programs are designed to create a benefit (e.g. miles) that loses some or all of its value if we don't continue to buy from that company.  Switching costs can be non-financial, too. My main email address, that I've used for many years, ends in the name of my cable company rather than and as a result there would be a non-trivial switching cost for me if we ever decide to change cable providers.

It's not realistic to avoid all switching costs, but excessive switching costs can be disastrous, especially if protective measures (e.g. a maximum annual fee increase) aren't built in.  If you can't afford to switch you have a terrible BATNA in any negotiation about pricing.

Avoiding Bad BATNAs

To reduce the number of situations in which you have a bad BATNA, the following steps are useful:

  1. Get used to thinking about the potential for disagreement.  Most people dislike conflict and in particular most people dislike thinking about future conflict at the start of a relationship.  At the same time, we recognize that disagreement, conflict and divergent interests exist in even the healthiest relationships.  (Just ask any successful business partners or happily-married couple!)  Get in the habit of asking yourself what the context for future disagreements is likely to be.  If it's tilted against you, the start of the relationship is the right time to act.
  2. Look forward, reason back.  At the start of any long-term relationship, imagine yourself (and, more importantly perhaps, your counterpart) in the future.  Does one of you have undue leverage over the other?  Are your interests diverging?  In my post about negotiating with a bad BATNA I gave the example of a homeowner who had some custom moulding made for his house and found himself with a bad BATNA when he wanted a small amount of additional work done using the same moulding.  He could easily have avoided that if he'd asked, "What will happen if it turns out I want some more work done?" and asked the contractor to agree at the start of the original job that any future work using that moulding would be done at a comparable rate.
  3. Build trust, social capital and mutual need.  There is a world of difference between negotiating with someone where you need them and where you need each other.  Similarly, integrated organizations can avoid painful games of brinkmanship if they develop social capital and norms that make such behavior unacceptable.  This is generally even more effective with individuals.
  4. Develop your BATNA before you need it.  If you're in the habit of looking forward and exploring the potential for future disagreements you can prepare for those disagreements by strengthening your BATNA so that if and when they arise you're in solid shape.  As a by-product, your strong BATNA may prevent disagreements from occurring at all, since your counterpart's incentive to force a renegotiation may not exist absent your weakness.

Above all, remember that your BATNA isn't something you're stuck with.  You can affect both how it starts and how it develops over time.

Friday, August 5, 2011

How to Create a Scoring Matrix

After my last post I got an email from a reader who said that a scoring matrix sounded like it could be really useful but that she had no idea from what I wrote how she should go about making one.  This post will walk through the creation of a simple scoring matrix and then discuss some practical ways one might use it beyond evaluating competing offers.

Let's take a common life decision, leaving one job for a new one.  In our hypothetical case, Jane has a job she's happy with but after exploring the market she identified some other interesting opportunities.  After pursuing them she's in final interviews with two other companies (we'll call them A and B) and is doing her homework to negotiate in the event any offers come through.

The first step is to set her "utility currency".  This is what you'll use to score each component of your offer.  For this exercise we're going to use money, but for someone for whom money wasn't a key motivator that might be the wrong way to go -- you could use generic points instead.

Jane currently has a salary of $100,000 and last year she earned $20K in performance bonuses.  She's having a good year this year and expects a bonus closer to $35K and a raise to $110K, so her BATNA has a compensation of around $145K.  She'll evaluate competing offers accordingly.

Jane identifies the following criteria as important to her decision:

  • Growth potential.  Jane's current company is large but it's also a family business.  Some non-family members have reached high levels within the organization but Jane is concerned that she may not be able to accomplish as much as she'd like there.  Company A offers more normal growth potential and Company B is particularly exciting because she'd be reporting to a woman who was a longtime client and who has an excellent reputation as a career mentor.  Jane estimates that the improved growth prospects are equal in value to $20K in salary at Company A and $40K in salary at Company B.
  • Commute.  Jane's commute currently takes over an hour each way and combined with long work hours has been taking a toll on her outside life.  Company A is only ten minutes away; Company B, unfortunately, is almost an hour and a half away.  Jane estimates the commute difference as worth $30K in salary at Company A and -$10K in salary at Company B.
  • Vacation time.  Jane's family has two traditional events each year, both of which take up a week of vacation.  Jane is also passionate about travel and really values the five weeks of vacation she has at her current job.  The HR representatives at A and B have told her that new employees only get three weeks of vacation.  Jane is very concerned about this; she would rate four weeks salary as -$15,000 and three weeks is almost a deal-breaker.  She reluctantly puts it as -$50,000.
A real job evaluation would likely consider more -- possibly many more -- factors than this.  How many you include depends on personal preference but you should try to include all significant factors.  Once you have a list and start scoring them you may decide to drop some.  For example, Jane might have included "quality of nearby lunch options" and then realized that this only scored at $500 which would be rounding error compared with other factors.

Jane's scoring matrix now looks like this:

Company A
Company B
Offer - $145K
Offer -$145K
Offer - $145K
Offer - $165K

The total line is how much better (or worse) an offer is than staying put.  From this Jane can tell that any offer above $145K from Company A or above $165K from Company B has a positive score and thus seems better than her BATNA of staying where she is.  More importantly it's a useful tool for negotiating with her new company and for creative problem-solving on her own.

Let's start with the commute.  Once Jane has it down on paper that she'd give up $30K in salary to shorten her commute she might think about moving.  There aren't any places she'd like to live that are closer to her current company but there's an attractive neighborhood close to Company B.  Rent is likely to be $15,000 a month higher but that's still a net improvement of $15,000 ($30K value of short commute less $15K in higher rent) instead of the negative $10,000 if she were to stay where she is.  Thinking through her priorities thus led her to realize that what looked like a drawback could be converted into a positive.

Fully understanding her interests might lead Jane to improve her BATNA as well.  Either before or after she gets an offer, Jane may decide to talk to her manager and explain that limited career prospects are playing a significant role in her considering a different firm.  It's possible that they'll want to keep her and will find a way to give her confidence that she can advance there.

Then there's vacation.  Jane really cares about her travel time; going from three weeks to four is worth the equivalent of $35,000 in salary to her and a fifth week is worth another $15,000!  Recognizing that should guide her in her negotiation but it also suggests a possible solution if the three week policy is indeed firm -- her contract could include two weeks of unpaid leave.  If her compensation is roughly $3K per week, then taking two weeks of unpaid time costs her $6K compared with her current job, which is much better than $50K.

Assuming Jane gets an offer from Company B, this problem solving suggests a different scoring matrix:

Company B (original)
Company B (revised)
Offer - $145K
Offer -$145K
-$6K (assuming 2 unpaid wks)
Offer - $145K
Offer - $84K

One last point.  A scoring matrix is only as accurate as the assumptions used to make it.  It should be reviewed and challenged and you should never forget that it's a tool.  For example, if Jane gets an offer of $90K her matrix gives that a positive score but such a significant drop in salary might indicate that her new firm sees her playing a less important role than she anticipates.